The massive move to outsourcing presents new challenges for organisations -- public and private. Here are 10 lessons learned from a comprehensive study of outsourcing in WA It was a decision the fictional mandarin Sir Humphrey Appleby, from the television comedy series Yes Minister, might have described as "courageous".
Somebody had to go first. Pioneers can be brave people with little knowledge and little experience, but in order to survive they soon learn very fast. In this case, a "pioneering spirit" saw a government agency with little understanding of the business requirements, and little experience in putting an outsourcing contract together, locked into a long-term contract that proved to be far from best practice. On the positive side, the trailblazing agencies in questions were probably the quickest learners in the industry. When the Western Australian government directed state agencies to focus better on "core activities" by outsourcing "non-core activities" to the private sector, it was pushing those agencies out to a kind of No Man's Land where few -- public or private sector -- had gone before. In fact, in early 1993 the WA government became one of the first in Australia to sign a long-term outsourcing contract, after one WA government department entered into a six-year fixed-price agreement with an outsourcing vendor.
That decision to push agencies out to the bleeding edge of outsourcing (at that time) should prove to be good news for others in both the public and private sector, following release of an in-depth study of six of the organisations that outsourced a significant portion of their IS/IT requirements. Author of the study Stephen Walsh says it sheds light on the reasons why some outsourcing arrangements either are unsuccessful or fail to achieve the expected benefits.
Walsh recently joined Westpac Banking Corporation's IT Division as an executive manager and worked for EDS for 11 years before beginning the work on his Masters, which underpinned the study. By identifying and addressing reasons for the success or failure of outsourcing arrangements in the public sector, Walsh has elicited some important lessons about how outsourcing should be done, which are every bit as applicable to the private as the public sector. His research project sought to gather and understand the opinions and perceptions of IS/IT managers and user managers in the Western Australian public sector. Walsh conducted semi-structured interviews with IS/IT managers, user management and outsourcing vendor executives involved in the outsourcing of the six organisations, focusing on the decision making process, the evaluation and tender process and contract negotiation and execution.
He also studied two major contracts in depth. One -- an open book and partnership arrangement signed in early 1996 between a number of public sector organisations and "Outsourcing Vendor 1" for provision of mainframe computing services -- had been running for more than 12 months at the time of the research interviews. There was an incentive-based pricing model and a pre-agreed set of rules to determine the pricing and budgets for services provided. Service levels and performance metrics focusing on customer satisfaction and traditional IS/IT metrics were included in the contract. The second -- a six-year fixed-priced agreement signed in early 1993 between government department "X" and "Outsourcing Vendor 2" -- embraced support for existing mainframe applications, running of mainframe data centres, replacement of existing mainframe applications and implementation of new systems on a client/server environment. There were no incentives for better performance, but the contract includes penalties if certain levels of performance are not met for the existing mainframe systems. No new functionality or additional business benefits are to be included in new systems. And it is a contract that -- Walsh's research suggests -- did not run at best practice.
"Both parties didn't understand the full business requirements; they had limited experience in putting together an outsourcing contract which was very inflexible. They entered into a fixed-price, fixed-delivery, long-term contract which definitely was not best practice," Walsh says. "To be fair to them, at the time they probably didn't have much other information to go on as it was one of the very early long-term contracts ever signed in Australia." Together the interviews helped him come up with a 10-point guide to outsourcing do's and don'ts.
1. The structure of the contract matters. Form a partnership.
When incentives and penalties apply and potential savings are shared, there's more chance the vendor's performance will meet expectations. Flexibility and an open book approach can lead to success with government contracts, where success is defined as meeting pre-determined customer satisfaction metrics within the set price and budget. And you can only build flexibility into contracts, Walsh says, by assuming both business needs and the business environment will change when drawing up the contract and structuring the deal. "A lot of people will spend a lot of time putting together, say, all the financial details of a contract that is meant to last for five or seven years. [But] it is guaranteed that after three years the assumptions will no longer be valid," Walsh says. My advice would be to focus on a formula or formulae instead of the absolute end dollar end result." Make sure the contract is flexible and that it matches business needs, with clear escalation and change management processes built in to allow for negotiations. And if you do enter into an open book partnering arrangement, make sure the culture of both organisations is sympathetic to such deals. The vendor who tendered for the WA government business on the basis of the "way they always did things back in the States" did not get the business.
