There can be few greater challenges for today's IT management than helping the business successfully navigate a course through a major acquisition. The board will want the deal completing as quickly, smoothly and as cheaply as possible. The staff will want the services they depend on to continue uninterrupted.
And there is the rub: more often than not, IT looks like an inhibitor, if not a deal-breaker, as the technical complexities involved in smashing together two radically different IT infrastructures push up acquisition costs and drag out the merger process.
"In some cases, there are good reasons why the technical challenges of doing the integration would actually outweigh the benefits of the deal. But from the IT side, you'd better be very confident in your relationship with the CEO if you're going to tell him that his deal isn't going to happen," says Chris Nottage, CTO for technical strategy at Lloyds TSB, which merged in 1996 and has made several acquisitions since.
That usually means that IT managers simply have to devise as foolproof a formula as possible for integrating the two distinct corporate infrastructures.
Sometimes technology helps. Whole sub-industries of vendors have grown up around application integration tools that try to make the co-existence and integration of packaged applications as painless as possible. But they do little for the stack of bespoke code that exists in most organisations.
Given that most mergers and acquisitions are predicated on the notion that the newly integrated entity will operate with a common IT infrastructure, shifting to that single infrastructure may take years. "Without that, though, the finances for most deals don't stack up," explains Andrew Morlet, a partner at consultancy Accenture.
And it is not just the applications that need to be integrated. There is no short cut for the laborious task of migrating data, with all the associated data profiling, cleansing and transformation. There are further issues to tackle too, such as data centre consolidation, email integration, network configuration and, ultimately, staffing levels and training.
It is clear from the outset of any large-scale acquisition that the entire business is under intense pressure. According to research carried out by Ohio State University, nearly half of companies undergoing merger-related integration suffer productivity losses in the first six to eight months.
Longer term, the outcomes hardly look much better: according to research by management consultancy McKinsey & Co, two thirds of deals result in zero growth over the subsequent three year period.
Sometimes, the integration is impractical or too expensive. Gerry Sheridan, head of HP's IT services unit in the UK, knows of one major company that has still not managed to integrate its email systems four years after a major acquisition.
Such failure, inevitably, increases tensions throughout the business. The relationship between IT and the senior management of other divisions can be severely tested throughout the course of a merger or acquisition.
According to a 2002 Booz-Allen Hamilton study, 42% of CEOs left within two years of closing what was deemed to be a ‘failed merger'; that compares to the 84% of CEOs who were still in place after a ‘successful merger'. And the CEOs' lieutenants fare even worse: more than 60% of all CIOs in failed mergers were gone within 24 months.
First steps
Avoiding that fate requires some careful understanding of merger dynamics. "Once a merger is announced, executive management wants to know two things from IT: How long will it take and how much will it cost?" says Julie Giera, vice president at analyst house Forrester Research. Those questions will usually be accompanied with assurances that the CIO will not be expected to absolutely commit to those initial estimates, explains Giera. After all, the CIO will have little detailed knowledge of the infrastructure of the target company. "Don't believe a word of it. As soon as any dates or dollars leave your lips, expectations are set."
But the nature of the acquisition can also influence the chances of its outcome: there are, in fact, a multitude of factors that can initiate M&A activity – and these directly affect the IT integration strategy (see box, ‘Acquisition strategies').
Once the merger strategy has been decided, IT management are positively encouraged to stick their necks out. Their delivery of a detailed integration plan, even if drafted half-blind, can be instrumental in getting shareholder approval of the deal.
For deals involving absorption, much of the financial planning can be done in the pre-deal stages – the models for data centre and network integration are reasonably mature and well understood. "The speed, the cost savings and the lower risk of absorption make it a very attractive integration choice," says Giera.
|
||
Lloyds TSB underwent exactly this sort of process when its share registration service acquired a similar unit from the Bank of Scotland. "The investment was in market share, we felt we had the stronger systems, therefore it was the natural choice to move the target's systems over to ours," says the bank's Chris Nottage, but that made data migration a priority.
