It was a clash of corporate philosophies from the first overtures of the merger. Over many years, the Halifax building society had established a fiercely independent view of how its IT operations should be managed: as a strategic weapon, a means to business responsiveness, something that should be tightly controlled from within the organisation. The approach by its merger partner Bank of Scotland could not have been more different. With the aim of reducing its IT overheads by around £15 million a year, the bank had decided in 2000 to pass the running of its IT management to IBM¹s Global Services organisation, as part of a £700 million contract that was due to run until 2010.

Perhaps they were not confident of seeing that promised payback, or perhaps they saw a requirement for the kind of flexibility that in-house IT offers, but director of group programmes Ian Kerr, who led the IT aspects of the September 2001 merger, and other senior management at the Halifax Bank of Scotland (HBOS), decided its days with large-scale outsourcing were over.

Ripping up its contract with IBM, as well as a four-year-old joint venture agreement with UK IT services company Xansa, HBOS began the year-long repatriation of the Bank of Scotland’s IT department in June 2002. When that decision was announced, it prompted questions among many other IT executives and consultants about the desirability of so-called mega-outsourcing deals. There were also concerns about whether some companies, under extreme budgetary pressure, had rushed into outsourcing contracts and were now failing to see either savings or efficiency gains.

HBOS may be an extreme case. Halifax by repute is so possessive of its IT that the company is known to have even written its own compilers. But the HBOS ‘backsourcing’ move is not unique. The Norwich Union insurance group, as it came together with CGU to form CGNU (now Aviva) in 2001, scrapped its primary £124 million, seven-year outsourcing deal two years in, again throwing out IBM Global Services, and bringing its IT operations back in house.

While the HBOS and Norwich Union ‘terminations’ were clearly merger-driven, there is a growing body of evidence that suggests that many other organisations ­worried that outsourcing is not working for them in both financial and strategic terms­ are reassessing their commitments.

Ready to jump

“It’s not just at the consideration stage, we are seeing organisations that have already made the decision to jump,” says Frank Philbin, associate director for implementation at Comunica, a communications infrastructure specialist with a focus on the financial services sector.

He qualifies that. “Typically it is not the whole outsourcing contract that is being scrapped but certain aspects. In particular, they are abandoning aspects that involve critical services, areas where the levels of risk to the business from outsourcing is very high.”

The conclusion companies are coming to: “Outsourcing can be a risk reducer, but it can also be a risk increaser. And organisations are particularly sensitive to risk at this moment.”

So are many boards of directors in the financial sector reassessing the contracts that they currently have? “Absolutely, absolutely. Decisions have been made over the past six to 12 months and are being brought into operation very quickly,” says Philbin. During 2003, a number of companies will tear up their contracts and in some cases then transfer to new organisations that can guarantee higher quality, he adds.

His perceptions are echoed elsewhere. Siki Giunta, CEO of service management software company Managed Objects, has been working with a large telecoms company that is coming to the end of a five-year contract and has no intention of renewing. “It is a complete turnaround. They actually see it as cost effective to bring it back in house,” says Giunta. She suggests that the margins in such contracts have grown larger over the past two years as service companies have come under financial pressure. As a result, customers are receiving poorer service.

One example that Philbin cites is in storage management. In recent months, he has seen a large company bring the on-going management of its storage area network back in house. The feedback from the user: not only was the outsourcing company not delivering the required quality of service, but the customer also felt a lack of control.

Despite that changing picture, for most organisations the notion of backing out of an outsourcing contract is unthinkable. In some cases, that has led to organisations feeling that they are being held hostage by their commitment to a single outsourcer. One look at how the UK’s Inland Revenue is struggling to open up fresh tendering for the management of its IT operations is enough to confirm that. Any notion that the incumbent, EDS, will lose the contract when it comes up for renewal in 2004 is treated with derision by services sector analysts.

