COMPETITION TRIBUNAL OF SOUTH AFRICA
Case No.:IR198Mar23
In the interim relief application between:
Apollo Studios (Pty) Ltd First Applicant
Motomatix (Pty) Ltd Second Applicant
And
Audatex SA (Pty) Ltd First Respondent
The Competition Commission Second Respondent
INTRODUCTION
[1] This matter concerns an application for interim relief in terms of section
49C of the Competition Act, 89 of 1998 (the “Act”). The applicants, Apollo
Studios (Pty) Ltd (“Apollo”) and Motomatix (Pty) Ltd (“Motomatix”), contend
that the termination by the f irst respondent, Audatex SA (Pty) Ltd
(”Audatex”), of the access that Apollo’s “PartSmart” software application
Panel : J Wilson (Presiding Member)
: M Mazwai (Tribunal Member)
: L Mncube (Tribunal Member)
Heard on : 12 April 2023
Order issued on : 14 April 2023
Reasons issued on : 08 May 2023
REASONS FOR DECISION
2
(“app”) had to Audatex’s online software platform constitutes a
contravention of section 8(1)(d)(ii), alternatively section 8(1)(c), of the Act.1
[2] The applicants accordingly sought an order interdicting Audatex from
carrying out its intended “offboarding” of the PartSmart app from the
Audatex platform pending the final determination of a complaint filed by
the applicants with the Competition Commission (“the Commission”), or a
period of six months, whichever occur red first. The substantive relief
initially sought by the applicants in their notice of motion was the following:
“2 That the first respondent (“Audatex”) is interdicted and
restrained from terminating the access of ‘PartSmart’ (owned
and operated by the first applicant) to the “Audatex platform”
(as described in the founding affidavit) pending the
conclusion of a hearing into the alleged prohibited practices
that are the subject of the applicants’ complaint to the second
respondent which is attached to the applicants’ founding
affidavit marked “FA1”, alternatively six months from the date
of this order, whichever occurs first.
3 In the alternative to paragraph 2 above, int erdicting and
restraining Audatex from terminating the access of PartSmart
to the Audatex platform pending the further hearing of the
application on a date and in accordance with a timetable to
be directed by the Competition Tribunal at which hearing the
applicants will seek relief in terms of paragraph 2 above.
1 The applicants also alleged contraventions of various other sections of the Act, but persisted at
the hearing only in their complaints under sections 8(1)(d)(ii) and section 8(1)(c) of the Act.
3
4 That the first respondent is ordered to pay the costs of the
application; alternatively that any party opposing the relief
sought is ordered to pay the costs occasioned by such
opposition.”
[3] We decided, after hearing the parties, not to grant the interim relief sought
by the applicants and issued an order dismissing the application on 14
April 2023. The reasons for our decision are set out below.
PROCEDURAL BACKGROUND
[4] On 20 March 2023, the appli cants submitted a complaint to the
Commission in respect of Audatex’s threatened “offboarding” of the
PartSmart app from Audatex’s platform with effect from 31 March 2023 ,
and, on 22 March 2023, they filed an application for urgent interim relief
with the Tribunal.
[5] The Tribunal convened an urgent pre -hearing on 27 March 2023 to
regulate the further process in the application. At the pre -hearing, the
Commission stated that it would not participate in the application
proceedings and would abide by the Tribunal’s decision. However,
Audatex stated that it intended to oppose the application and to file an
answering affidavit. Having regard to the dates available to the Tribunal
to hear the application, and the time required by the par ties to file their
answering and replying papers respectively, it was ultimately agreed by
the applicants and Audatex, and directed by the Tribunal, that the interim
4
relief application would be heard on 12 April 2023 subject to a without
prejudice undertaking by Audatex not to terminate PartSmart ’s access to
its platform “until midnight on 14 April 2023 or such earlier date and time
as the Tribunal issues an order dismissing the application (if applicable) ”.
Directions were also given for the filing of ans wering and replying
affidavits, and heads of argument, by the parties.
[6] As a result, the “interim interim” alternative relief sought in paragraph 3 of
the applicants’ notice of motion fell away , and what remained for
determination was the primary relief sought in paragraph 2 of the notice of
motion.
[7] Audatex filed an answering affidavit on 4 April and the applicants filed a
replying affidavit on 7 April 2023. Thereafter, on 10 April, Audatex filed an
application for leave to file a rejoinder affi davit on the grounds that the
replying affidavit contained new factual material as well as allegations of
dishonesty that required a response. On 11 April, the applicants filed an
application for leave to file a further affidavit of their own, in which the y
sought to respond to certain allegations in Audatex’s rejoin der affidavit,
and also tendered confidentiality undertakings in respect of information on
the Audatex platform to which they might have access.
[8] The hearing of the interim relief application proceeded on 12 April 2023 .
At the commencement of the hearing, the applicants and Audatex
5
indicated that they did not oppose each other’s applications for leave to file
further affidavits in the proceedings, and those further affidavits were
accordingly admitted into evidence before the Tribunal. After hearing the
parties, and considering the matter, the Tribunal issued its order
dismissing the application on 14 April 2023.
[9] After the issuance of the Tribunal’s order, the applicants directed a request
to the Tribunal to provide our reasons as soon as possible because they
wish to consider the merits of an appeal. T hese reasons are accordingly
provided on an expedited basis.
RELEVANT FACTUAL BACKGROUND
The parties and their businesses
Apollo and PartSmart
[10] Apollo is a software development company that was established in 2020.
It is the owner of the PartSmart app , which is automotive parts
procurement software that was developed in 2018 by the co -founder of
Apollo in collaboration with [Name].
[11] PartSmart automates and facilitates procurement of replacement
automotive parts for insurance repairs. It enables insurers and automotive
repair companies to obtain competing quotes for, and to place orders, for
6
automotive parts supplied by Original Equipment Manufacturers (“OEMs”)
and third party parts suppliers.
[12] PartSmart competes with various other parts procurement software
providers, including [Competitor Names]. Audatex also lists [Competitor
Names] as competitors of PartSmart.
[13] At the time of the applicant’s complaint and interim relief application,
PartSmart was integrated with Audatex’s online software platform, as
discussed further below . As discussed further below, [Competitor
Names] were also integrated with Audatex’s platform shortly before
Audatex removed PartSmart’s access to its platform.
[14] PartSmart contracts only with insurers, and currently has [Number]
customers, namely [Customer Names].
Audatex
[15] Audatex is a subsidiary of a glo bal group of companies, and has been
operating in South Africa since the 1980’s.
[16] Audatex owns and operates an online software platform that facilitates
electronic vehicle damage estimation and processing of automotive
insurance claims. Audatex describes its business model as follows:
“The Audatex offering enables vehicle repairers to generate
comprehensive repair estimates, taking account of parts
7
manufacturing prices, part numbers, paint material and labour rates.
Once an estimate has been generated, the system facilitates
interaction between insurers, claims assessors and vehicle
repairers. The system allows insurers, assessors and repairers to
interact so as to effectively reach commercial outcomes relating to
repairs to be undertaken on a vehicle.”
[17] Audatex’s customers include insurers, repairers and third party service
providers, including parts procurement providers such as PartSmart.
[18] It appears from the papers that, where the relevant insurer is an Audatex
customer, the vehicle damage claim process is basically the following:
18.1 First, the insured informs its insurer that it requires repair work to
be undertaken on its vehicle.
18.2 The insurer dispatches an affiliated asse ssor to assess the
damage to the vehicle . The assessor feeds into the Audatex
system the vehicle details, and the areas on the vehicle that are
damaged. The Audatex system then produce s an estimation
report which contains details of the specific parts req uired and
the part identification number of each such part, the estimated
cost per part, and the estimated cost for labour and paint.
18.3 That estimation report is then accessible on the Audatex platform
to insurers and to third party service providers on the platform,
8
including parts procurement service providers such as PartSmart
and its competitors.
18.4 If the insurer has a contract with a parts procurement software
provider that is integrated with the Audatex platform, that provider
will then, on the basis of the estimation report, source prices for,
and confirm availability of, the parts required, and present those
quotes to insurers for their consideration and authorisation, all via
the platform.
18.5 The authorised claim instructions are then delivered to an
automotive repairer for the repair of the vehicle with the parts,
and at the prices, authorised by the insurer.
[19] Audatex states that there are two distinct, albeit related, aspects to its
operations. The first aspect, which Audatex refers to as its “ core value
offering”, is an integrated on -line vehicle damage estimating solution to
insurers, repairers and assessors , which generates initial estimates for
repairs to vehicles, primarily pursuant to insurance claims. These
estimates a re generated off the back of an embedded database
comprising information identifying all the standard parts in more than 1,300
vehicle models.
[20] Audatex stresses the proprietary nature of its parts database. Audatex
says it has, with significant investment over many years, compiled this
9
database by identifying the parts specific to each vehicle model, their
manufacturer details and manufacturer pricing per part. The database has
been developed based on information that is available from vehicle
manufacturers, industry research and other investments (such as
personnel, software development and the like). The database also has to
be continuously maintained and updated, and also localised to the
conditions of the particular countries in which the Audatex solut ion is
offered.
[21] The second aspect of Audatex’s offering is integrated software which
automates certain business processes of, and integrates with, participants
in the vehicle repair value chain. This enables insurance companies,
accident management compan ies, vehicle manufacturers , parts
procurement service providers and repairers on the Audatex platform to
manage the claims and authorisation process in an automated and
integrated manner.
Relationship between Audatex and PartSmart
[22] The applicants contend t hat a critical feature in the development of
PartSmart was its access, via an application programming interface (“API”)
and user accounts created for PartSmart by Audatex, to the Audatex
platform.
10
[23] The applicants contend that, if an insurer uses the Audatex platform
exclusively (as is overwhelmingly the case) for claims processing, then
PartSmart has no other source for this data in real time, meaning that its
ability to service its insurer customers and to compete in the market is
severely constrained.
Motomatix
[24] Motomatix was founded in 2022, and is a wholly -owned subsidiary of
Apollo. In August 2022, Motomatix was appointed as the exclusive
distributor of GT Motive solutions in South Africa. It appears from the press
release announcing that appointment that GT Motive is an international
supplier of repair data and solutions, and was previously represented in
South Africa (since 2017, according to Audatex) by another distributor
called Estimatic Management Solutions. It is common cause that the GT
Motive platform is a direct competitor to the Audatex platform
internationally, and in South Africa.
[25] It appears that the GT Motive platform has, so far, made limited headway
in the market. According to Audatex, the GT Motive platform has only
[Number] customers together with certain firms that have vehicle fleets
and do their own repairs, such as [Customer Name].
[26] However, the applicants anticipate that GT Motive will become more
competitive under licence to Motmatix. According to the press release
11
referred to above, the agreement between Motomatix and GT Motive
involves “ a large investment by Motomatix to elevate the data, model
coverage and technology to greater levels which will support the growth of
the business in the upcoming years ”. The applicants also state in their
founding affidavit that Motomatix now presents a “viable alternative” to the
Audatex platform in the South African market. Audatex’s internal strategic
documents confirm that it regards GT Motive as a competitive threat in
South Africa.