Recognise the operation of a partnership outsourcing arrangement will require a cultural change for many public sector organisations and put a business strategy in place that focuses on managing rather than doing. And a good working relationship between client and vendor is vital. Build a working relationship that transcends the legal contract and use a partnership agreement describing how to meet contract objectives as the day-to-day working document.
Share both the contract and partnership agreement with the business areas in the client's organisation where the services will be delivered.
2.Don't assume business needs and workplace practices will stay static during the life of the contract. "This point is particularly important in today's environment because work practices are rapidly changing in the public sector," Walsh says. In the case of two of the organisations in question, business needs were determined in 1991/1992; a "coordinating" government department signed an inflexible outsourcing contract in 1993; and the vendor will only finalise implementation of the new systems this year. Not surprisingly, there have been significant changes to business requirements and workplace practices since 1991, none of them covered by the contract. Keep in mind that anywhere between six and 24 months can elapse between deciding to outsource and the commencement of service delivery. In an environment where the agreed contract may become impossible or impractical to implement, an atmosphere of "give and take" is essential on both sides. "The vendor's willingness to be flexible during contract negotiations will be a good indication how it will act once the contract is consummated," Walsh says.
But never forget that flexibility increases uncertainty for the vendor -- and that in turn brings them increased risk, which could lead to price hikes for the customer. "The government department -- and any commercial organisation -- needs to accept that there is a payback for the vendor and they need a period of time to get that payback," Walsh says.
3.Communicate the goals of the outsourcing decision widely or departments will find it hard to react positively.
This is particularly true if service levels are sacrificed for corporate-wide financial gains. Speaking to IT staff and user departments gave Walsh two different perspectives on this issue. In departments where the outsourcing arrangements were deemed to be less successful, Walsh says, the user departments had "zero to limited" knowledge of why the organisation was outsourcing. "Were they outsourcing to save costs? Were they outsourcing for strategic reasons? Were they outsourcing because they didn't have the personnel and skills to run the IT shop? No one in the user department seemed to know." This gap between the strategic management reasons for outsourcing and end-user understanding of the decision led to a lot of frustration and cynicism amongst end users. In organisations where management made clear the reasons for the decision to outsource (government direction, cost savings and improved service levels), the picture was very different. Instead of complaining about suddenly having to pay for services or not having an IT department they could direct to their own purposes, end users were more likely to consider carefully before requesting work.
The lesson? The reasons for the outsourcing decision and the benefits to the organisation should be communicated to all managers and all staff.
4.Have a plan. Accept the process will take more people and more time than anticipated. Be very specific about the services required, terms and conditions.
Assign the right type of people and required amount of resources to run the evaluation and negotiation phases. The outsourcing process is very time consuming, and staff must be dedicated full-time to the project. Expect to have a core team of three to five people working full-time for approximately one year on the project, Walsh says. Apply good risk management procedures and processes to all decision making and evaluation phases. And take the time to producing the service level agreement -- far better to spend a little longer to get it right up front than be continually fixing it afterwards. Any service item not in the original deal will be subject to additional charges from the vendor. "Minor" issues will cause distraction and become "major" issues after the contract is signed. Take the case of the organisation that decided to postpone the question of arrangements for remote locations until after the contract was signed. "When the outsourcing deal was consummated those staff had no idea who they were working for," Walsh says. "The business people had no idea who was actually supplying them with IT services, and these were, from a government services point of view, servicing large regional communities. In the end it actually took as long, in so far as negotiations and management time went, to sort out those issues as it did to negotiate the whole contract. It was a big distraction."Even if it takes longer up front, sort everything out before you sign the contract.
5.A "one-size-fits-all"- type of contract will not work. Encourage agencies to form consortiums and to develop common goals before entering into outsourcing arrangements.
Avoid imposing rigid "whole-of-government" contracts on individual government agencies, because no generic contract covering many agencies (or subsidiaries or strategic business units) can address specific business needs. Instead generic contracts must be adapted to suit the business needs of each individual agency. Take the case of government department "X", which signed a contract with vendor 2 and imposed the terms and conditions on 11 other agencies.
Participants have deemed the contract a failure, resulting in huge levels of dissatisfaction, with 11 agencies attempting to do their own thing and an inevitable blow-out in costs. With the benefit of that unfortunate experience in their armoury, four government departments working in consortium on a later deal sought assistance from the contract and management service group to negotiate the contract and put together the RFP but drove the deal themselves.