“Of the organisations that aren’t satisfied with their outsourcing contract, only 5% to 10% could even contemplate taking it back in house because of the trauma and cost involved in doing so,” says Stuart Payne, principal consultant with outsourcing consultancy Morgan Chambers.

The feasibility of re-constituting the IT department­ or even parts of that operation ­is clearly hampered by the dispersal of the skill base, and in many cases, the disposal of the organisation’s data centres and other facilities. Staff are literally caught in the middle. Passed on to the service provider by their former employer, they are often unwilling to rejoin. But they do usually have that legal right. Under UK TUPE (Transfer of Undertakings­ Protection of Employment) regulations, anyone at the outsourcing group who has been spending more than half their time on that account has the option of going back.

‘Backsourcing’ bugs

Areas of concern for organisations seeking to take IT back in house.

TUPE ­ Under UK employment law (the Transfer of Undertakings ­ Protection of Employment Act 1981) all employees who spend more than half their time working on an individual outsourced contract have the right to join the ‘backsourcing’ organisation. But will the company want them all and how much will they cost?

IP ­ Who owns the intellectual property developed by the outsourced company on behalf of the organisation?

Kill-fee­ How much will the termination charge be? Depending on the age of the contract, this could be as much as 5% to 10% of the contract value.

Transition period ­ How long will it take to move the systems back to the organisation? In the case of HBOS, that process will take at a year.

Software licences ­ How will the company deal with the cost of acquiring software? The outsourcer is likely to have benefited from various volume discounts that will not be available to the organisation. Moreover, the supplier will charge the organisation for any of its proprietary software that is required in the running of the organisation’s IT operations.

David Lister, CIO at Boots, portrays it as a delicate issue. “You have two options: bringing them back in or moving them on. Bringing them back in has to be for keeps and you would need to have a very convincing story to the people about what’s different this time around, otherwise you are going to bring back in a whole raft of distrust. Additionally, you are going to lose people because a lot of people that have gone to an IBM from a Boots have come to see themselves as an IT professional working for a professional IT firm,” says Lister.

However, according to Comunica’s Payne, repatriation of key personnel in the event of contract termination is now a feature of most outsourcing contracts, whether that is moving them back into the organisation or to another third-party supplier.

But it is not the desire to have those staff back that is prompting most backsourcing decisions. These are driven by the degree to which outsourcing has fallen short of business requirements and, indeed, expectations. “There are plenty of outsourcing contracts which are non-performing,” reckons Payne. Many of those are what he calls ‘golf-course deals’, contacts that were clinched with an executive handshake, and lacked adequate due diligence. “Those are the ones that are going pear-shaped,” says Morgan Chambers¹ Payne.

Get-out clause

But how can organisations structure deals so that elements can be taken back in house if the need arises? That was certainly on Lister’s mind recently when drafting Boots’ outsourcing strategy. In a deal with uncanny parallels to the Bank of Scotland’s contract, Boots agreed in September 2002 to pay IBM Global Services a total of £710 million over 10 years for the IT services organisation to handle all of the retailers’ IT infrastructure, including its data centres and networks, in-store systems, telecoms, databases and applications maintenance. At the same time, Boots opted to pass applications development services to Xansa in a £90 million, seven-year deal.

These services contacts are on a grand scale and involve the management of 3,000 servers, 10,000 desktops and several mainframes, with 400 staff transferring from Boots’ Information Systems and Technology group to IBM and 200 joining Xansa. The Boots senior management team has mostly focused on the cost savings. The company thinks it can save £130 million in projected IT spend over the life of the IBM contract and £30 million from the Xansa deal.

But, if those savings do not materialise, there has to be a get-out clause. “Absolutely any contract has to define how termination would come about and how you would manage those elements back [in house],” says Lister. “You need to make sure it is clear [to the outsourcer] that it is an option, but also with the hope that you never have to use it.”

The majority of contracts today include provision for a ‘reversal plan’, says Philbin. “That was not the case a few years ago.”