[27] As discussed below, the competitive relationship between Audatex and GT
Motive lies at the heart of the dispute between the parties.
Events leading up to Audatex’s decision to “offboard” PartSmart
[28] In July 2020 (approximately two years after PartSmart had first integrated
with Audatex), Apollo and Audatex concluded an "Audatex SA Reselling
(Partner Product) Agreement (the “reseller agreement”) for a n initial term
of three years. In terms of that agreement, Apollo granted Audatex the
right to incorporate PartSmart into Audatex's offering to Audatex's
customers, and thus to act as a reseller of the PartSmart product on a
revenue-sharing basis. Whilst the licence granted to Audatex was non -
exclusive, clause 2.4 provided that approval from Audatex was required
before PartSmart could be implemented with any short -term insurer in
South Africa.
12
[29] Audatex contends that, pursuant to th e reseller agreement, Apollo
obtained preferential, and largely unfettered, access to Audatex's system,
including aspects of its proprietary parts database , and to pull that
information from the system.
[30] In August 2022, Motomatix announced that it had acquired the licence to
the GT Motive software in South Africa, and this led to the termination of
the reseller agreement by mutual agreement . The termination of the
reseller agreement was memorialised in a letter agreement ultimately
signed by both Audatex and Apollo dated 11 October 2022. In that letter,
it was recorded, inter alia, that:
30.1 The reseller agreement was terminated by mutual
agreement with effect from 30 September 2022 (clause
2.1).
30.2 Following termination of the reseller agreement:
30.2.1 Audatex would cease to be a reseller of the PartSmart
software (clause 2.2.1).
30.2.2 Apollo and Audatex would enter into a new agreement,
effective 1 October 202 2, in terms of which Audatex
transferred to Apollo all of Audatex’s obligations to provide
PartSmart to [Number] “legacy clients”, namely [Customer
Names] (subject to a 12 month exit period insofar as such
13
substitution could not be effected in relation to any legacy
client for any reason) (clauses 2.2.2 to 2.2.4).
30.2.3 Apollo would be entitled to provide PartSmart directly to
clients, including short-term insurers (clause 2.2.5).
30.2.4 “Insofar as clients who utilise the PartSmart software already
use same in integrat ion with Audatex’s claims estimation
software (“the Audatex Software”), or insofar as a client
requests integration of PartSmart with the Audatex Software,
Audatex shall be entitled (but not obliged) to charge such
client an integration fee per claim, or otherwise as Audatex
considers appropriate” (clause 2.2.7).
30.2.5 Nothing would preclude Apollo from “permitting integration of
PartSmart with any claims estimation software which
performs a function similar to that performed by [Audatex] ”
(clause 2.2.8).
[31] Notwithstanding the terms of this agreement, Audatex’s contends that it
only fully appreciated the closeness of the working relationship between
Apollo and Motomatix when it received an email from Motomatix on 12
January 2023 seeking certain informati on relating to Audatex’s parts
database. Audatex says that it was immediately alarmed by the apparently
close relationship between Apollo and Motomatix, and therefore decided
14
that, as of 31 March 2023, it would no longer grant PartSmart access to
the Audatex system.
The applicants’ complaint and Audatex’s response
[32] It is common cause that Audatex’s decision to offboard PartSmart from its
platform was a consequence of Apollo’s subsidiary, Motomatix, becoming
the exclusive distributor in South Africa of GT Motive, a direct competitor
of Audatex. However, the underlying rationale for this move lies at the
core of the parties’ dispute before the Tribunal.
[33] The applicants contend that Audatex’s conduct is intended to foreclose
PartSmart from competing effe ctively as a supplier of parts procurement
software; and, indirectly, also to impede the ability of Motomatix’s GT
Motive system to compete with the Audatex system.
[34] The applicants argue that Audatex is a dominant firm for purposes of
section 7 of the Act with a market share of over 80% in “the market for the
provision of software for the management and processing of insurance
claims in the automotive repair industry ”; and that Audatex’s decision to
offboard PartSmart from its platform constitutes an abus e of dominance
under section 8(1)(d)(ii), alternatively section 8(1)(c), of the Act.
[35] Audatex denies this. It contends that it is simply seeking to prevent a direct
competitor, Motomatix, from continuing to have access to proprietary
information on its p latform through PartSmart , and from using that
15
information to compete unlawfully against Audatex. Audatex alleges that
PartSmart does not require access to the Audatex platform to compete
effectively with its rivals, and also denies that the offboarding of PartSmart
from the Audatex platform will have any negative effect on Motomatix.
[36] We proceed to assess these respective claims under section 49C and the
relevant prohibited practice sections of the Act.
THE REQUIREMENTS FOR INTERIM RELIEF: SECTION 49C
[37] The grant of interim relief under the Act is governed by section 49C of the
Act. Section 49C(2)(b) states that the Tribunal:
“may grant interim relief if it is reasonable and just to do so, having
regard to the following factors:
(i) the evidence relating to the alleged prohibited practice;
(ii) the need to prevent serious or irreparable damage to the
applicant; and
(iii) the balance of convenience.”
[38] Section 49C(3) explains that:
“In any proceedings in terms of this section, the standard of proof is
the same as the standard of proof in a High Court on a common law
application for an interim interdict.”
16
[39] Following the well -known common law appr oach set out in Webster and
Mitchell2 as qualified in Gool3,4 the Tribunal set out its approach in applying
section 49C(2) as follows in York Timbers:5
“[W]e must first establish if there is evidence of a prohibited practice,
which is the Act’s analogue of a prima facie right. We do this by
taking the facts alleged by the applicant, together with the facts
alleged by the respondent that the applicant cannot dispute, and
consider whether having regard to the inherent probabilities, the
applicant s hould on those facts establish the existence of a
prohibited practice at the hearing of the complaint referral.
If the applicant has succeeded in doing so we then consider the
‘doubt’ leg of the enquiry. Do the facts set out by the respondent in
contradiction of the applicant’s case raise serious doubt or do they
constitute mere contradiction or an unconvincing explanation. If
they do raise serious doubt the applicant cannot succeed.
As far as the remaining factors in [section 49C( 2)] are concerned,
viz irreparable damage and the balance of convenience, these are
not looked at in isolation or separately but are taken in conjunction
with one another when we determine our overall discretion.”
2 Webster v Mitchell 1948 (1) SA 1168 (W) at 1189.
3 Gool v Minister of Justice and Another 1955 (2) SA 682 (C) at 688E-F.
4 This approach was recently confirmed by the Constitutional Court in Eskom Holdings SOC Ltd
v Vaal River Development Association (Pty) Ltd and others (Case No. CCT44/22), 23 December
2022, at paras 247-248 and 253.
5 York Timbers Ltd v SA Forestry Company Ltd (15/IR/Feb01) [2001] ZACT 19 (9 May 2001) at
paras 64-66.
17
[40] In Natal Wholesale Chemists,6 the Tribunal pointed out that:
“[I]n terms of Section 49C(2), the Tribunal no longer has to consider
whether each of the requirements has been established in isolation,
but rather looks at all the factors listed in Section 49C(2) as a whole
to see whether a case for interim relief has been established. . . .
Section 49C(2) follows the approach at common law as applied by
Appellate Division in the case of Eriksen Motors (Welkom) Ltd v
Protea Motors, Warrenton 1973 (3) 685 (A) . The court held that in
deciding whether to exercise its discretion to grant interim relief the
court should not look at the prerequisites in isolation but should
consider all of them in conjunction with each other. The court went
on to state that these prerequisites
"... are not indiv idually decisive, but are interrelated, for
example, the stronger the applicant's prospects for success
the less the need to rely on prejudice to himself. Conversely,
the more the element of "some doubt", the greater the need
for the other factors to favour him."
[41] Guidance regarding the approach to be adopted in applications for interim
relief in terms of section 49C has been provided in the CAC’s recent
judgments in BCX7 and eMedia8.
6 Natal Wholesale Chemists v Astra Pharmaceuticals and others [2001] ZACT 7 (12 March 2001)
at para 34.
7 Business Connexion (Pty) Ltd v Vexall (Pty) Ltd and Another (182/CAC/Mar20) [2020] ZACAC
4 (15 July 2020).
8 eMedia Investments (Pty) Ltd v Multichoice (Pty) Ltd and another (201/CAC/JUN22) [2022]
ZACAC 9; [2022] 2 CPLR 23 (CAC) (1 August 2022).
18
[42] In BCX, the CAC emphasised that interim relief under the Act has certain
features that are distinct from interim relief at common law:
“Two features of the power to grant interim relief have particular
salience. First, the prohibited practices in chapter two of the Act are
concerned with practices that affect markets, a market or a segment
of the market. Unlike disputes in private law which, for the most part,
concern the rights enjoyed and duties owed by individuals to one
another, prohibited practices in chapter 2 concer n the conduct of
firms and their effect on competition in the market. Even those
practices that are not defined by reference to their effects are
nevertheless rendered unlawful by reason of their presumptive
harmful effects upon competition. As a result, interim relief granted
by the Tribunal has effects upon the state of competition in the
market. Second, when the Tribunal grants an interim relief order, it
is not a status quo order. The order requires that the respondent
firm desist from the prohibited pr actice (in whole or in part). The
purpose of the order is to alter the competitive relationship between
firms in the market. If the interim order is to be effective, it is
intended to permit of competition taking place in the market that has
hitherto not taken place. That may have effects within a market or
across markets, and may affect different market participants:
customers, competitors and suppliers. When the Tribunal grants an
interim order it alters the status quo in the market and is intended to
change the way firms compete in the market, with consequences
that may well resonate within and between markets.
An interim relief order under the Act does not provide a remedy to
permit a person claiming a right to enjoy the exercise of that right
until the right is finally determined. Rather, the Tribunal is
19
empowered to regulate how competition in the market is to take
place for a six or twelve month period. That is a different
competence to that of a court adjudicating a dispute of right; it is a
regulatory competence to decide whether the state of competition
in the market must endure, notwithstanding the evidence that a
prohibited practice is taking place, or whether the Tribunal should
order a change.”9
[43] The CAC proceeded, in the light of the se observations, to explain the
approach to be adopted in respect of each of the three factors in section
49C(2) that are relevant to the grant of interim relief:
“The evidence of a prohibited practice, as I have sought to explain,
is not concerned with the rights of the applicant but the competitive
position of competitors in the market, judged against the regulatory
criteria of the prohibited practices defined in chapter 2 of the Act.
The need to prevent serious or irreparable damage to the applicant
posits an enquiry into the effects of the alleged prohibited practice
upon the applicant and it is for this reason a party specific enquiry.