That deal seems to be working well, Walsh says. "And the same model was used in a recent contract between nine WA government departments and a major outsourcer," he adds. "Many other lessons learnt were also applied to this new contract, including the idea of forming the consortium beforehand, of having penalties and incentives on both sides, and the idea of sharing risks and rewards."While Walsh concedes it is early days yet, he says by leveraging the lessons learnt from earlier efforts this group of departments has minimised the risks.
To achieve the benefits of "whole-of-government"-type contracts such as economies of scale and common systems you'll need excellent project management and change management at agency, outsourcer and "coordinating" department level. And the vendor must be accountable to each individual government agency for performance-related issues, as well as at a macro level to the "coordinating" government department which signed the contract.
6.Be careful about the size of the contract. Separate regional development issues from IS/IT services.
Can the vendor really handle such a large contract? A large contract that locks in many government agencies may become a monopoly for the vendor. During Walsh's interviews, some participants suggested competition between vendors as the main incentive to provide good quality services and keep costs down. Since outsourcing vendors need a certain critical mass of business in each state to be profitable, and since unprofitable vendors can't provide quality service, be cautious with the divide and conquer approach. "Keep regional development 'carrots' separated from the provision of IS/IT services," Walsh advises. "If the IS/IT service proposal cannot stand on its own two feet then do not enter into the contract. Regional development promises cost the vendor money and distract senior management in both the client and vendor organisations, thus moving the attention away from the provision of IS/IT services."7.Place less weighting on the financial criteria during the evaluation of tender phase. Apply greater financial rigour to the preparation of in-house costs and to the comparison of in-house costs to the outsourcer's costs.
Sticking to a five-year financial plan in a partnership arrangement when business needs and technology are expected to change so much is simply unrealistic. While Walsh says one should not abandon financial comparisons, it is essential to accept that financial projections after year one are unlikely to hold. "All financial projections are based on assumptions. Place a greater emphasis on questioning the assumptions rather than comparing numbers," he says. When the Victorian Auditor General reviewed the state's outsourcing contracts they found that not all the cost of negotiating and managing the contract had been included in the financial models. They also found many IT departments would find it difficult to bring the work back in-house, Walsh says. "Many commercial organisations don't include legal costs, up-front management costs, transition costs including staff costs or the cost of transitioning back," he says.
"IS/IT managers need to take proactive action to reduce the risk of future audit problems. The involvement of appropriate financial personnel throughout all phases of the outsourcing decision will help."8.Have continuity of staff on both sides.
In an ideal world, the person who structures and negotiates the contract on behalf of the vendor should become the account manager, while the person who negotiates the contract on behalf of the client should become the contract administrator. Walsh concedes there's a danger the negotiators may become antagonistic towards each other during negotiation, thus rendering a friendly and professional relationship during contract execution impossible. Ideally he says, the account manager should stay in the role for at least three years.
"The reason for this commitment is to reduce the possibility of the vendor assigning your preferred account manager to the next big contract signed by the vendor," Walsh says.
9.Keep information systems planning, contract and project management skills in-house.Allow vendor involvement in IS/IT planning where appropriate.
Use IS/IT strategic planning to ensure all parties are working towards common business goals. Walsh says that excellent contract management and project management are the keys to getting value from outsourcing arrangements, and points out that IS/IT strategic planning, contract management and project management skills need to be kept in-house. Involve the outsourcing vendor in IS/IT planning and business planning if you want to optimise benefits.
Naturally some situations will arise where it is inappropriate to involve the outsourcing vendor in the planning or decision-making process. Such situations may include the extension or renewal of the outsourcing vendor's contract, the appointment of other outsourcing vendors and the tendering of additional data processing or applications development work10.Have processes and tools to manage technological indivisibility. Make sure there are appropriate channels of communication between the end users and the outsourcing vendor.
In one WA contract involving two organisations the outsourcing vendor manages the mainframe and Unix components, a government coordinating department manages the WAN, and the individual organisations manage PCs and LANs. The result is a situation where accountability is diminished and responsibility dispersed. So difficult has the problem become that the vendor has recently implemented tools to monitor automatically all components, even those outside its contractual responsibility. A well thought out communication process is also essential for day-to-day activities. Walsh says it's important to ensure end users are allowed to speak directly to the outsourcer's staff to resolve routine issues, with all routine requests handled electronically. "Question if the amount of paperwork and approvals required for routine tasks under some outsourcing contracts really add value to your organisation," he says.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.