Anticipated acquisition by BT Group plc of IP TradeSADecision on relevant merger situation andsubstantial lessening of competitionME/6658/16The CMA’s decision on reference under section 33(1) of the Enterprise Act 2002given on 25 April 2017. Full text of the decision published on 11 May 2017.Please note that [ ] indicates figures or text which have been deleted orreplaced in ranges at the request of the parties for reasons of commercialconfidentiality.SUMMARY1.BT Group plc (BT), through its subsidiary BT (Netherlands) Holding BV, hasagreed to acquire IP Trade SA (IP Trade) (the Merger). BT and IP Trade aretogether referred to as the Parties.2.The Competition and Markets Authority (CMA) believes that it is or may be thecase that the Parties will cease to be distinct as a result of the Merger, that theshare of supply test is met and that accordingly arrangements are in progressor in contemplation which, if carried into effect, will result in the creation of arelevant merger situation.3.The Parties overlap in the supply of turret systems (and associatedmaintenance and support) used in voice trading applications to customersbased in the UK. In its assessment, the CMA has considered it appropriate toinclude within the product frame of reference all types of turret systems used invoice trading applications (including hardware, hybrid and cloud-basedsolutions), on the basis that some customers considered all types of turretsystem to be substitutable. This further represents the most conservativeapproach to identifying potential competition concerns, as the Parties would ineach case not overlap if the CMA were to establish frames of referencedelineated by the types of turret system described by the Parties. In terms of1

geographic frame of reference, while the CMA found that there are reasons tobelieve that the supply of turret systems has global characteristics, on acautious basis the CMA has assessed the potential impact of the Merger in theUK. Accordingly, the CMA has assessed whether the Merger may give rise tohorizontal concerns in the supply of turret systems (and associatedmaintenance support) used in voice trading applications in the UK.4.In the UK, customers purchase turret systems either directly from suppliers orindirectly through third party distributors and resellers. While BT supplies itsturret systems directly to end-customers, IP Trade supplies many of its turretsystems through third parties. These third parties may themselves competewith BT both in the supply of turret systems and the supply of certain installationand integration services associated with turret systems. Consequently, inaddition to assessing potential horizontal concerns arising from the Merger, theCMA has investigated whether the merged entity could harm competitionthrough foreclosure, by ceasing to supply, or supplying on less favourableterms, its turret systems to such third parties.5.With regard to the horizontal concern, the CMA found that the merged entitywill have only a modest combined shared of supply (of between 20% and 30%)with the Merger resulting in only a small increment to BT’s share of supply (lessthan 5%). The CMA also found that:(a) For larger customers, IP Trade does not have the scale or geographic reachto be considered a viable alternative to BT; and(b) For many smaller customers, whilst there is some evidence of competitionbetween the Parties, they are not close competitors, and a number ofalternative suppliers will remain post-Merger.6.With regard to the vertical concern, the CMA found that the Merger would notenable BT to foreclose third parties from supplying turret systems, given that anumber of alternative turret system suppliers would remain post-Merger whichthird parties could switch to.7.In light of the above, the CMA believes that the Merger does not give rise to arealistic prospect of a substantial lessening of competition (SLC) in the supplyof turret systems (and associated maintenance and support) used in voicetrading applications in the UK.8.The Merger will therefore not be referred under section 33(1) of the EnterpriseAct 2002 (the Act).2