However, here too the analogue of interim interdicts as an equitable
remedy at common law must be approached with care. The
common law remedy asks what well -grounded apprehension of
irreparable harm will be suffered by the applicant if interim relief is
not granted and the applicant succeeds in proving the right, now
prima facie established. This concerns an interference with an
applicant’s rights and the harm that may be suffered by an applicant
as a result of such interference until the court can finally determine
the question of rights. Interim relief under s49C requires an enquiry
9 BCX, supra, at paras 17-18.
20
that is similarly structured, but distinct in a number of respects. The
need for intervention is a function of the probability of serious or
irreparable damage occurring, if no intervention is ordered by the
Tribunal before it can make a final determination as to whether the
alleged prohibited practice has taken place. It is the damage to the
competitive position of the applicant that the prohibited practice may
cause that marks out this enquiry. Other forms of damage to the
applicant are not relevant because the Act’s purpose is to maintain
and promote competition in the market.
Finally, the balance of convenience in s49C is a direct borrowing
from the common law. It weighs the prejudice the applicant will
suffer if the interim interdict is not granted against the prejudice to
the respondent if it is granted. This requires an equitable reckoning
as to who bears the greater burden of error. If the interim order is
granted and no case is ultimately established to prove the alleged
prohibited practice, what prejudice will have been suffered by the
respondent, and how might that prejudice be mitigated? So too, if
the interim order is refused and the prohibited practice is ultimately
proven, what prejudice will the applicant suffer in the interim. Here
too, the currency of prejudice is reckoned by recourse t o the
consequences for the competitive positioning of the parties in the
market. A respondent that is required to desist from conduct that
gives it a legitimate competitive advantage suffers prejudice. An
applicant that is required to endure an unlawful co mpetitive
disadvantage also suffers prejudice. How to weigh prejudice in the
balance is a difficult task. Hence the warranted caution with which
21
the Tribunal and this court have approached the exercise of the
power to grant an interim interdict.” 10
[44] In eMedia, the CAC explained that, in considering the three factors listed
in section 49C(2), consideration must be given to whether there are “clear,
non-speculative and uncontroversial facts” that support the grant of interim
relief, and that “whilst there will inevitably be disputes of fact”, that should
not prevent the Tribunal from taking a “ robust approach” on the evidence
before it.11
[45] The CAC also added that the basis of the power to grant interim relief must
be contextualised within the jurisprudential framework of the Act. The CAC
referred in this regard to the preamble to the Act, and to its purposes as
set out in section 2 .12 The CAC also referred to the following passages
from the Constitutional Court’s recent judgment in Mediclinic:
“It oug ht never to be acceptable for any of us, including the
corporate citizens of this land, to indulge, talk less of over -indulge,
in the unconscionable practice of seeking to record the highest profit
margin possible by any means necessary, in wanton disregar d for
what that would do to the rest of humanity. Neither should the
historic exclusion of some from meaningful participation, particularly
in the mainstream economy, be normalised. For, this seems to be
one of the most stubborn injustices of our past that require a more
10 BCX, supra, at paras 20-22.
11 eMedia, supra, at paras 80-81.
12 eMedia, supra, at paras 82-83.
22
deliberate, intentional and systematic confrontation appropriately
enabled by independent, incorruptible, efficient and effective law
enforcement and justice-dispensing institutions.”
and
“Institutions created to breathe life into these critical provisions of
the Act must therefore never allow what the Act exists to undo and
to do, to somehow elude them in their decision -making process.
The equalisation and enhancement of opportunities to enter the
mainstream economic space, to stay there and operate in an
environment that permits the previously excluded as well as small
and medium -sized enterprises to survive, succeed and compete
freely or favourably must always be allowed to enjoy their pre -
ordained and necessary pre -eminence. The legitimisation through
legal sophistry or some right-sounding and yet effectively inhibitive
jurisprudential innovations must be vigilantly guarded against and
deliberately flushed out of our justice and economic system.”13
[46] The CAC proceeded to explain that:
“It follows therefore that these are the guidelines this Court and
indeed the Tribunal must follow when applying the provisions of the
Competition Act. The approach calls for a transformative
constitutional approach and must be consistent with the scheme of
the Competition and apply a context -sensitive approach. This is a
striking feature that must be considered in this application. Unless
this transformative approach is applied even at an interim stage of
13 Competition Commission of South Africa v Mediclinic Southern Africa (Pty) Ltd and another
(CCT 31/20) [2021] ZACC 35 (15 October 2021) at paras 3 and 7, quoted in eMedia, supra, at
para 87.
23
proceedings, then the historical and insidious u nequal distribution
of wealth in South Africa will continue.” 14
[47] We have sought to apply this transformative and context-specific approach
in the present matter. However, given the particular facts of this case, we
have concluded that there are not “ clear, non -speculative and
uncontroversial facts” supporting the grant of the interim relief sought by
the applicants.
GENERAL LEGAL PRINCIPLES APPLICABLE TO THE INTERPRETATION
OF SECTION 8
[48] The applicants limited their case before the Tribunal to an alleged
contravention of section 8(1)(d)(ii), alternatively section 8(1)(c), of the Act.
[49] The requirements of these subsections fall to be understood in the context
of section 8(1) as a whole, which states, in material part, that it is prohibited
for a “dominant” firm to:
“(c) engage in an exclusionary act, other than an act listed in
paragraph (d), if the anti -competitive effect of that act
outweighs its technological, efficiency or other pro -
competitive gain; or
(d) engage in any of the following exclusionary acts, unless the
firm concerned can show technological, efficiency or other
14 eMedia, supra, at para 84.
24
pro-competitive gains which outweigh the anti -competitive
effect of its act –
(i) . . .
(ii) refusing to supply scarce goods or services to a
competitor or customer when supplying those goods
or services is economically feasible.”
[50] The definition of dominance is of course set out in section 7 of the Act. 15
An “exclusionary act” is defined (in its now expanded form) in section 1(1)
of the Act as “an act that impedes or prevents a firm entering in to,
participating in, or expanding within, a market ”. The term “ participate”
refers, in turn, to “ the ability of or opportunity for firms to sustain
themselves in the market”.
[51] The Tribunal set out its legal approach to the application of sections 8(1)(c)
and 8(1)(d) (then, sections 8(c) and 8(d)) of the Act in SAA,16 and that
approach has since been confirmed and elaborated upon in various cases,
15 Section 7 states that “A firm is dominant in a market if (a) it has at least 45% of that market; (b)
it has at least 35%, but less than 45%, of that market, unless it can show that it does not have
market power; or (c) it has less than 35% of that market, but has market power ”. “Market power”
is defined in section 1(1) as “ the power of a firm to control prices, to exclude competition or to
behave to an appreciable extent independently of its competitors, customers or suppliers ”.
16 Competition Commission and South African Airways (Pty) Ltd (18/CR/Mar01) [2005] ZACT 50;
[2020] 2 CPLR 821 (CT) (28 July 2005), at paras 132-137.
25
including the CAC’s judgments in Comair,17 Computicket18 and Uniplate19.
The relevant principles are therefore now well -established, and it is
unnecessary to repeat th em all here. For current purposes, it suffic es to
highlight the following.
[52] First, b oth section 8(1)(c) and section 8(1)(d) distinguish between an
“exclusionary act” (as defined above) and an “ anti-competitive effect ”.
Exclusionary conduct might give rise to an anti-competitive effect, but that
effect must be proven in order to establish a violation of those sections .20
As the CAC explained in Uniplate:
“[T]he exclusionary act of the firm is something separate from the
anti-competitive effect of that act. There must be a causal
relationship between the exclusionary act and its anti -competitive
effect. If the exclusionary act is taken as proof of its effect, the onus
resting upon the Commission will not have been discharged. That
would be a case based on conduct without regard to consequence,
and does not meet the requirements for liability under s 8(d).”21
[53] As noted above, the definition of an “exclusionary act” in the Act has now
been expanded to include impeding or preventing a firm from “participating
17 South African Airways (Pty) Ltd v Comair Ltd and Another (92/CAC/MAR10) [2011] ZACAC 3;
2012 (1) SA 20 (CAC) (11 April 2011).
18 Computicket (Pty) Ltd v Compe tition Commission of South Africa (170/CAC/Feb19) [2019]
ZACAC 4 (23 October 2019).
19 Uniplate Group (Pty) Ltd v The Competition Commission of South Africa (176/CAC/Jul19)
[2020] ZACAC 10 (25 February 2020).
20 SAA, supra, at paras 136-137; Computicket, supra, at paras 17 and 25-26.
21 Uniplate, supra, at para 23.
26
in” a market, in the sense of having the ability or opportunity to sustain
itself in the market.
[54] In addition, for the r easons explained in Mediclinic and eMedia, the
competition authorities must be particularly alive and sensitive to the anti-
competitive effects that may be felt in South African markets as a result of
exclusionary conduct by dominant firms, especially insofar as it is directed
at historically disadvantaged persons and small - and medium -sized
enterprises.
[55] Second, t he existence of an “ exclusionary act ” is assumed in section
8(1)(d)(i) if the conduct referred to in that section is established, whereas
that element must be proved by the complainant in section 8 (1)(c).22 We
consider further below the specific requirements of section 8(1)(d)(ii) of the
Act.
[56] Third, in section 8(1)(d), the requirement of a substantial anti -competitive
effect must be proven by the complainant, and is met either (i) if there is
“evidence of actual harm to consumer welfare ” or (ii) “ if the exclusionary
act is substantial or significant in terms of its effect in foreclosing the
market to rivals”.23
22 SAA, supra, at para 132; Computicket, supra, at para 16.
23 SAA, supra, at para 132; Comair, supra, at para 112; and Computicket, supra, at paras 18-19.
27
[57] In Computicket, the CAC emphasised that, in order to establish an “ anti-
competitive effect” for purposes of section 8(1)(d) , it must be shown that
the harm in question , “ whether in the form of actual or pote ntial harm,
strengthens the dominant firm’s position to the extent that competitive
rivalry is significantly impeded or is likely to be so impeded by the
impugned conduct of the dominant firm.” 24
[58] This again is an area where the competition authorities must be alive to
the role that smaller firms may play in promoting competition in
concentrated markets in South Africa. As the CAC pointed out in
Computicket:
“Rivalry may be diminished because a small firm plays an important
role in constraining the dominant firm in a part of the market,
whether as to the product or territory. An effect of this kind is not
ousted from consideration. And this is so because under the
discipline of section 8(d), the effects of the exclusionary c onduct
(harms and gains) must be weighed to determine ultimately whether
there has been an abuse. The aggregative judgment is made in
weighing the effects. It is not made by insisting that what can count
as a gain or a harm must itself meet some criterion of sufficiency or
can only be counted if it is an actual rather than a potential harm or
must be a market-wide effect.”25
24 Computicket, supra, at para 36. See also paras 29-33.
25 Computicket, supra, at para 31.
28
[59] In the light of these general principles, we now turn to consider each of the
elements required to establish a contravention of sec tion 8(1)(d)(ii) and
section 8(1)(c) with reference to the particular facts in this case.
