Current procurement for public sector IT needs to be better and more cost-effective. Massive monolithic procurements that only a few can afford to bid for have been repeatedly discredited. The intention of engaging small and medium-sized enterprises, encouraging innovation and splitting up procurement into bite-sized chunks have not yet matured in practice.

OGC Gateway Reviews and the identification of a Senior Responsible Owner (SRO) have been useful initial measures – and in some cases have offered improvements. But proper scrutiny and early decisions to cancel/modify a project can all too easily be circumvented. Also, procurement does not always interact well with the political process: when legislation over-specifies a solution or makes last-minute changes, it can have significant cost and delviery implications on a project.

There needs to be a better categoration of procurement and different strategies for handling them. Commodities (such as operating systems, email, office application software or some network hardware) can be procured in a very different way to heavily bespoked requirements relating to one-off legislation. And common infrastructure (such as secure networking) may benefit from national-scale procurement activity, whilst local government web sites may not.

Pricing may be the critical factor in an individual procurement, or it may be risk, innovation or capability. Procurement rules (such as time trading, turnover levels) and other barriers to entry may be appropriate for some largescale, bespoke and high risk projects, but not for others. There are many tools and strategies that can be used, but they must be applied intelligently based on the characteristics of what is needed.

Government should also make use of what is available commercially in the market where there is a sufficient match. The rapid evolution of generic software is such that bespoke government procurements can easily be outpaced by the market. Where partial solutions exist in the form of open source products – or where there may be an unmet market demand for software – government can consider funding open source development and other types of community contribution to meet its needs. But government needs to recognise its own significant distorting impact on the wider IT market, and use its own buying power to help invigorate local markets rather than stifle them.

Government should use, participate in and adopt open standards. This will break up systems into components and interfaces, and systems can be evolved/upgraded rather than made obsolete. Well-defined interfaces can enable bite-sized procurement, competition and engagement from SMEs. Open standards can also be developed outside the procurement process and contractual situations, leading to well-specified and lower risk procurement of components.

Key Principles

1. Effective competition is needed in the public sector IT marketplace
2. Complex and commodity procurements should be handled separately, with appropriate procurement approaches
3. Commission bespoke work only when necessary, with a presumption of adopting and driving market solutions over proprietary and custom
4. Break procurements into components that solve specific problems

Other points/clarifications

Government should look to the principles of open source as a way of achieving value for money – and as a participant both giving and benefiting from community projects. But there may be areas where the economics or simple availability suggests proprietary solutions. However, government needs to be fully aware of the potential impacts of any market interventions it makes. Where there are thriving businesses developing products and services, to fund an open source alternative or release a government-owned competitor can be counterproductive. Government should undertake market impact assessments before choosing to fund projects for general release (closed or open source), taking responsibility for its own buying/funding decisions.

Specific recommendations

Things to do:

1. Develop and use a categorisation of IT procurement types, with approaches for each. Break commodity, low risk purchases away from one-off, high risk procurements: do not continue to bundle them together or place with a single supplier just “because it is easier” (use large, high reliability suppliers where appropriate; use agile, specialist providers for suitable tasks. Stop bundling the two together and manage them separately)
2. Ensure the strict separation of procurement, review and audit, providing as much publicly-available data and scrutiny as possible.
3. Ensure appropriately skilled technical staff are involved in procurement processes and provide independent technical assessments
4. For existing contracts:
   1. Set up a transparency portal with details of all existing contracts and all spend over £25K
   2. Identify % market share of all major IT suppliers to the public sector
   3. Identify risk/exposure of “too big to fail” IT suppliers in the public sector
   4. Identify whether companies pay corporate tax in the UK – and, where they do not, publish the amount lost annually to the UK taxpayer/exchequer
   5. Request existing suppliers to agree to the open publication of contractual information on the transparency portal and, where they do not agree to do so, make clear on the portal (a) they have declined (b) total value of taxpayers’ business with that organisation
   6. Review existing contracts against actual requirements, costs, benefits, get-out-clauses
   7. Work with existing suppliers to drive a least 20% cost out of existing services within 24 months
   8. Make Permanent Secretaries, CIOs and the IT department account properly for all expenditure rather than, as at present, not knowing what is being spent. This should be done within 1 year, the current lack of proper accounting systems is unacceptable
   9. Assess the extent to which TUPE is, as some suppliers claim, preventing them delivering innovation, better services at lower cost
   10. Request feedback on which products and technologies currently in place under existing contracts are proprietary and/or otherwise not-interoperable
5. For new contracts:
   1. freeze all new procurement while other aspects of eg goverance, information architecture and technical architecture are fixed, along wth clear stategies around eg identity, privacy and security. Exceptions may be made, but must be thoroughly reviewed.
   2. assess the extent to which agile procurement methods are permitted under EU legislation and promote their rapid adoption in the UK in place of the current model
   3. move to small procurement chunks rather than multi-year, complex requirements
   4. drop OCG requirement for at least 3 years trading figures on smaller procurements and bring in a more realistic risk-assessment model (rather than current risk avoidance)and “keep ajar” framework procurements for new entrants
   5. stage contracts, with competitive tendering at each stage
   6. ensure all projects have a senior, accountable business sponsor not just a technical owner
   7. establish clear acountability for failed procurements/projects/programmes (make it possible to remove staff – and suppliers – who have not performed well as easily as can be done in the private sector)

Stop, challenge or review:

1. Stop bespoke bundling of different requirements into a single procurement as the status quo, and aggressively explore market alternatives
2. Stop pointless red tape that impedes rather than encourages effective competition and an efficient, open marketplace