ASSESSMENTParties9.BT is a provider of communications products and services to consumers,businesses and the public sector. BT also sells wholesale products andservices to communications providers in the UK and around the world. Globally,BT supplies managed networked IT services to businesses and the publicsector. The turnover of BT in the financial year 2015/2016 was around 18.9billion worldwide and around 14.8 billion in the UK.10.IP Trade is a provider of turret systems for voice trading and command andcontrol applications. It is based in Liège, Belgium. The turnover of IP Trade inthe financial year 2015 was around [ ] worldwide and around [ ] in theUK. 1Transaction11.BT, through its subsidiary BT (Netherlands) Holding BV, intends to acquire IPTrade for an enterprise value of [ ] million (subject to price adjustments inaccordance with a Share Purchase Agreement signed on 19 January 2017).Jurisdiction12.As a result of the Merger, the enterprises of BT and IP Trade will cease to bedistinct.13.The Parties overlap in the supply of turret systems for use in voice tradingapplications in the UK, with a combined share of supply in excess of 25% bynumber of turret positions. The CMA therefore believes that the share ofsupply test in section 23 of the Act is met.14.The CMA therefore believes that it is or may be the case that arrangementsare in progress or in contemplation which, if carried into effect, will result inthe creation of a relevant merger situation.15.The initial period for consideration of the Merger under section 34ZA(3) of theAct started on 7 March 2017 and the statutory 40 working day deadline for adecision is therefore 4 May 2017.Finalised 2016 turnover figures for IP Trade were not available at the time of the Parties’ submission of theMerger Notice.13

Counterfactual16.The CMA assesses a merger’s impact relative to the situation that wouldprevail absent the merger (ie the counterfactual). For anticipated mergers theCMA generally adopts the prevailing conditions of competition as thecounterfactual against which to assess the impact of the merger. However,the CMA will assess the merger against an alternative counterfactual where,based on the evidence available to it, it believes that, in the absence of themerger, the prospect of these conditions continuing is not realistic, or there isa realistic prospect of a counterfactual that is more competitive than theseconditions. 217.In this case, there is no evidence supporting a different counterfactual, andthe Parties and third parties have not put forward arguments in this respect.Therefore, the CMA believes the prevailing conditions of competition to be therelevant counterfactual.Frame of reference18.Market definition provides a framework for assessing the competitive effects ofa merger and involves an element of judgement. The boundaries of the marketdo not determine the outcome of the analysis of the competitive effects of themerger, as it is recognised that there can be constraints on merger parties fromoutside the relevant market, segmentation within the relevant market, or otherways in which some constraints are more important than others. The CMA willtake these factors into account in its competitive assessment. 319.Turret systems are specialised desktop phone devices with a switch to controlcommunications to and between desktop devices. They combine phone andintercom capability with multiple speakers. Turret systems are used in twoend-use applications: voice trading (where the Parties overlap) and commandand control (where the Parties do not overlap). 420.Turret systems are often integrated as part of a customer’s other internal andexternal communications systems, such as instant messaging, chat, desktopMerger Assessment Guidelines (OFT1254/CC2), September 2010, from paragraph 4.3.5. The MergerAssessment Guidelines have been adopted by the CMA (see Mergers: Guidance on the CMA’s jurisdiction andprocedure (CMA2), January 2014, Annex D).3 Merger Assessment Guidelines, paragraph The latter involves the use of turret systems by command centres of organisation dealing with public safety,utilities, transportation, government and industry, where turret systems are used to coordinate time sensitive,event-driven response in, for example, air traffic and emergency response operations. In the UK, IP Tradesupplies turret systems for both voice trading and command and control applications whilst BT only suppliesturret systems for use in voice trading applications.24

sharing and audio and video conferencing (sometimes collectively referred toas the customer’s ‘unified communications’ systems).21.Turret systems used for voice trading are incorporated into a wider installedsolution encompassing voice recording, private wires (‘always-on’ point-topoint lines that connect traders to their most frequent business contacts), and,if required, a separate intercom facility. Together, this is known as a ‘TraderVoice Solution’.Product scope22.The CMA found that the Parties overlap in the supply of turret systems usedin voice trading applications. The CMA investigated whether it wasappropriate to further delineate this frame of reference with regard to:(a) the different types of turret systems currently supplied to customers;(b) the different types of customer that purchase these systems;(c) the supply of maintenance and support services to existing customers ofturret systems; and(d) the supply of installation and integration services.23.Each of these possible delineations is discussed further below.Possible delineation by type of turret system24.The Parties submitted that there are three different types of turret systemcurrently supplied to customers:(a) Hardware turret systems: a physical desktop turret that connects to ahardware switch (currently supplied by BT and IPC);(b) Hybrid turret systems: a desktop or ‘soft’ (ie PC application) turret thatconnects to a software switch (which runs on servers located at thecustomer’s site or in off-site data centres) and can be integrated into thecustomer’s existing communications systems (currently supplied byEnepath, IPC, IP Trade, Speakerbus, Wesley Clover and Unify). 5The CMA understand that Wesley Clover has minimal presence in the UK. As the CMA has assessed the effectof the Merger within the UK (see paragraphs 37 to 41 below), this supplier is not taken into account in thecompetitive assessment.55