RELEVANT MARKETS AND DOMINANCE
Relevant markets
[60] In order to succeed under section 8(1)(d)(ii) or section 8(1)(c) of the Act,
the applicants must of course first establish that Audatex is a dominant
firm (as defined in section 7) in a relevant market.
[61] In their complaint to the Commission, and in their founding affidavit in their
interim relief application, the applicants contend ed that there is a single
broad relevant market in this matter, namely “the market for the provision
of software for the management and processing of insurance claims in the
automotive repair industry”. The applicants allege further that Audatex,
Apollo and Motomatix all provide software and related services to the
automotive insurance repair market, and accordingly compete with each
other in this broad market.
[62] In its answering affidavit in the present application, Audatex disputed what
it referred to as the conflation of the activities of Audatex and Motomatix,
on the one hand, and Apollo on the other . Audatex contends that it
competes with Motomatix (but not with Apollo) in the market for identifying
the parts required for particular models of vehicles, estimating the cost to
29
replace those parts and linking insurers, claims assessors and vehicle
repairers (which is refers to as the “estimation market”) . Audatex alleges
that there are also various other players in this market, namely
[Competitor Names].
[63] Audatex contends that Apollo, for its part, competes in a separate market
for the provision of parts procurement software product s (which Audatex
refers to as the ” parts procurement market ”), together with at least five
other players, namely ([Competitor Name s]). In their papers, t he
applicants only list [Competitor Names] as competing providers of parts
procurement software.
[64] In their written and oral argument before the Tribunal, the applicants
departed from their previous stance that all the parties compete in a single
broad market, and instead adopted the respondents’ stance that Apollo
competes in a separate market to Audatex and Motomatix. The applicants
refer to the latter market as “the market for electronic vehicle damage
estimation and processing of automotive insurance claim s"; and to the
former market as “the market for value-added services offered to users of
the Audatex platform”.
[65] Based on the evidence before us , it appears , prima facie, that there are
indeed two distinct relevant markets in this matter, one in which Apollo
participates with (at least) other providers of parts procurement software,
30
and another in which Audatex competes with Motomatix/ GT Motive (and
possibly other players) in the provision of estimation and claims processing
services. As indicated above, there now seems to be consensus between
the parties in this regard. Purely for the sake of convenience at this stage,
we will adopt Audatex’s shorthand references to these two markets,
namely the “parts procurement market” and the “estimation market”.
[66] It furthermore appears that these market s are (at least partially) vertically
related, to the extent that the provision of parts procurement services takes
place on the estimation and claims processing platforms of parties such
as Audatex and GT Motive. The extent of the vertical integration between
these services is of course closely related to the alleged exclusionary and
anti-competitive effects of Audatex’s conduct , and is a matter of dispute
between the parties.
Dominance
[67] Notwithstanding their narrowing of the definition of the relevant market in
which Audatex competes, the applicants maintained that Audatex is a
dominant firm in the estimation market. They do so primarily on the basis
of Audatex’s own claim in its marketing material that it processes over 80%
of all insurance claims in South Africa , and has over 1200 r epairers as
clients.
31
[68] Audatex does not dispute these claims. However, it claims that there is no
distinct market for insurance repair work since its platform (and the GT
Motive platform) is also used by non -insurers (such as vehicle rental
companies).
[69] In our view, this is not a convincing rebuttal of the applicants’ dominance
claim. Notwithstanding that Audatex’s platform may be used by non -
insurers, the estimation and processing of insurance claims is what is
relevant in this matter. This lies at the heart of Audatex’s business model,
and the only customers of third party providers such as Apollo are
insurance customers.
[70] Audatex has also put up no evidence to suggest that GT Motive and the
other competitors it has listed in the estimation market are anything like its
size in the relevant market. On Audatex’s own version, GT Motive
currently only has [Number] customers.
[71] For all the above reasons, we are satisfied that the applicants have made
out a prima facie case that Audatex is a dominant firm in the estimation
market.
SECTION 8(1)(d)(ii) – REFUSAL TO SUPPLY GOOD S OR SERVICES TO A
COMPETITOR OR CUSTOMER
[72] There does not appear to be any dispute that Audatex’s conduct evidences
a refusal to supply Apollo. Audatex made it clear that it intended to remove
32
PartSmart’s access to the Audatex platform with effect from 31 March 2023
(subsequently extended to 14 April 2023 as a result of its undertaking in
these proceedings). Rather, Audatex’s case before the Tribunal, simply
put, is that it is not legally obliged to provide PartSmart with access to its
platform.
[73] In addition, i t appears to us, prima facie, that the provision of access to
Audatex’s platform constitutes a “ service” provided by Audatex to Apollo,
and that Apollo is a “customer” of Audatex in respect of that service, within
the meaning of section 8((1)(d)(ii). The app licants contend that the
essential service that Audatex provides to third party software providers
such as PartSmart is access to the flow of claims information on the
Audatex platform.
[74] Audatex argues that Apollo is not its “customer” because the arrangement
between the parties after the termination of the reseller agreement did not
provide for Audatex to provide any specific “ services” to Apollo. In our
view, however, this attributes an unduly narrow, and uneconomic, meaning
to the term “services” in section 8(1)(d)(ii) of the Act. It is common cause
that Apollo and Audatex were parties to a commercial arrangement in
terms of which Audatex permitted PartSmart to integrate with the Audatex
platform and to provide its own procurement services to customers o ver
that platform. It appears from the papers that there are various third party
service providers who are party to similar commercial arrangements with
33
Audatex, to the mutual commercial benefit of the Audatex platform and the
third party service providers. For all of these third party providers
(including PartSmart), Audatex is, in our view, providing a “service” (within
the meaning of section 8(1)(b)(ii)) by connecting them with their customers
in a real-time platform environment.26 As the Tribunal noted in GovChat,
technologies such as internet platforms are generally regarded as
“services” in competition law and industry.27
[75] A separate question is whether Apollo is (also) a “competitor” of Audatex.
As discussed above, the applicants initially claimed that Apollo competed
with Audatex and Motomatix in a broad “ market for the provision of
software for the management and processing of insurance claims in the
automotive repair industry ”. However, the applicants subsequently
accepted that Apollo competes in a distinct parts procurement market with
other third party service providers. Audatex stated in its answering affidavit
that it does not offer any parts procurement services in competit ion with
PartSmart,28 and this was not disputed by the applicants. There is also no
26 See eMedia, supra, at paras 111-113.
27 GovChat (Pty) Ltd and another v Facebook Inc. and others (Case No. IR165Nov20), 11 March
2021, para 108.
28 Prior to Audatex’s conclusion of the reseller agreement with Apollo, it offered a parts
procurement solution called Global InPart. However, Audatex explained that this product is no
longer in the market locally and that it no longer has a product that competes with PartSmart in
South Africa.
34
indication in the parties’ papers that there is any competitive interaction
between the activities of Apollo and those of Audatex.
[76] The applicants argued that Apollo is neve rtheless still a “ competitor” of
Audatex for purposes of section 8(1)(d)(ii) on the grounds that Apollo owns
Motomatix, and that Audatex regards them as one entity. However, we
think this is a mischaracterisation of Audatex’s position. The concern
expressed by Audatex was that Apollo and Motomatix effectively operate
as one entity in that they have overlapping employees and there is “ no
apparent limitation on the sharing of information between them ”. In our
view, this does not amount to a contention that Apollo competes in the
same market as Audatex. The market definitions adopted by all the parties
in argument indicate the contrary. The concern expressed by Audatex
was, rather, that Apollo was able to share with its subsidiary, Motomatix,
proprietary information that it (Apollo) had access to as a part procurement
service provider on the Audatex platform. That does not make Apollo a
competitor of Audatex. And, in any event, the applicants deny Audatex’s
allegations in this regard.
[77] Of course, it is common cause that Motomatix is a competitor to Audatex,
but there is no suggestion that any good or service was supplied (or
refused) by Audatex to Motomatix.
35
[78] We therefore conclude that the applicants have made out a prima facie
case that Apollo is a customer, but not that it is a competitor, of Audatex
for purposes of section 8(1)(d)(ii) of the Act.
SECTION 8(1)(d)(ii) – DID AUDATEX PROVIDE A “SCARCE” SERVICE TO
APOLLO?
[79] The Act does not define what is meant by the term “ scarce” goods o r
services in section 8(1)(d)(ii) term. According to the Oxford English
Dictionary, the word “scarce” denotes something that is “not abundant” or
is available “in deficient quantity ” to meet demand. In GovChat, the
Tribunal found that the technology app at issue in that matter “ cannot be
easily duplicated without significant capital investment and therefore can
be considered as ‘scarce’ or hard to come by”.29
[80] Audatex relies on the approach to refusal to supply cases in European law,
which asks whether the relevant good or service is “objectively necessary”
for the complainant to be able to compete effectively. However, caution
must be exercised in incorporating tests from other jurisdictions into the
particular structure of the South Africa n Act. As we understand it, the
European law on refusals to deal does not distinguish, as the South African
Act does, between the so -called “essential facility” doctrine (section
29 GovChat, supra, at para 113 (quoted with approval in eMedia, supra, at para 114).
36
8(1)(b)) and a refusal to supply “scarce” goods or resources (section
8(1)(d)(ii)).
[81] At the same time, however, a good or service can not be regarded as
“scarce” for the purposes of section 8(1)(d)(ii) simply by virtue of the fact
that it is provided by a dominant firm. This would conflate the separate
requirements of dominance and scarcity in section 8(1)(d)(ii). Dominance
relates to the market share and market power of the respondent firm in a
particular market . Scarcity relates to an insufficient availability in the
market of the relevant services that are being provided by the dominant
firm. Even if a firm is dominant in a certain market, that does not
necessarily mean that there is an insufficient availability in the market of
all the services that it offers. Of course, if the services in question are
unique to the dominant firm, or are a specific attribute of its dominance, 30
that may be the case, but that is an matter that has to be established by
evidence.
[82] The applicants make a general claim along these lines in the ir papers.
Relying on GovChat, they suggest that, because Audatex is the platform
of choice for insurers, assessors, and repairers, it offers unique access to
the flow of claims information from insurers using the Au datex platform,
which is not reasonably capable of being replicated.
30 See GovChat, supra, at paras 109-110; eMedia, supra, at para 114.
37
[83] In our view, however, there is an important difference between the facts in
GovChat and those in the present case . In GovChat, the applicants
rendered messaging services, the effectiveness of which was found to be
reliant on the network effects created by the WhatsApp platform. In the
present case, by contrast, the applicants limited their case to PartSmart’s
ability to service its [Number] existing insurance customers, namely
[Customer Names]. The dispute between the parties therefore revolved
around the extent to which Apollo is dependent on PartSmart’s access to
the Audatex platform in order to compete effectively for the business of
these [Number] specific customers.