The key to successful data migration is, says Nottage, the ability to develop a common definition of data. In the first instance, this involves looking for shared architectures, business functions and software functions.
Lloyds uses data extraction, transformation and loading (ETL) tools from Ascential Software – now part of IBM – to improve data integrity. "When you're talking about the merger of large business units, with the associated volumes of data, you need to reduce systems to a common set just to be able to operate," says Nottage.
Fusion by-products
On the application side, there are fundamentally three levels of complexity, says Accenture's Morlet: the basic application infrastructure such as emails and desktop operating systems; enterprise resource planning (ERP) and other business applications; and, most complex, the bespoke applications. But speed is of the essence: "We tend to find that the longer businesses leave applications running in parallel the more likely they are to regret it," says Morlet.
Many organisations will have undergone email consolidation programmes as part of a natural update cycle, and while these can still be traumatic – costs can vary between $125 to $500 per mailbox to migrate, according to IT analysis group Ferris Research – it is at least a process that is well understood and therefore manageable.
Likewise, other enterprise applications, such as ERP and customer relationship management, while more complex, are at least relatively common and may even overlap, says Morlet. "Even with two separate SAP implementations, there are likely to be migration issues – they may have different data definitions – but at least there is some commonality.
"For those businesses with a substantial element of bespoke development, integration is inevitably more complex – in some cases it can be a real nightmare," he adds – to the point where some organisations junk both of their existing systems and move to a third. That was exactly what Halifax and Bank of Scotland (HBOS) did over 2002-2003 with its credit card processing services applications, the head of business change for HBOS Card Services, Andy Carter, told a recent Information Age conference.
As that shows, the decision about which application should remain in the organisation is not always straight forward. Back in 2004, Dutch IT services group Atos Origin bought Sema, the IT arm of oil and gas services giant Schlumberger for E1.5 billion, effectively doubling the size of its operations.
That also meant that there was no immediately obvious choice about which company's IT infrastructure would meet the requirements of the new operations, says Xavier Flinois, board member at Atos Origin and former CEO of Sema.
"We were in a privileged position in that we had ready access to a talented IT consulting team, but the process still involved some difficult IT decisions," explains Flinois.
Atos started by identifying what the new corporate structure would look like, then designed an IT ecosystem to best suit the business. Then came the hard nuts-and-bolts task of profiling existing infrastructure – call centres, ERP systems, networks and so on – and deciding, one by one, which was most suitable to retain.
"At that stage, the project team was attached to the management board, so therefore there was good communication and support when it came to doing the heavy lifting of migrating ERP systems and networks," adds Flinois.
As part of the merger, Atos had promised to deliver E200 million in annual savings within two years. "For us, that meant rolling out the new financial system very quickly," says Flinois. "As it transpired, we had comp-leted the financial side a quarter of the way through the entire integration programme."
However, while speed of integration is cherished by CEOs overseeing a merger, there can be pitfalls, says Forrester's Giera. "Many organisations move too quickly to shut down the acquired company's IT organisation, forgetting that access to customer files, transaction histories and other similar data will be crucial in resolving post-integration customer service problems and complaints. CIOs should plan to keep the acquired company's files and systems operating for at least 30 to 60 days post-implementation, and should make accommodations for retention of data should research be required after that."
Another problem that besets many businesses is a rush to optimise IT infrastructures, says Simon Rowland, global head of project management consultancy PIPC, who worked on the IT integration of Royal Bank of Scotland and NatWest. "You need to get the integration complete before you look at optimising systems. Of course there are always plenty of senior managers that are pushing for new application developments the moment the merger is announced. But unless you've got the integration complete, you'll be setting yourself up for trouble."
And trouble needs to be avoided. As HBOS's Andy Carter puts it: "Mergers put the business through a lot of pain." The kind of pain that can make – or break – a career.