(c) Cloud turret systems: a turret which runs on a desktop device andconnects to other communication systems via the cloud (currentlysupplied by Cloud9 and Greenkey).25.The Parties submitted that all types of turret system should be treated asbeing part of the same product frame of reference.26.The CMA’s investigation confirmed the differences in functionality betweenthe three types of turret systems used in voice trading applications, asdescribed by the Parties. 6 The CMA also found that customer preferences forthese different types of turret systems varied. Whilst some customerconsidered all three types of turret systems to be substitutable, othercustomers believed that cloud turret systems did not currently offer the fullfunctionality of a hardware or hybrid turret systems nor were as proven asthese devices. 727.The CMA notes that if it were to establish frames of reference delineated bythe type of turret system described by the Parties, the Parties would in eachcase not overlap and thus such an approach would not be the mostconservative as regards identifying potential competition concerns.28.On a cautious basis, the CMA has therefore included all types of turretsystems used for voice trading applications within the same frame ofreference, but has taken into account differences in the respectivefunctionality (and substitutability from the customer’s perspective) of differentturret systems in the competitive assessment.Possible delineation by customer type29.The Parties submitted that a distinction could be made between the size andgeographic scope of customers and their turret system preferences.However, the Parties considered that no delineation between different typesof customers was necessary for the purposes of defining the product frame ofreference.30.The CMA’s investigation found that customer preferences for different types ofturret system (and/or for different turret system suppliers), could varyThe CMA’s investigation found that customers (and competitors) were not typically familiar with the terms‘hardware’, ‘hybrid’ and ‘cloud’, suggesting that these terms are not standard within the industry. However, theCMA’s investigation found that third parties were familiar with the key differences in functionality thatdifferentiated these three types of turret system, as described by the Parties in the preceding paragraphs. Forease of reference, the CMA has therefore adopted the terms ‘hardware’, ‘hybrid’ and ‘cloud’ in this decision.7 The CMA’s investigation also found that, for a small number of customers, and typically for a subset of userswithin the customer, a Private Branch Exchange (PBX, a private telephone network) may also be considered asubstitute to turret systems. However, as PBX systems were not considered substitutes by the majority of thecustomers, the CMA has not considered widening the product frame of reference to include these products.66

according to the size and nature of the customer’s business. In particular, theCMA’s investigation was consistent with the Parties’ submission, insofar ascustomers that required a large number of turret system positions stated thatonly the largest market players (namely IPC and BT) had the scale necessaryto meet their requirements. In contrast, the CMA found that smaller customerswere more likely to identify both Parties as potential suppliers (alongside anumber of other alternatives).31.However, the CMA found that there was no clear threshold (in terms ofnumber of positions required) at which customer preferences changed.Therefore, adopting a cautious approach (on the basis that the degree ofoverlap between the Parties is less significant for certain customer segments),the CMA did not further segment the frame of reference by customer size, buttook into account the Parties’ relative competitiveness vis a vis the size of thecustomer in the competitive assessment.Possible delineation between the supply of turret systems and their maintenance andsupport32.The Parties submitted that maintenance and support of turret systems aretypically provided by the supplier of the turret system (whether themanufacturer or reseller) and this was confirmed by customer responses tothe CMA’s investigation. The CMA therefore did not establish a separateframe of reference for maintenance and support of turret systems.Separate frame of reference for the installation and integration of turret systems33.The CMA understands that, in some cases, a customer will require theservices of a third party to install the turret system and to integrate the turretsystem with the other elements of a Trader Voice Solution and the customer’swider unified communications systems.34.The Parties submitted that, where the customer has purchased a BT turretsystem, BT (or its contracted distributor or service agent) will provide theinstallation of the turret system and BT may also, occasionally, oversee theinstallation of the entire Trader Voice Solution. 8 By contrast, the Partiessubmitted that IP Trade does not provide installation/integration services,although resellers of IP Trade turret systems typically do.8 The Parties submitted that it was rare for BT to oversee the installation of the entire Trader Voice Solution.However, BT will in most cases procure the voice recording capability element of the Trader Voice Solution whenalso providing the turret systems.7