[84] Therefore, unlike in GovChat, the broader network benefits that may be
associated with the Audatex platform are not directly relevant to the case
advanced by the applicants in their papers , and such benefits are only
potentially indirectly relevant insofar as they might impact on the
willingness of insurer customers of Audatex to deal with providers that are
not integrated on the Audatex platform. However, as discussed further
below, the evidence in this case was that Apollo’ s [Number] insurance
customers are willing to continue using the PartSmart app notwithstanding
its removal from the Audatex platform.
[85] In relation to the [Number] customers in question, it is not clear from the
applicants’ papers precisely what services (and information) they say are
uniquely available to PartSmart from Audatex. In argument, the applicants
38
claimed that what PartSmart requires access to is the real -time flow of
claims information between insurers, claims assessors and automotive
repair agents on the Audatex platform, including the estimation reports and
claims data of PartSmart’s [Number] customers. They contended, in
essence, that even if PartSmar t had access to the claims information of
their insurance customers outside the Audatex platform, it would be too
cumbersome and unwieldy for them and for their customers to process
efficiently.
[86] Audatex contends, however, that the basic functionality of the PartSmart
app is the same whether PartSmart receives the claims information it
needs from the Audatex platform, from Motomatix’s GT Motive platform
(referred to below), or from some other system (including the insurer itself).
Audatex claims, therefore, that there are various alternatives to the
Audatex platform that are available to PartSmart to continue competing
effectively for its customers’ business:
86.1 First, these insurance companies can simply terminate their
customer relationship with Audatex and switch to
Motomatix’s competing platform, GT Motive.
86.2 Second, Audatex has offered to extract t he repair
estimation report from the Audatex software, with the full
39
parts list per estimation and cost estimate, in order for the
insurers to pass that information directly on to PartSmart.
86.3 Third, Audatex has offered to assist the relevant insurers to
obtain the same estimation information referred to above
through an automated process on the insurer's own IT
system. Audatex states that t his will result in a similar
interface as is currently in place, allowing an insurer, in its
claims management system, to receive the estimation
report prepared by the repairer (using the Audatex
software).
[87] As regards the first option, Audatex stated that the GT Motive platform
provides the same functionality as the Audatex platform, and that this was
demonstrated by the f act that two significant short -term insurers
([Customer Names]) are already using the GT Motive system. In addition,
Audatex stated that, on 15 March 2023, one of PartSmart’s [Number]
insurance customers ( [Customer A]) gave notice to terminate its
relationship with Audatex after receipt of Audatex’s notification that it
intended to remove PartSmart’s access to the Audatex platform.
[88] The applicants acknowledge that [Customer A] has terminated its
relationship with Audatex and confirm that it is moving to GT Motive. They
40
say, though, that this does not necessarily mean that [Customer A] will
continue to use PartSmart, and that it may switch to one of its competitors.
[89] However, this is a purely speculative statement, which is unsupported by
any evidence in the record. The applicants’ response also does not dispute
the alleged functional substitutability of the Audatex and GT Motive
systems. In addition, a bsent any contrary evidence, the fact that
[Customer A] has decided to move to the GT Motive platform suggests
that it regards the GT Motive system as providing an equivalent offering to
that previously provided to it by the Audatex system. And, if [Customer
A] is not able to utilise PartSmart effectively on the GT Motive platform for
any reason, we would expect the applicants to say so, which they have not
done. Given the close relationship between Apollo and Motomatix, this
seems unlikely.
[90] As regards the second option, the applicants contend it would be hugely
onerous and cumbersome to process claims data in a manual rather than
automated fashion, especially given the ongoing changes that take place
in the claims estimation and repair process. They say that attempting to
do so would undermine the efficiencies that make PartSmart attractive to
insurers, which are currently able to automate the entire assessment,
procurement, and repair process via the Audatex platform. Given that
none of PartSmart’s insurance customers has indicated any interest in this
option, we disregard it in our assessment below.
41
[91] As regards the third option, the applicants say that this is currently non-
functional, technically unworkable, and would in any event render
PartSmart’s offering inferior to those of its competitors who are still
integrated on the Audatex platform . However, Audatex contends, in
response, that two of Apollo's customers ([Customers B and C ]) have
already selected this option , and that Audatex is actively d eveloping
alternative solutions with them.
[92] Audatex note s that a transition to this option “ requires investment by
Audatex and the insurer's appropriately qualified developers and other
technical personnel, which Audatex is in the process of providing”. It adds
that Audatex ha s in fact been requested by [Customer B] to make
changes to Audatex's own underlying IT system to better facilitate
functionality on the [Customer B] system and interface/communication
between that system and Audatex.
[93] Audatex states that, as a consequence, PartSmart will be able to receive
the same information regarding repair estimations, and to input into the
relevant insurers' system the same parts procurement information, that
PartSmart receives and inputs into the Audatex software.
[94] As regards [Customer B], the applicants attached to their founding
affidavit a supporting affidavit from Mr [X], the National Claims
Relationship Manager at [Customer B]. In that affidavit, Mr [X] had
42
stated, inter alia, that [Customer B] wished to continue using PartSmart
as integrated with the Audatex platform, and that [Customer B] was
opposed to the offboarding of PartSmart . He said that moving to
alternative parts procurement solutions with the same parts feed that
existed between A udatex and PartSmart was not an acceptable solution
to [Customer B] because it “totally disregards process impacts, change
management as well as integration capabilities buil[t] between PartSmart
and [Customer B] back-office system to allow full benefit a nd
functionality”. As a result, the offboarding of PartSmart would “cause harm
to [Customer B], its repairers, suppliers, and policyholders because of
higher parts prices, parts procurement costs and hence higher premiums
and excesses”.
[95] In response, however, Audatex attached to its answering affidavit an email
dated 7 March 2023 from Mr [Y], Head of Claims at [Customer B]
Commercial and Personal (and, according to Audatex, Mr [X’s] direct
manager), in which Mr [Y] stated that he “[Email Contents]”. Mr [Y]
continued that [Customer B] wished to “[Email Contents]” and therefore
that:
“[Email Contents]”
[96] Audatex states further that Audatex and [Customer B] subsequently had
various workshops and engagements and that, as at the date of Audatex's
answering affidavit, it was expected that [Customer B]’s “black box
43
solution” would be implemented “within a matter of days ”. Audatex
explained that this solution is a form of middleware hosted by [Customer
B] which Audatex interfaces with, and which will in turn provide PartSmart
with the information it requires and also enable PartSmart to input its own
information, such as parts adjustments, prices, lead times, and the like.
[97] Audatex adds that [Customer D] is underwritten by [Customer B], and
that the above solution will therefore apply equally in respect of [Customer
D]. This is not disputed by the applicants.
[98] In the ir replying affidavit, the applicants dismiss AudaBridge (the
technology being used for the [Customer B] “black box” solution) as old
software that is “grossly inferior” to the Audatex platform, with very limited
functionality. They state that AudaBridge also involves significant
integration costs for insurers, requires multiple manual steps, and does not
in fact work. The applicants attach affidavits from Mr Phillippus van Zyl (a
Technical Specialist at Apollo) and Mr Jacques van der Merwe (Apollo’s
Technical Director) stating that the alternative solution being developed by
Audatex with [Customer B] is not currently functional. Mr van der Merwe
estimates that it will take “until approximately the end of July” to work with
[Customer B] to implement the AudaBridge solution.
[99] The applicants state further that, even then, “PartSmart’s functionality will
be severely limited meaning that insurers such as [Customers B, C and
44
D] will almost certainly switch from PartSmart to a rival offering that does
have access to the Audatex platform such as [Competitor Name]”.
[100] However, Audatex responds in its rejoinder affidavit that the only way for
a third party to connect to Audatex is through AudaBridge or AudaConnect.
It says th at AudaConnect was designed for repairers and parts
procurement software providers like Apollo . AudaBridge, on the other
hand, was designed for insurers to interface and integrate with the Audatex
system. As such, AudaBridge, whilst an older technology, is Audatex’s
only technology for its interface with insurers, and its functionality has been
developed for that purpose.
[101] Audatex says that Apollo, u niquely, enjoyed access to both the
AudaConnect and AudaBridge interfaces by virtue of its historically deep
integration with the Audatex system. All other third party providers only
have access to the AudaConnect interface.
[102] It therefore appears that the alternative solution being developed by
[Customer B] would involve PartSmart losing the benefits of its historical
AudaConnect connection and having to rely instead on the AudaBridge
connection between [Customer B] and Audatex’s platform.
[103] As regards the difficulties alleged in respect of the black box integration
with [Customer B], Audatex states that there are only a few functionalities
45
that still require development, and that [Customer B] and Audatex are
currently working to address them.
[104] The position in respect of [Customer C] seems to be broadly similar.
Audatex has attached to its answering affidavit correspondence between
itself and Mr [Z], the Manager: Special Projects at [Customer C], in late
March 2023. In an email dated 29 March 2023, Mr [Z] wrote:
“In view of the tight deadlines, our support teams are currently under
considerable pressure to urgently set up the Proxy to communicate
with AudaBridge. While we are doing everything to urgently setup
the Proxy, we are cognisant that the parts feed communication
ultimately impacts financial decisions within our business and
therefore need to ensure that these platforms work as intended.
Furthermore, we have received feedback today indicating that the
parts feed communication between Audatex and PartSmart will not
terminate on the 31st of March, as previously stated. Instead, an
extension has been agreed.
As our representative for Audatex, could you please confirm the
accuracy of this information and amend the attache d parts feed
communication accordingly if necessary.”
[105] Thereafter, in response to Audatex’s confirmation that it had given an
undertaking not to offboard PartSmart until 14 April 2023, Mr [Z] wrote the
following on 30 March 2023:
“With the new extension until midnight on 14 April 2023, we will
adjust our plans accordingly to ensure quality is maintained. Our
46
teams will continue to work towards setting up the Proxy to
communicate with AudaBridg e. We plan to continue testing and
improving the Proxy until the 6th of April, at which point we will
request the PROD Credentials to switch over to the live
environment. Could you request that your team prepares the PROD
Credentials in anticipation of the 6th.”
[106] According to Audatex, this email indicates that (i) [Customer C] wishes to
retain PartSmart as a supplier notwithstanding its removal from the
Audatex platform, and (ii) [Customer C] anticipated that it would be able
to achieve this by establishing a direct feed between itself and PartSmart
by 6 April 2023.
[107] The applicants again say , in their replying affidavit, that the alternative
proposed by Audatex in respect of [Customer C] is not in fact functional,
and will not be functional by 14 April 2023. They attach in this regard an
email from Mr [Z] dated 6 April 2023 in which he objects to Audatex “using”
[Customer C] in its answering affidavit without [Customer C’s] consent;
states that the extract provided to him by the applicants (it is not clear
which one) “creates the impression that Audatex had quoted [Customer
C] out of context”; and that [Customer C] will be engaging with Audatex
and will decide its next steps based on Audatex’s response. However, the
letter does not say anything about the functionality of the solution referred
to by Audatex in its answering affidavit.