Lister has some experience here. In 1998 at Guinness Brewing in the UK, where he was group IT director, he made the decision to bring service and applications management back in house from Computer Sciences Corp (CSC), dissolving a contract that had been running for over two years. The reason? “There was no differentiation in capability coming from CSC. It did not seem to be any different from the capability that had been outsourced years earlier.”

Secondly, the contract was proving more expensive than an in-house alternative. “We realised we could achieve the outcome we wanted using our own internal resources ­and save money as well,” says Lister. “It was an obvious route to take,” he adds.

It was obvious in this case because of the limited nature of the contract. “Because the contract was set up as a short-term, cost-based transaction, CSC had not really done anything other than take our people on and hire their services back to us at cost plus,” says Lister. “So bringing the whole team back into Guinness was a relatively easy thing to do as they were basically running ‘as was’.”

He contrasts that with the outsourcing model set in place with IBM. “It is one of shared benefit, working towards a long-term outcome,” says Lister.

“At Guinness, with CSC the attitude was ‘do what I do now and carry on doing it’. The Boots deal with IBM involves a change plan that will be executed over seven years. And IBM’s role in that is to deliver a whole series of defined elements. So in order to deliver that IBM will need to change people, develop people, and bring in new people.”

That approach shows a striking difference with many contacts set up in the mid-1990s. Those were mostly centred on managing existing operations at a reduced cost, and specified nothing about how those systems would change, says Lister.

Whose innovation?

The contract between IBM and Boots highlights another area where organisations and their outsourcers can fall out. Boots is the first company in the UK to sign up for IBM’s Collaborative Innovation Programme under which the two organisations will work on exploiting emerging technologies, such as wireless and mobile systems, to build new applications.

That does, however, create issues over intellectual property. In an outsourcing deal, when the IT service company is developing unique systems or enhancing the organisation’s processes on behalf of the client, there will be grey areas over who owns the resulting innovation.

One major pharmaceutical company, for example, which outsourced its IT in 2001, is reportedly finding pockets of resistance. Management in certain parts of the company have reportedly become concerned that the outsourcing company, which is running its specialist analysis software on the pharmaceutical company’s data, is claiming it owns some of the resulting intellectual property. As that involves the critical area of drug discovery, alarm bells have started to ring and the pharmaceutical company has recently bought a series of high-end Unix servers in order to again run part of its own drug discovery programme in house. Outsourced development was certainly a concern for Boots. “We spend a lot of time thinking about the issue of how the outsourcing deal might impact our ownership of intellectual property,” says Lister.

There are certainly grey areas, he says, and an outsourcing contract has to be able to deal with them. “We wanted strong protection, on things that would be specific to Boots, such as a new way of targeting customers,” he says. “But if, for example, IBM invents a new kiosk for us, we would regard it as something that could be duplicated by another developer, and would allow IBM to exploit that innovation after a year or so,” says Lister.

It all comes down to obtaining an optimal balance between insourcing and in-house IT, says Philbin. “In some cases companies see it as lower risk going with an outsourcer, with others it’s a riskier bet. And balancing those is the difficulty, particularly when markets are so volatile.”

How do organisations decide what elements are desirable to outsource, and which are strategic and need to be held onto? There are aspects, that should never be outsourced, says Lister: in particular, control over IT strategy and IT architecture ­ things that shape and maintain the outcome of the business plans. However, he adds: “There is a logical sequence here. First and foremost, you have to think about your desired outcomes:­ what is it that you actually need in three, five or seven years? Second, the question is one of capability:­ what capabilities do you require to deliver the outcome? Third is timing: ­ can you develop the capability yourself, in time, at an acceptable cost, or are there alternative sourcing models that can deliver those particular capabilities? Whether you go with an outsourcer or use in-house resources, it is all about delivering the outcome,” he concludes.

“If you are clear on the outcome, you are also clear about what you need to build the capability to deliver it,” says Lister. “You know what the journey looks like. So the sourcing discussion is about how you can achieve that outcome, but to do it faster, better, cheaper.”