35.As part of its investigation, the CMA explored whether vertical concerns couldarise as a result of the Merger given that, pre-Merger, IP Trade supplies turretsystems to resellers, which may compete with BT for the supply of turretsystems and the installation/integration of those turret systems within TraderVoice Solutions. For the purpose of its assessment of these potentialconcerns, the CMA therefore established a separate downstream productframe of reference for the installation and integration of turret systems as partof a Trader Voice Solution.Conclusion on product scope36.In light of the above, the CMA has assessed the potential impact of theMerger on the supply of:(a) turret systems (and associated maintenance and support) used in voicetrading applications; and(b) the installation and integration of turret systems as part of a Trader VoiceSolution (Turret Integration Services).Geographic scopeTurret systems37.The Parties submitted that the geographic frame of reference for the supply ofturret systems used in voice trading applications is worldwide, on the basisthat the same products are sold around the world and that many customers(particularly large customers) purchase these products on a global basis.38.The CMA’s market investigation found that some customers with globalrequirements procured turret systems on a global basis, or at least requiredthe ability to use the same supplier’s product across its global operations.However, responses received from customers also indicated that the strengthof individual suppliers varied across different global regions, and that only twosuppliers in the UK (namely BT and IPC) had the geographic reach to serveglobal clients.39.On this basis, the CMA believes that the supply of turret systems used invoice trading applications may be global. However, on a cautious basis, theCMA has assessed the potential impact of the Merger on competition for thesupply of turret systems used in voice trading applications in the UK.8

Turret Integration Services40.The CMA’s investigation found that a supplier of Turret Integration Serviceswill need to have a national presence in order to serve customers as it willneed staff on the ground to provide the Turret Integration Services at thecustomer’s offices. The CMA has therefore assessed the potential impact ofthe Merger on competition for the supply of Turret Integration Services in theUK.Conclusion on frame of reference41.For the reasons set out above, the CMA has considered the impact of theMerger in the following frames of reference:(a) the supply of turret systems (and associated maintenance and support)used in voice trading applications in the UK; and(b) the supply of Turret Integration Services in the UK.Competitive assessmentHorizontal unilateral effects in the supply of turret systems42.Horizontal unilateral effects may arise when one firm merges with acompetitor that previously provided a competitive constraint, allowing themerged firm profitably to raise prices or degrade quality on its own andwithout needing to coordinate with its rivals. 9 Horizontal unilateral effects aremore likely when the merger parties are close competitors.43.The CMA assessed whether it is or may be the case that the Merger hasresulted, or may be expected to result, in an SLC in relation to unilateralhorizontal effects in the supply of turret systems (an associated maintenanceand support) used in voice trading applications in the UK.Shares of supply44.9The CMA’s analysis, based on data provided by the Parties and theircompetitors found that the Parties’ combined share of supply of turret systemsused in voice trading applications was between 20% and 30% with anMerger Assessment Guidelines, from paragraph