47
[108] What is the Tribunal to make of these various factual disputes at the interim
relief stage? Whilst we are unable to make any final determination on them
without the benefit of further evidence , it appears to us that the following
inferences can be made at the prima facie level. Both [Customer B and
Customer C] were concerned about the offboarding of PartSmart from the
Audatex platform , and may have preferred to retain the benefits they
enjoyed arising from the full integration of PartSmart with the Audatex
platform. However, precisely because of the benefits that both systems
afford them, both insurers were willing to develop alternative solutions with
Audatex and PartSmart rather than dispense with either of them.
Furthermore, whilst the re appear to be certain technical challenges with
the implementation of these alternative solutions, and there is a question
whether they are yet fully functional, the available evidence does not
suggest that either insurer is likely to move from PartSmart to one of its
competitors. On the contrary, the insurers appear to be resolute in finding
solutions that involve their ongoing use of PartSmart rather than one of its
competitors.
[109] On the available evidence, therefore, it does not appear to us that
PartSmart is likely to lose any of its [Number] customers to competitors
as a result of its offboarding from the Audatex platform. [Customer A] has
already decided to move to the GT Motive platform (and seems likely to
retain PartSmart as its parts procurement service provider in that context)
48
whilst PartSmart’s other customers appear likely to continue using it
through an alternative solution being developed with Audatex.
[110] Having regard to this evidence, we do not believe that it can be concluded,
on a prima facie basis, that the services that Audatex has ceased to supply
to Apollo are “scarce” services within the meaning of section 8(1)(d)(ii) of
the Act, or that Apollo’s removal from the Audatex platform will “exclude”
them from the parts procurement market, even on the expanded definition
of that term now contained in the Act.
[111] If we are incorrect in this conclusion, the above evidence is also relevant
to the assessment of anti -competitive effects, irreparable harm and
balance of convenience, as discussed further below.
SECTION 8(1)(d)(ii) – ECONOMIC FEASIBILITY
[112] In terms of section 8(1)(d)(ii), a dominant firm is on ly required to supply
scarce goods to a competitor or customer “when supplying those goods or
services is economically feasible”.
[113] Given that Audatex has supplied a fully integrated service to PartSmart for
a number of years, there can be no dispute that th e provision of that
service is economically feasible – save, of course, for the concern now
expressed by Audatex regarding access to competitively sensitive
information brought about by Apollo’s acquisition of Motomatix and the GT
Motive distributorship in South Africa. This is the basis upon which
49
Audatex justifies its decision to offboard PartSmart from the Audatex
platform.
[114] It is therefore necessary to consider whether PartSmart, as integrated with
the Audatex platform, had access to competitively sensitive information of
Audatex; and whether Audatex had a reasonable apprehension that such
information would come into the hands of Motomatix absent its
offboarding.
Audatex’s concerns
[115] Audatex contends that PartSmart enjoyed a uniquely deep integration with
the Audatex platform, which provided it with unfettered access to
Audatex’s parts database. Audatex states that this database has been
built up over decades by Audatex employees a cross the world , and
contains information relating to 1,349 vehicle models and over 17 million
parts.31 In respect of each model, the database reflects (i) a physical
description of each part, (ii) a unique part identification number, and (iii)
the estimated cost of each part . In addition, because new vehicles are
constantly being released to the market, Audatex has a team of employees
whose sole function is to monitor the market for new vehicles and to update
31 Audatex says that the average vehicle is made up of more than 10,000 parts.
50
Audatex’s parts database so that it rem ains current and relevant to
Audatex’s customers.
[116] Audatex adds that its parts database in South Africa has been localised to
parts for South African vehicles and local prices, which requires constant
investment to remain current. As such, access to the A udatex parts
database provides a unique insight into cars on the road in South Africa,
per insurer, that are being repaired. Audatex states that this would be of
particular benefit to Motomatix in localising GT Motive’s database for the
South African market.
[117] Audatex argues further that there are important policy justifications behind
a firm ’s entitlement to protect its intellectual property from misuse by
competitors. In particular , if free -riding were permitted, the incentive to
invest would be drastically diminished.
[118] The applicants do not dispute the legitimacy of a justification of this sort,
but they contend that it is a contrived one in this case. The applicants
advance two contentions in this regard. The first is that PartSmart did not
have access to competitively -sensitive inform ation on the Audatex
platform; and the second is that Audatex’s claim was raised for the first
time in its answering affidavit, and is inconsistent with its previous course
of conduct. We deal with each of these contentions in turn.
PartSmart’s access to confidential information
51
[119] The applicants contend that the parts database is not proprietary to
Audatex because it is a listing of information supplied by the various
vehicle manufacturers. Audatex’s response is that, whilst it is correct that
each manufact urer owns the vehicle information that it supplies to
Audatex, Audatex’s proprietary rights arise from, and relate to, the fact that
it consolidates each manufacturer’s information into a single database and
updates it on an ongoing basis.
[120] Audatex states that, while much of the underlying information can be
sourced from the various OEMs, that would be a hugely time -consuming
and costly exercise, which would require significant investment by a
competitor. The access that Motomatix enjoyed to the Audatex pl atform
through PartSmart would allow it to compile (or improve) its own parts
database based on recreating Audatex ’s parts database, extracting the
information from Audatex ’s estimation reports and inputting that into a
competing database, based on the research and investment already
performed by Audatex.
[121] Apollo’s Technical Director, Mr van der Merwe, states that Apollo’s
integration with Audatex does not allow it to access or download Audatex’s
parts database, or the source code that underpins the workf lows and
processes of the platform. He says that the only information that Audatex
makes available to PartSmart is the estimation reports (setting out parts
52
required and estimates on part, labour and paint) that the platform
generates.
[122] However, Audatex ’s r esponse is that this is precisely the problem.
Through the uniquely deep integration that PartSmart enjoyed with the
Audatex platform (which gave it access to both the AudaBridge and
AudaConnect interfaces) , Apollo ha d access in real -time, and on an
ongoing basis, to thousands of Audatex’s estimation reports , to all
amendments to those reports over time, and to the workflow associated
with them. Audatex says that each estimation report contains information
drawn directly from Audatex’s parts database, including parts prices, parts
names, parts descriptions, VIN numbers and parts number, and markup
and guide numbers. It says that Motomatix (through Apollo) could simply
copy this information and paste it into its own competing parts database to
develop and improve it.
[123] Audatex says that Apollo was also able to identify the manner in which
Audatex parts are classified or categorised within the Audatex platform, by
reference to the vehicle model in question. Audatex says that it would not
share its classification data for its parts with a competitor, as this would
enable the competitor to better identify parts within the database, source
details regarding those parts within in the local market , and replicate the
database.
53
[124] There is some support in the record for Audatex’s contention that, pursuant
to the reseller agreement, Apollo obtained a degree of integration with
Audatex’s system that was significantly deeper than that afforded to other
third party service providers, including PartSmart’s competitors. As noted
above, the termination of the reseller agreement was memorialised in a
letter agreement ultimately signed by both Audatex and Apollo dated 11
October 2022. The final signed version of that agreement is attached to
Audatex’s answering affidavit. However, an earlier version of the
agreement, which was dated 7 October 2022 and signed by Apollo (albeit
not by Audatex), was relied on by the applicants in their founding affidavit.
That version of the agreement contained the following provisions:
124.1 “the depth of each integration between PartSmart and the
Audatex Software is extensive in that it may be
characterised as being two -way integration ” (clause
2.2.8.1);
124.2 “whilst the Audatex Software is from time to time integrated
with parts pricing software other than PartSmart, such
integrations may be characterised as one-way integrations”
(clause 2.2.8.2);
124.3 “in light of the aforegoing, and with a view to maintaining the
unique functionality afforded to clients by way of the
Audatex/ PartSmart integration, Apollo agrees that for as
long as PartSmart remains the only parts pricing software
with which Audatex effects a two -way integration , Apollo
shall make payment to Audatex of a flat monthly fee, the
54
amount of which shall be negotiated and set out in the New
Agreement” (clause 2.2.9); and
124.4 “nothing sh all preclude Audatex from enabling one -way
integration between the Audatex Software and any parts
pricing software other than PartSmart” (clause 2.2.10).
[125] Whilst these provisions were (for reasons that were not explained) deleted
from the final letter agree ment signed by the parties, their inclusion in a
draft signed by Apollo prima facie supports Audatex’s claim that PartSmart
did enjoy a uniquely deep level of integration with the Audatex platform.
[126] The applicants argue that the same information provided to PartSmart by
Audatex is also provided to other providers who have been retained on the
Audatex platform, including its rivals [Competitor Names]. However, it is
not evident whether the deeper level of integration enjoyed by PartSmart
provided it with greater access to confidential information than these other
service providers. And, of course, unlike PartSmart, those other providers
are not associated with a direct competitor of Audatex.
Audatex’s previous course of conduct
[127] The applicants argue that another indicator of the contrived nature of
Audatex’s confidentiality concern is that it was raised for the first time in
Audatex’s answering affidavit. They say that Audatex knew about the
relationship between Apollo and Motoma tix, and the fact that Motomatix
has acquired a licence to GT Motive, as far back as August 2022; and that
55
Audatex had various discussions during September and October 2022
about their future commercial relationship precisely because Motomatix
had acquired a licence to GT Motive. The applicants say that Audatex did
not, in any of those exchanges, raise any concern a bout Apollo or
Motomatix having access to Audatex’s confidential information by virtue of
PartSmart’s access to the Audatex platform.
[128] The applicants rely in particular in this regard on the terms of the letter
agreement between the parties dated 11 October 2022. As set out above,
it was recorded i n that letter that, following termination of the reseller
agreement with effect from 30 September 2022, Apollo and Audatex would
enter into a new agreement, effective 1 October 2022, in terms of which
Audatex transferred to Apollo all of Audatex’s obligations to provide
PartSmart to PartSmart’s [Number] customers. C lause 2.2.7 then
provided that:
“Insofar as clients who utilise the PartSmart software already use
same in integration with Audatex’s claims estimation software (“the
Audatex Software”), or insofar as a client requests integration of
PartSmart with the Audatex Software, Audatex shall be entitled (but
not obliged) to charge such cli ent an integration fee per claim, or
otherwise as Audatex considers appropriate”.
[129] The applicants contend, with reference to this clause, that it was expressly
stated and confirmed by Audatex ’s representatives that, despite the
termination of the reseller a greement, PartSmart’s customers would
56
continue to be able to use PartSmart, without interruption and, by
implication, that PartSmart would continue to have access to the Audatex
platform via the PartSmart-Audatex integration.
