In a recent article in online technology ezine Delimiter – “Qld Health Preps Huge IT Outsourcing Deals” – Renai LeMay points out that the Qld government will need to spend an unadjusted $7.4 billion over the next 5 years in order to replace and upgrade 90% of its outdated ICT portfolio.
The question is whether it has learnt from the Qld Health outsourcing debacle with IBM and will it move forward with its best foot? It is unlikely, given that HSIA seems thrust into the contracting process too early with too little. With the pace seemed to be set by the Costello audit, the HSIA is now engaging in early vendor ‘discussions’ (IBM notably excluded) without even a detailed set of business requirements, system parameters and financial boundaries.
In letting vendors shape the early development of these outsourcing contracts the government is on track to lose billions. This will happen for 3 primary reasons:
- Service Management will be overlooked – The failure of virtually all contracts is not in the structure but the management. Although most experienced litigators have gasped at contracts (usually government) which seemed to have been designed to fail (largely due to something bordering on corruption or undue influence), most are not so designed. Outsourcing contracts, more than any other, require detailed attention to the management of the service. Ultimately, the difficulty lies in the paradigm shift from network support to service management. In other words, the personnel in charge of servicing are now in charge of service management. In most cases, these jobs are made redundant as the vendor takes them over (unlike TUPE laws in the UK). Consequently, the knowledge is lost. Purchasors can protect themselves against service decline by making the vendor buy key personnel. Better yet, Purchasors should transition these key roles from performing the work to managing the service/contract.
- Deals will be too long – Vendors will push longer outsourcing contract lifecycles. Although, prima facie, there is nothing long with a long contract it is imperative that such contracts are designed to be managed, i.e. the focus is on the delivery and management Schedules and not the boilerplate of the Operational clauses.
- Tech bundles will lack modularity – Qld government loves to re-organise. After each government their is always a paradigm shift in the Machinery of Government (MOG). Departments shift and with them so do budgets and internal processes. These are generally managed within departmental parameters with necessary roles and functions often going unfunded and unfilled for entire cycles of government. To reinforce Renai LeMay’s point, it goes without saying, therefore, that the government needs to develop a modular, multi-bundle architecture wherein departments can buy and sell service credits between themselves without limiting the overall strategic cost savings.
In summ, there are numerous ways in which Qld government can guard against the inevitability of cost overruns and poor, overly simplistic outsourcing contracts. On the other hand, the vendor which offers these first will have a significant advantage. in many cases, the difficulty will be in convincing QGCPO that novel and innovative contracting vehicles are there for the benefit of the Purchasor and not just vendor voodoo.
In a recent article in online magazine IT News, government CIOs are actively warned against signing long term service deals. Mike Lafford, from Gartner, advises government CIOs against long term contracts.
“Please, please, please, don’t sign ten year deals.“
– Mike Lafford, Gartner
Logic favours a longer contract in order to squeeze more value for money from a vendor by allowing them greater economies of scale (i.e. more guaranteed revenue = more borrowing power = bigger, better service infrastructure). However, Mr Lafford notes that the contracting process is so long and tedious that vendors artificially force up their margins to cover the enormous costs of business development and tendering. Government departments, therefore in particular, do not see cost benefits.
– Current outsourcing contracts show a level of sharp practice and degree of innovation usually reserved for used car sales –
Lafford, however, does concede that there are 3 artificial factors which increase both the risk and costs for vendors, namely:
- Transition Costs are High – Departments are told they place an unusually high management burden on vendors by ‘creeping’ into contracts. Long evaluation periods and bespoke management structures all serve to drive up vendor costs. Purchasors should align their management structures with the, relatively, inflexible service delivery of the vendors. In addition they can give the vendor a 3-6 month Service Credit holiday with a Buy-Back/Call Option at the en. In this situation all current infrastructure remains in place and the Buy-Back option (purchase for the base cost of transition) is contingent on a stipulated average service performance level well below market benchmarks.
- Vendor Risk in Technology/Business Change – Primary risk is derived from reduced volume/numbers, e.g. in desktop support. There are 2 possible ways to deal with this: (i) sign a deal involving multiple departments. Hedge overflow costs buy buying and selling capacity amongst departments/teams. Secondly, (ii) pay for the minimum not the maximum and then buy overflow capacity. Departments and teams are always looking at headcount reduction so don’t buy at your maximum capacity but do expect/ensure good discounts (on an increasing scale) for more capacity.
- Inflexibility in Contracts – The current thesis is that long contracts are inflexible contracts. This is nonsense, although technology contracts can be far more complex than PPP contracts, for instance, for the reasons stated above, i.e. lack of clarity in the future requirements. The focus of most contracts is on the ‘operational clauses’, i.e. boilerplate governing standard business relationships. Failure of outsourcing contracts, however, is almost always with the failure of contract management (by which we usually mean “service management”). In order to develop a contract which changes over time, the focus of the contract must be in the Service Management Schedules.
In summ, governments and businesses alike have to realise that someone pays for risk. If they place risk with the vendor then the business/department ultimately buys it back. To that end, I would usually recommend a 5+5 contract. The trick, however, is not with the development of the KPIs but rather with the management of the service. To that end, Purchasors should not underestimate the detail they need to go in to when developing their service management structure.