increment of below 5% on any plausible frame of reference, whether on a UK,EEA or worldwide basis. 1045.The CMA’s analysis also found that:(a) IPC was the largest supplier of turret systems used in voice tradingapplications by a very significant margin, whether assessed on a UK, EEAor worldwide basis (market share [60-70]%); and(b) a number of smaller market players, namely suppliers of hybrid turretsystems (eg Enepath, Speakerbus and Unify) and cloud turret systemsuppliers (eg Cloud9 and Greenkey), accounted for the remaining shareof supply.46.The CMA’s analysis also found that:(a) for customers requiring a smaller number of positions (eg up to 100250), 11 the Parties’ combined share was slightly lower (between [10-20]%and [20-30]% in the UK), with IPC retaining the majority of positions andother suppliers responsible for the remainder (in particular Speakerbusbut also cloud turret system supplier, Cloud9); and(b) for customers requiring a larger number of positions (eg above 250), theParties’ combined share was slightly higher (between [20-30]% and [2030]% in the UK) albeit that the increment arising from the Merger wasminimal (or non-existent); with IPC mostly accounting for the remainder.Closeness of competition47.The Parties submitted that they were not close competitors, on the basis ofdifferent customer focus and product offering. The Parties submitted that over[the vast majority] (by revenue) of BT’s bids in 2015/16 was accounted for bysales to large customers. 12 The Parties submitted that IP Trade lacked theglobal scale necessary to compete realistically for these customers, with allDue to the way in which customer data is stored in IP Trade’s systems, UK position numbers submitted for IPTrade include positions in the Republic of Ireland, Jersey and Guernsey, and therefore overstate IP Trade’spositions in the UK, while EEA position numbers submitted by IP Trade include continental Europe, UK, Russiaand Switzerland. This note applies equally to the data used for the share of supply calculations in paragraphs 45and 46.11 The Parties made a notable distinction in their submissions between customers with over 1000 positions (witheither global or multi-national presence), and customers with less than 1000 positions (and generally nationalpresence). While the CMA’s investigation did not identify a clear cut-off point (in terms of the number of positions)at which customer preferences changed, the CMA considers that a lower number (in the hundreds, rather thanthousands) is likely to be a better reference point for distinguishing between larger and smaller customerpreferences. Accordingly, in its assessment of shares of supply discussed further in paragraph 46, the CMA usedtwo alternative reference points for its analysis: (i) customers with above/below 250 positions and (ii) customerswith above/below 100 positions.12 Which it defined as those typically with over 1000 positions and multi-national or global presence.1010

but one of the contracts won since 2012 being for customers with less than100 positions. For smaller customers, the Parties submitted that growingcustomer preferences for hybrid turret systems meant that BT (which does notoffer a hybrid turret system) was at a competitive disadvantage to hybrid turretsystem providers, such as IP Trade.48.The Parties submitted that the limited competition between the Parties wasdemonstrated through the Parties bidding data, which showed that the Partieshad only competed for the same customers on a small number of occasions inthe past five years.49.The CMA assessed closeness of competition on the basis of:(a) bidding data provided by the Parties, customers and other turret systemsuppliers;(b) views received from customers and other turret system suppliers; and(c) the Parties’ internal documents and industry reports.Bidding data50.The CMA’s analysis of the bidding data provided by the Parties, customersand other turret system suppliers indicated that the Parties did not competehead to head for the vast majority of tender episodes in which either Partyparticipated. 13 Where the Parties have competed, in almost all cases theyhave faced at least one other competitor and often two or three (most oftenIPC, but also others including Enepath and Speakerbus). 1451.The CMA found that:(a) on the basis of bidding data submitted by BT, the Parties only competedhead-to-head in [a very small proportion] of all tenders ([ ]) in which atleast one Party participated. Of these, the Parties faced at least one othercompetitor (always IPC) in over half of the bids, and faced either two orthree competitors (including IPC, Enepath and Speakerbus) in a furtherthird;(b) on the basis of bidding data submitted by IP Trade, the Parties onlycompeted head-to-head in [a very small proportion] ([ ]) of all tenders inUK tender data submitted for IP Trade also included data on the Republic of Ireland, Jersey and Guernsey;see footnote 10.14 The bidding data also, in a small number of instances, listed PBX suppliers Avaya and Cisco as competingbidders. As noted in footnote 7, PBX systems may consider a substitute to turret systems for some customers.1311