[130] Audatex says in response that, while it was aware in August-September
2023 that there was a connection between Apollo and Motomatix
(including that Mr Moodley was involved in both companies) , it was not
aware at the time that Motomatix was a wholly-owned subsidiary of Apollo;
of the level of integration between these two firms ; or that employees of
Apollo who Audatex had shared technical details with over many years
(including Mr V an der Merwe) would similarly be involved in technical
aspects of the GT Motive product.
[131] Audatex says it only became aware of all of these facts after it received an
email request from Motomatix itself on 16 January 2023 for information
from Audatex’s parts database . In that email, one Phil Van Zyl , from a
Motomatix email address , requested Audatex to provide an “ updated
Glass Identification Listing”. Mr Van Zyl followed up on that email on 16
January 2023 and then, on 24 January 2023, the same request was
subsequently sent to Audatex from [Customer B]. The same Mr Van Zyl
is referred to in the applicants’ replying affidavit as a Technical Specialist
employed by Apollo.
57
[132] Audatex claims that this information, if provided, would entail a file
containing the make, model, guide number and part price of the seri es of
products in question. It says that even Apollo was only supposed to
receive the parts details for an actual completed repair estimation, rather
than underlying database information; and that there was no basis
whatsoever for Motomatix, a direct competitor, to seek this information.
[133] Audatex says that it was immediately alarmed by the apparently close
relationship between Apollo and Motomatix, and therefore decided that, as
of 31 March 2023, it would no longer grant Apollo access to the Audatex
system.
[134] The applicants’ response to this is that the glass listing was requested to
update PartSmart’s [Customer B] and [Customer C] integration so that,
when an estimation report listed replacement glass products, the prices
would be set to nil, requiring rep airers and suppliers to quote for
replacement. They say that [Customer B] and [Customer C] had
requested this because they regarded the default glass pricing supplied by
the parts manufacturers as uncompetitive.
[135] However, this response does not explain why the relevant information was
ostensibly being sought by Motomatix, not Apollo. Nor does it suggest that
Audatex was, or should have b een, aware that the information was not
being sought for the benefit of Motomatix.
58
[136] Of course, the fact that Motomatix was requesting this information by email
suggests that it was not readily attainable by Apollo from the Audatex
platform. However, it nevertheless seems reasonable to us that Audatex
would have been concerned that Motomatix – a direct competitor and
subsidiary of its customer Apollo – was requesting proprietary database
information from Audatex , especially given that the same Mr Van Zyl is
also a Technical Specialist in the employ of Apollo.
[137] It therefore appears to us, prima facie , that Audatex ha d a legitimate
concern about the risk of its proprietary information being disclosed by
Apollo to Motomatix, Audatex’s direct competitor.
[138] There is also no explanation, on the applicants’ version, of why, if Audatex
was aware in August/September 2022 of all the links between Apollo and
Motomatix/ GT Motive that it refers to, it waited until February 2023 to
decide to offboard PartSmart.
[139] In the circumstances, we do not believe that there is a sufficient basis for
us to reject, at this stage of the proceedings, the explanation provided by
Audatex for its decision to offboard PartSmart.
[140] The applicants did not dispute that the need to protect proprietary
information would constitute a legitimate basis , in terms of section
8(1)(d)(ii), for a dominant firm not to supply services to a customer.
59
[141] For the above reasons, we do not believe that the applicants have made
out a prima facie case that it was economically feasible for Audatex to
continue supplying Apollo services on the integrated basis that it
previously enjo yed, given the risk it reasonably apprehended that
PartSmart would use its access to the Audatex platform to appropriate
confidential information for the benefit of Audatex’s direct competitor in
South Africa, GT Motive.
[142] There is a separate question is whe ther it was necessary, in order to
protect Audatex’s confidential information, to offboard PartSmart entirely
from the Audatex platform, or whether that could have been achieved by
affording PartSmart a lesser form of access to the platform at the time .
However, this issue was not canvassed on the papers, or advanced in
argument by the applicants, and we therefore do not have any basis to find
that a less extreme form of relief was available to Audatex when it decided
to offboard PartSmart.
[143] Two proposa ls were raised by the applicants at the hearing. First, the
applicants tendered various undertakings on behalf of Apollo and
Motomatix, and their respective employees, not to share or use , for the
benefit of Motomatix , any confidential information that wa s accessible to
PartSmart on the Audatex platform. Second, the applicants applied to
amend their notice of motion to reformulate the relief they sought in
paragraph 2 of their notice of motion as follows:
60
“That the first respondent (“Audatex”) is interdic ted and restrained
from terminating the access of ‘PartSmart’ (owned and operated by
the first applicant) to the “Audatex platform” (as described in the
founding affidavit), via its existing AudaConnect accounts: i)
[Partsmart Account] , ii) [Partsmart Account], iii) [Partsmart
Account], iv) [Partsmart Account] and v) [Partsmart Account],
and its existing AudaBridge account: vi) [Partsmart Account], with
their existing functionality and operability alternatively functionality
and operability no less effectiv e or favourable than provided from
time to time to any rival of the first applicant (with any changes in
functionality or operability being on reasonable notice to the first
applicant), pending the conclusion of a hearing into the alleged
prohibited practices that are the subject of the applicants’ complaint
to the second respondent which is attached to the applicants’
founding affidavit marked “FA1”, alternatively six months from the
date of this order, whichever occurs first, and subject to the
undertakings and indemnity tendered in the applicants’ affidavit
dated 11 April 2023.”
[144] However, given that both of these proposals were made at such a late
stage, Audatex did not have an opportunity to respond to them. Audatex
also correctly pointed out that it wo uld have no means to monitor
compliance with th e undertakings. A s regards the revised relief , the
applicants confirmed in argument that the effect of the proposed
amendment would not be to limit in any material way the current access
that PartSmart enjoys to the Audatex platform, but simply to specify it in
more detail. The revised relief therefore would not have taken the matter
any further. In addition, Audatex disputed that certain of the account
61
information contained in the revised relief was inaccurate. For all these
reasons, we decided to decline the application to amend.
SECTION 8(1)(d)(ii) – ANTI-COMPETITIVE EFFECTS
[145] As discussed above, it is necessary under both section 8(1)(d)(ii) and
section 8(1) (c) to demonstrate that the impugned conduct has an anti -
competitive effect even if it meets the other requirements of those sections.
[146] In their papers, the applicants contended that the offboarding of PartSmart
would have anti-competitive effects relating to both Apollo and Motomatix.
However, i n their heads of argument, and in their oral argument, the
applicants pressed only the former. We will therefore deal primarily with
the market in which PartSmart operates, but will also make some brief
comments in relation to Motomatix.
The parts procurement market
[147] As discussed above, the applicants ’ position at the hearing – consistent
with Audatex ’s position and with our prima facie findings – was that
PartSmart competes in a market for the provision of parts procurement
services together with other parts procurement service providers such as
[Competitor Names] (but not with Audatex).
[148] We have assessed whether there is prima facie evidence that the
offboarding of PartSmart will impact on its ability to compete for the
62
[Number] customers on which the applicants’ application is based, and
we have found that there is insufficient evidence for such a finding.
[149] The applicants make a passing allegation in their founding affidavit that
Audatex informed a potential customer of PartSmart, [Customer Name],
that it should “ reconsider” dealing with Apollo because PartSmart’s
integration to the Audatex system would shortly be disabled. However, the
applicants do not disclose whether or not PartSmart in fact subsequently
acquired [Customer Name] as a customer, and for what reasons.
[150] Audatex furthermore denies that has promoted any parts procurement
provider over another because it has no incentive to do so. It says that it
has simply indicated, to insurers who enquire, which parts procurement
service providers are integrated with its platform and which are not.
[151] Insofar as other potential insurance customers are concerned, there is no
evidence at all in the papers regarding the extent to which they would be
willing to deal with PartSmart if it is not integrated with the Audatex
platform.
[152] However, even if Audatex’s con duct were to have an exclusionary effect
on PartSmart, the applicants must still show that such conduct has an anti-
competitive effect in the parts procurement market.
[153] We note that the applicants’ theory of harm in this regard is not a typical
case of ver tical leveraging or foreclosure, because they do not dispute,
63
and do not adduce any evidence to controvert, Audatex’s contention that
it is not an actual or potential competitor in the parts procurement market,
and accordingly has no obvious incentive to foreclose competition in that
market. The applicants also do not advance the standard theory of harm
that Audatex’s conduct is creating, protecting or extending any market
power for Audatex in the parts procurement market. The standard test for
anti-competitive effects referred to in Computicket (would the dominant
firm’s conduct “strengthen [its] position to the extent that competitive rivalry
is significantly impeded or is likely to be so impeded ”?) 32 is therefore not
applicable to the applicants’ theory of harm.33
[154] In th ese circumstances, the applicants’ theory of harm must be that
Audatex is, by its conduct, “incidentally” causing anti-competitive effects in
the parts procurement market, even though it is not itself an actual or
potential competitor in that market, and despite the absence of any theory
or evidence as to how Audatex will derive any benefit in th e parts
procurement market from its conduct.
[155] Assuming that such a theory of harm is available under section 8(1) of the
Act (which it is not necessary or appropriate for us to determine in this
32 Computicket, supra, at para 36.
33 See also Bulb Man, supra, at paras 52-56.
64
application),34 the absence of any obvious incentive to foreclose is still a
significant factor in assessing the likelihood of anti -competitive effects in
the relevant market.
[156] This is particularly so where Audatex has provided what we have prima
facie found to be a legitimate reason for its conduct, namely the protection
of its confidential information from a direct competitor.
[157] In addition, the applicants have not adduced any evidence regarding the
nature and dynamics of competition in the parts procurement market.
Even if Audatex’s conduct did have an exclusionary impact on PartSmart,
the available evidence is that there are, on the applicants’ version, at least
three other significant competitors in the market ( [Competitor Names])
and Audatex also refers to [Competitor Names]. We know nothing about
the competitive strengths and weaknesses of PartSmart relative to these
different players or, in particular, what the exclusion of PartSmart would
have on the market power of these other players.
[158] We also note in this regar d that, following the termination of the reseller
agreement between Apollo and Audatex, Audatex integrated t hree of
PartSmart’s rivals ([Competitor Names]) onto its platform. The fact that
these players were regarded as significant competitors prior to th eir
34 The recent inclusion of the words “or customer” in section 8(1)(d)(ii) is consistent with such an
interpretation. However, it may arguably still be necessary to show that the refusal to supply a
customer protects or extends the market power of the dominant firm.
65
onboarding by Audatex raises further questions about the importance to
suppliers in the parts procurement market of being integrated on the
Audatex platform. Furthermore, there are now t hree alternatives to
PartSmart on the Audatex platform, whereas previously PartSmart
enjoyed an exclusive competitive advantage in this regard.
