All great public corporations have to start somewhere. Take MCI Communications Corp, soon to be wedded to British Telecommunications Plc. Back in 1983 phone companies other than Ma Bell were hardly bigger than seedlings. But what set MCI apart from around that time was the excellence of its customer service function, an asset which proved a powerful aid to its growth when allied to superlative marketing. No-one knew that would happen in 1983; all MCI was sure of was that it had purchased a Wang Laboratories Inc mini based call center automation product involving one of the very first CTI (computer telephony integration) interfaces. This system, the building of which marked MCI as a pioneer in call center activity, helped put together a set of customer databases that a sales agent could access to make MCI’s outward pointing ‘face’ to the customer as unified and friendly as the technology then allowed. The software was the one of the very first products shipped by a then two year old service automation specialist called Early, Cloud & Company, of Newport, Rhode Island. (Early, Cloud – or ECC – is not named, incidentally, for any meteorological phenomena but for two of its founders, Jack Early and Don Cloud.) The success of that early system helped both parties immensely, though ECC never became an $18.5bn giant, but instead a more modest $23m operation when the closely held concern was purchased outright by 25% stakeholder IBM in October 1995. Post-purchase, the company now forms the core of IBM’s customer service and relationship management solution for large (100+ seats) organizations – fitting in with other Call Center offerings such as professional services from ISSC, and IBM’s own CallPath CTI product. Call centers can involve a range of customer contact management systems, including IVR (interactive voice response units), integrated fax, automated dialing capabilities and workflow management systems. CTI’s role is characteristically to link the call center telephone switch to a host for applications like automatically transferring customer information to an agent’s workstation during or prior to a call. In other words, ECC is in the business of helping banks work better with customers – though sometimes that’s just to help ’em sell us more credit cards. We recently caught up with ECC to gauge how it was faring in the embrace of (Less) Big (Than It Was) Blue. Though it would always have been a stretch to think of ECC as anything other than a close camp follower of IBM. Consider the typical ECC customer; after the mid 1980s, for sure, we’re talking large banking and financial services institutions running big call centers off of MVS and ancillary mainframes (and banker sweethearts Tandem, natch). Of its currently active customers (somewhere north of 120, we believe), the largest slice of its 224 product licenses, 65, is for its TCS (Teleservicing Control Systems), its 3270-screen based call center automation product on IBM MVS/CICS. TCS is just one of three ECC product strands, however. Also in its armory is CallFlow, a more modern call center automation software product targeted for CTI-enabled telephone sales operations, which has also spawned CallFlow Client, a desktop, more package-like, product, again, usually aimed at banks. This is now on a non-mainframe platform, though try not to hit the ground with shock when you hear it’s only Big Daddy IBM’s own AIX on RS/6000 at the moment. Probably most well known ECC product (outside the banking vertical applications side) is of course MDp (Message Driven processor), based on IBM’s venerable but still powerful CICS transaction processor. This GUI client/server offering is message-oriented middleware with work flow automation elements allows integration of legacy systems and enterprise-wide client/server, giving rise to a better flow of information between different applications, according to IBM. TCS, CallFlow and MDp had all helped put ECC in a dominant position in its chosen market even before Messrs Early, Cloud and the other private shareholders decided to sell out to IBM 18 months ago. In the words of The Tower Group, a research firm operating out of Newton, Massachusetts, ECC has established itself as the dominant provider of call center solutions to large US banks. Tower thinks Call Centers may be a half billion dollar market last year, growing to $850m by 2000, so IBM/ECC are in the right place at the right time, it seems. Given that IBM/ECC if the only IBM mainframe based call center provider, and that banks therefore face either home-brewing their own solutions or purchasing a boutique system from an Andersen Consulting type systems integrator. Does that make IBM/ECC seem like it’s doing dull old Big Iron stuff? Pas de tout. MDp can conceivably be purchased separately to form the basis for a generic message driven architecture. But more customary has been for MDp to be sold alongside IBM’s message oriented middleware (MOM) product, MQSeries. Banks have for long been seeing the need to pull data from disparate databases into one front-end customer focused application, a la MCI in 1983 – but for most of the 1980s and early 1990s the system that resulted tended to be fairly screen- scraperish. These CIOs are now looking at (and investing in) message driven architectures as the next step forward, apparently finally convinced that they offer a better way to achieve the flexibility promised by classical client/server or n-tier architecture than RPC (remote procedure call) or other database- centric distributed configurations. There is also a kind of birds-of-a-feather trend in banks and MOM; if for no other reason, than that in this era of mega mergers, integration of an acquired bank’s IT may be easier if the basic IT plumbing all kind of looks the same. And seven of the top ten US banks use IBM/ECC products. So goodness abounds? Maybe not wholly. IBM/ECC has now started to officially ship MQSeries server software for free along with the client version, which is a natural progression in a sense – This helps position the two products better – there was some confusion in the marketplace; we want to position MDp as an application enabler for MQSeries, says Group Marketing Manager Mark Desrosiers. However, it does raise the question of how well the Early, Cloud part of ECC is doing in the febrile grip of Lou Gerstener’s mob. In Tower’s words, ECC has maintained an independent identity in the marketplace but is moving to integrate its solutions with related IBM technology and professional services organizations, a move which is causing a few creaks in the rigging. IBM has been successful in retaining ECC’s staff by maintaining the independent identity and management of ECC. ECC’s knowledge resources, however, will be stretched thin as IBM’s worldwide organization generates demand for ECC’s solutions in banking as well as other industry segments …it will be very difficult to spread ECC’s limited resources across a rapidly growing number of sites during the next two years. ECC also faces a challenge retaining its flexible, market driven entrepreneurial spirit while leveraging IBM’s considerable resources, thinks the wise thinkers over at Tower. ECC must now support other IBM products (eg VisualAge Smalltalk, OS/2 and RS/6000 AIX, etc) at the same time it plans to support non-IBM products like Visual Basic, Windows NT and Unixes from HP and Sun. But support …will significantly lag the IBM-based products and will likely not be as strong because of the IBM orientation of the sales and support organization. One particular pressure point: Windows NT, which IBM/ECC’s fave customers are increasingly wanting to put in their network diagrams, if only so far at the desktop level. ECC says there is an NT client product on the way, but not until June at the earliest. And though ECC is well placed at the high end, it stands much fiercer competition in the lower tiers of financial institutions, where Unisys, Olivetti, AIT in the UK, and Hogan Systems all stand baring their fangs at the reserved Rhode Islanders. So is ECC in trouble? Our culture was one of the aspects of the way that we work that IBM knew all about before it purchased us. We’re fairly independent and that’s been sustained, assures Desrosiers. But ECC is growing at over 30% per year (now up to $30m revenue, with 27% from international), and some customers feel the support has suffered, according to Tower. So maybe ECC isn’t in trouble – but is in danger of being spread too thin. This resource shortage has only been exacerbated as ECC trains IBM support resources in all locations, though to be fair it’s hard to see anybody doing an egregiously better job in this space. It’s ironic that one of the reasons ECC gave when it agreed to IBM’s offer was that it felt it needed more gas in its tank to grow at more than the 20% growth rate it was then achieving, hence the assimilation with the Borg in Armonk. And yet it may still falter as a result of having to support products which must inevitably, at least sometimes, prove to be less than best of breed but must be championed simply because they come from guys whose names are in the same company directory. But if MQSeries continues to steamroller along – the handwaving and the bleats of the ‘other’ MOM vendors like Momentum and BEA notwithstanding – the ECC part of the IBM/ECC zygote seems well placed for the time being.