which one Party participated. Of these, the Parties faced a single othercompetitor (IPC) in all cases; and(c) on the basis of bidding data submitted by a number of competitors (whichthe CMA notes was less extensive than that submitted by the Parties), theParties competed head-to-head in [10-20]% ([ ]) of all tenders in whichat least one Party participated. Of these, the Parties faced at least oneother competitor in just under half of the bids, and faced either two orthree competitors in the remainder of bids.Third party views52.Of the responses received from BT customers, the vast majority of customersdid not consider IP Trade to be a substitute for BT. In particular, largercustomers noted that IP Trade did not have the scale to supply and supporttheir turret system requirements, while others (both larger and smallercustomers) were either not familiar with the IP Trade product, or consideredthe functionality of the BT turret system to be superior. Those customers thatdid consider the Parties to be close substitutes were typically smallercustomers. As discussed further below, these customers each listed a numberof alternative suppliers of turret systems alongside BT and IP Trade.53.The majority of customer responses received from IP Trade customersindicated that BT would be a viable alternative. However, the majority of thesecustomers further stated that compatibility with their existing unifiedcommunications systems was a key requirement for their turret systems,which BT did not offer as a provider of hardware turret systems. One IP Tradecustomer raised concerns that the price of IP Trade’s turret system may risepost-Merger. However, given that this customer highlighted the differencesbetween BT (and IPC) on the one hand, and IP Trade on the other, in terms ofprice, service levels and customer focus, the CMA believes that the Partiesare unlikely to have been particularly constrained by each other’s pricing preMerger.Internal documents and industry reports54.BT internal documents and market reports submitted by the Parties supportedthe view that BT and IP Trade were not close competitors. These documentstypically distinguished, on the one hand, between the larger, more establishedplayers BT and IPC and, on the other, smaller players such as IP Trade, Mitel,Speakerbus, and Unify. For example:(a) a series of internal competitor studies prepared by BT focus on thecompetitive dynamic between BT and IPC and a comparison of their12

respective product offering first, before assessing in more summary formthe offering of players including Enepath, IP Trade, Mitel, Speakerbus andUnify; 15(b) one of these competitor studies examines BT’s wins and losses in2015/16, identifying [ ] losses to IPC (including [ ] to Etrali), 16 against 1loss to IP Trade; 17(c) a 2016 industry report prepared by a technology consultancy firmdescribes on the one hand IPC, BT and Etrali as the ‘dominant, multiregional’ players and the ‘big three’ voice trading technology and networksuppliers, while describing IP Trade and Speakerbus together as newentrants and ‘market challengers’. 18 The report notes further that the turretsystems market has been largely static for the preceding five years, with‘little shift of market share between the traditional technology and serviceproviders (and new market entrants)’, underlining the persistent gapbetween these two groups of turret systems suppliers; 19 and(d) a report prepared by a third party (which BT contributed to) groups on theone hand BT and IPC as turret systems providers with more advancedturret expertise, and on the other, IP Trade and Speakerbus as ‘low costturret providers’ with lesser turret expertise. 20 The report also indicatedthat IP Trade was not competing closely with BT (and IPC), stating that IPTrade lacked the geographic coverage to significantly gain market sharefrom these players. 21Competitive constraints55.The CMA assessed what constraints would remain on the Parties postMerger.56.The Parties submitted that, following the Merger, BT’s strongest competitorwould remain IPC, the largest provider of turret systems. The Partiessubmitted that IPC and BT were the only players with the scale and scope tocompete for large, multi-national competitors and that the Merger would allow[ ].IPC acquired Etrali Trading Solutions in February 2016.17 [ ].18 Wesley Clover was also listed in this latter category. However, as noted in footnote 5, the CMA understandsthat Wesley Clover has minimal presence in the UK and is therefore not relevant for the CMA’s assessment ofcompetition within the UK.19 [ ].20 [ ]. Wesley Clover was also listed in this latter category. However, see footnote 18.21 Silverfleet Capital, Etrali strategic analysis: Final Report, July 2015.151613

BT to better compete against IPC for this customer segment by allowing BT tomatch IPC’s ability to provide both hardware and hybrid offerings.57.For smaller customers who did not require a provider with global scale, theParties submitted that post-Merger there would remain a number ofcompetitive alternatives. The Parties submitted that these customers

11. BT, through its subsidiary BT (Netherlands) Holding BV, intends to acquire IP Trade for an enterprise value of [ ] million (subject to price adjustments in accordance with a Share Purchase Agreement signed on 19 January 2017). Jurisdiction . 12. As a result of the Merger, the enterprises of BT and