[159] Finally, we have had regard to the applicants’ allegations, supported by Mr
[X] of [Customer B], that the offboarding of PartSmart would caus e
prejudice to insurers and to policyholders. However, those assertions are
all made in conclusory terms, and no explanation has been provided as to
how th e alleged anti-competitive effect would come about in the
circumstances referred to above. We also note that Mr [X]’s affidavit does
not make any reference to the development of the “black box” solution
referred to in Mr [Y]’s email of 7 March 2023 and elaborated upon in
Audatex’s answering affidavit.
[160] In the circumstances, we do not believe we have a sufficient evidential
basis to make a prima facie finding that Audatex’s conduct would result in
an anti-competitive effect in the parts procurement market.
[161] We come to this conclusion cognisant of the requirement to bring a
transformative and context -specific approach to abuse of dominance
complaints such as those at issue in this case. In this case, however, the
theory of harm advanced by the applicants is (as explained above) not a
66
typical one of leveraging, and , after careful consideration, we have come
to the conclusion that the available evidence is insufficient to reach a prima
facie finding of anti-competitive effects in the parts procurement market, or
that the transformative objectives of the Act would be undermined if interim
relief is not granted.
The estimation market
[162] Insofar as the estimation market is concerned, a theory of horizontal
foreclosure is available to the applicants. In addition, the evidence
suggests that Motomatix/ GT Motive may be a potentially significant
entrant into the South African estimation marke t, and an increasingly
competitive threat to Audatex’s incumbent position going forward.
However, this foreclosure theory has not been pursued with any vigour by
the applicants. This is perhaps because there is a complete dearth of any
evidence as to how harm to PartSmart in the parts procurement market
would, indirectly, result in harm to the competitive position of Motomatix in
the estimation market.
[163] Two arguments are hinted at in the applicants’ papers. The first is that
PartSmart’s insurance custom ers are unlikely to use PartSm ’rt's offering
even if integrated via the GT Motive platform if their recent experience of
using PartSmart is that it is no longer operable. The second is that harm
67
to PartSmart will inhibit the ability of Apollo to invest in the GT Motive
platform through Motomatix.
[164] However, the applicants do not provide any evidence in support of the first
argument, and it is inconsistent with the conduct of PartSmart’s [Number]
customers, as discussed above. The applicants also do not expl ain why
GT Motive w ould not be able to integrate with other providers of parts
procurement services such as [Competitor Names] if it so wishes.
[165] As regards the latter argument, the applicants raised this for the first time
in their replying agreement, and again only in conclusory terms. They
provided no evidence regarding the growth plans or funding requirements
of Motomatix; the extent to which Motomatix is reliant on Apollo for those
requirements; or the percentage contribution of PartSmart to the revenues
of Apollo. The applicants do not state whether Apollo has any businesses
other than PartSmart, or who the shareholders of Apollo are. All that is
stated is that Apollo and Motomatix are South African owned , and are
Level 4 B-BBEE contributors.35
[166] Audatex also alleges that there are various other players in the South
African market, namely [Competitor Name s]. We again do not know
35 This is to be contrasted with eMedia, where the applicants provided considerable evidence in
support of a similar theory of harm.
68
anything about the offerings of these players, or the nature and extent of
their competitive relationship with Audatex and GT Motive.
[167] Having regard to the above, we find that the applicants have not made out
a prima facie case of anti -competitive effects in either the parts
procurement market or the estimation market under section 8(d)(ii) of the
Act.
[168] In the light of this finding, we do not need to consider whether Audatex has
made out an efficiency defence for its conduct . We note, however, th at
Audatex’s concerns regarding the protection of its proprietary information
may be cognisable as such a defence if we are incorrect in assessing it
under the heading of economic feasibility in relation to section 8(1)(d)(ii) of
the Act.
[169] For all the reasons set out above, we conclude that the applicants have
not established a prima facie contravention of section 8(1)(d)(ii) of the Act.
SECTION 8(1)(c) OF THE ACT
[170] The above findings are also dispositive, in our view, of the applicants’
alternative claim under section 8(1)(c) of the Act . We do not believe that
the applicants have made out a prima facie case that Audatex’s conduct
is exclusionary (i.e., that it will impede or prevent PartSmart from
participating in, or expanding within, the parts procurement market), or that
69
it has had an anti-competitive effect in the parts procurement market or in
the estimation market.
[171] A difficult question is where, in the assessment of section 8(1)(c), it is most
appropriate to locate the consideration of Audatex’s justification relating to
the protection of confidential information. However, the applicants did not
dispute that, if this justification is made out, it would be cognisable under
section 8(1)(c) as it is under section 8(1)( d)(ii) of the Act . It is therefore
unnecessary, and inappropriate in the context of the current application, to
make any further determination in this regard.
HARM AND BALANCE OF CONVENIENCE
[172] The Tribunal has previously stated that it would be reluctant to grant
interim relief in c ircumstances where a prohibited practice has not been
made out on a prima facie basis.36 It any event follows from what we have
said above that we do not believe that the refusal of interim relief in this
case will result in serious or irreparable harm to the competitive position of
PartSmart or Motomatix, or that the balance of convenience warrants the
grant of interim relief.
[173] As discussed above, we are required to consider whether there are “clear,
non-speculative and uncontroversial facts ” that are relevant to the
36 See Replication Technology Group (Pty) Ltd v Gallo Africa Ltd (Case No. 92/OR/Sep07) at para
29 (and the cases cited in fn 7).
70
assessment of these factors, and to take a “ robust approach ” on the
evidence before us.37
[174] The “serious or irreparable damage ” that is relevant to the enquiry under
section 49C is damage to the competitive position of the applicant.
Likewise, the prejudice relevant to the assessment of the balance of
convenience is prejudice to the respective competitive positions of the
parties in the relevant market(s).38
[175] We do not believe that the applicants have made out a case that the
offboarding of PartSmart from the Audatex platform will cause serious or
irreparable damage either to Apollo or to Motomatix.
[176] As regards Apollo, the evidence suggests that all [Number] of PartSmart’s
existing customers are likely to remain so, either on the Motomatix platform
or by way of the alternative solutions being developed b y the relevant
insurers and Audatex. The fact that the alternative solutions have involved
development cost on the part of the insurers, and that certain technical
difficulties are still being addressed, does not represent harm to the
competitive position of Apollo. The evidence does not suggest that these
challenges have dissuaded PartSmart’s customers from continuing to use
it – quite the contrary.
37 eMedia, supra, at paras 80-81.
38 BCX, supra, at paras 21-22.
71
[177] The applicants also have not made out any case that Apollo’s ability to
compete for new customers will be seriously or irreparably damaged by its
removal from the Audatex platform. The applicants did not put up any
evidence that they have in fact lost, or will lose, any potential customers if
they are not integrated on the Audatex platform. It is also relevant in this
regard that both parties regard [Competitor Name s] as effective
competitors in the parts procurement market notwithstanding that they
have only recently been onboarded onto the Audatex platform.
[178] As regards Motomatix, the applicants do not make out any case that its
competitive position in the estimation market will be seriously or irreparably
harmed by the offboarding of PartSmart. The only arguments advanced
by the applicants on this score are that customers might be dissuaded from
using PartSmart on the GT Motive platform, and that less revenue will be
available from Apollo to fund the growth of the GT Motive platform in South
Africa.
[179] However, the available evidence is inconsistent with the former argument,
and the applicants do not explain why GT Motive would not be able to
integrate with other provid ers of parts procurement services such as
[Competitor Names] if it so wishes. As regards the latter argument, t he
applicants raised this for the first time in their replying agreement, and
again only in conclusory terms. The applicants did not put up any evidence
regarding the impact that the offboarding of PartSmart is likely to have on
72
the revenues of Apollo, or regarding the knock-on effect, if any, that would
have on the growth plans of GT Motive. This is to be contrasted with
eMedia, where the applicants put up considerable evidence on the direct
revenue impact of the exclusionary conduct in that case, and on the
indirect effect that would have on the ability of the applicant’s platform to
compete with that of the dominant firm.
[180] The absence of eviden ce of serious or irreparable damage to Apollo or
Motomatix is also relevant to our assessment of the balance of
convenience. On the other side of the scale is what we have found to be
prima facie credible evidence that, if PartSmart retains access to the
Audatex platform, that may be used to appropriate confidential information
of Audatex for the benefit of Motomatix’s competing GT Motive platform.
Any such springboarding would clearly be detrimental to Audatex’s
competitive position in the estimation market for the reasons explained by
Audatex.
[181] As noted above, the applicants tendered various undertakings at the
hearing on behalf of Apollo and Motomatix not to share or use, for the
benefit of Motomatix, any confidential information that was accessib le to
PartSmart on the Audatex platform. However, as Audatex pointed out, it
would have no means to monitor and enforce compliance with the se
undertakings.
73
[182] The applicants also applied at the hearing to amend their relief in the
manner discussed in paragraph 143 above. However, the applicants
confirmed in argument that the effect of the proposed amendment would
not be to limit in any material way the current access that PartSmart enjoys
to the Audatex platform, but simply to specify it in more detail. The revised
relief therefore would not have taken the matter any further.
[183] We have also considered the position of insurers and end-customers in
our assessment of the balance of convenience. As discussed above, the
evidence suggests that PartSmart’s insurer customers may have preferred
for it to remain integrated with the Audatex platform. However, [Customer
A] has already moved to the GT Motive platform, whilst [Customer B] and
[Customer C] are already far advanced in developing alternative solutions
with Audatex that will allow them to retain the benefits of both Audatex and
PartSmart in their claims processing operations. In the circumstances, it
does not appear to us that the grant of interim relief at this stage will
prevent material prejudice to such insurers or end-customers. There is no
non-speculative evidence that the alternatives in question will result in end-
customers paying higher insurance prem iums or otherwise being
prejudiced as a result of the offboarding of PartSmart.
[184] In all the circumstances , we do not believe that the considerations of
irreparable or serious harm and balance of convenience, weighed with the
74
weak evidence of a prohibited practice by Audatex, warrant the grant of
the interim relief sought by the applicants in this application.
ORDER
[185] For the above reasons, we dismissed the applicants’ application for interim
relief.
[186] We did not make any order for the payment of any costs. We do not believe
we have any power to order costs in interim relief proceedings such as
these, nor would we regard it as appropriate to do so in this case given
that the applicants’ complaint is, in our view, bona fide, not frivolous, and
its merits have yet to be finally determined.39
39 BCX, supra, para 54.
08 May 2023
Presiding Member
Mr Jerome Wilson
Date
Concurring: Ms Mondo Mazwai and Professor Liberty Mncube
Tribunal case managers: Matshidiso Tseki and Sinethemba Mbeki
For the Applicant: Adv Robin Pearse SC and Adv Luke Kelly instructed
by Dingley Marshall Lewin Attorneys.
75
For the First Respondent: Adv Alfred Cockrell SC and Adv Shannon Quinn
instructed by Werksmans Attorneys.
For the Second Respondent: Ms Betty Mkatshwa