Originally Posted by
Bigbux
Hi PDR
I'd love to read more, I'm guessing this was for a technical degree rather than a legal or commercial subject?
Well there's 54,000 words of it (226 pages with 14 jokes to keep the reader sane). It's an MSc dissertation, but its topic focuses around the interaction between the commercial and engineering processes in the UK defence sector. In fact at one point it was subtitled "Looking for a way to undertake a sound engineering process in a commercial environment that is explicitly designed to prohibit it" but my supervisor made me delete it. Some parts of it are confidential, but I could pass you a redacted copy of you're having trouble sleeping.
One of the interesting and controversial findings was that competition doesn't work. I don't know why this was so controversial, because the Audit Office had already found just this when looking at defence procurement - the programmes that were on time, on budget and delivering to expectation were generally single-sourced partnered procurements, whereas those which were late, over budget or failing to deliver were generally competed procurements (neither is 100% true, but the strong trend is clear in the NAO reports). In fact most of the problems on those programmes have roots that can be traced to a perceived commercial need to maintain "competitive tension" when it had no contribution to the Mission; that conflict just added complexity for its own sake and measurement of parameters no one actually cared about.
If I were designing a contract for development and I read the above in the brief my initial thoughts would be:
A penalty clause is not designed to guarantee availability/performance. It is an incentive to the supplier to show some interest, and part fund the fix required by the user. Effective penalty clauses are usually combined with a number of other risk measuring/reducing mechanisms in a contract - such as KPIs/escalation/geared incentives etc.
The point of that particular section is to illustrate the difference between military and commercial contracting. It is extremely rare in commercial contracting to have a non-performance which cannot be both measured and resolved in purely financial terms. The military mission is different, so a purely commercial model won't necessarily be appropriate.
Any responsible project owner would be expected to conduct rigorous due diligence into all of the major risk areas of their project. In the commercial world, evidence that dd has been carried out is usually a pre-requisite to secure funding.
In the UK defence sector it's not always possible. In some cases due to the technical risk or mission uncertainity and in others simply due to time and cost. For example a few years ago the MoD ran a competition (using the "competitive dialogue" process) to contract out its Air Traffic Management systems in a project called "Marshall". There were three downselected tenderers looking to take on, manage, update and/or replace a vast range of equipment on 106 MoD sites. Huge volumes of site survey data (drawings, specs, equipment lists, maintenance records, photos etc) were provided to the tenderers, but most of it was many years old and it was "unwarranted". Each tenderer was permitted a single "due diligence visit" comprising no more than 3 people per company for one or two days to each site. These visits were made together - tenderers had to seek information and ask questions in the presence of their competitors "to ensure a level playing field", so of course many questions went unasked. No further visits or questions were permitted, and on the basis of that the bidder would be expected to take on a £1.7bn risk with a 22 year maturity date. That's why one of the bidders pulled out before the second dialogue phase and the one that ultimately won the contract is now haemorrhaging money and looking to re-negotiate the deal.
If you are competing a requirement (say, for a new engine/aircraft/armoured vehicle) then VfM would be difficult to achieve if you simply paid each supplier a large sum to collate reliability data. You would probably want to make confidence levels in reliability a significant element in your evaluation criteria.
Good plan. How? Reliability predictions are meaningless and reliability measurement is extremely difficulty and expensive to measure for anything but very large fleets.
If the project has to award to a single chosen contractor without competition - then funding reliability could reduce risk - but you have to make sure you ask for relevant data
You're assuming the data exist. Reliability data only exist for off the shelf items which are currently fielded in large quantities for long periods in applications and environments which are identical to the one you're interested in. If any of these aspects are different the reliability data are meaningless. And in any event no mission owner is actually INTERESTED in reliability - they are interested in mission capability, mission success and mission affordability. In some cases this will have a dependency on Mission Reliability (the probability that the system will complete the required mission) whilst in others it will have a dependency on Logistic Reliability (the amount of stuff that will break which I need to expend resources to fix). Which of these dominates depends on the nature of the mission, and is unique to each instance. There are also cases where low reliability is unimportant provided the system has excellent
maintainability - it can happily fail every hour if making it serviceable take 5 seconds with no skills, tools or spares.
Finally, "mutual trust" and "partnering" are loose, woolly phrases bandied around in the forlorn hope that an inadequate contract will protect the MoD from poor performance.
Again, experience says otherwise. There are several MoD programmes which succeed precisely because integrated industry/military teams operate with common objectives towards a common goal, with a simple KPI structure that is directly related to the mission rather than commercial dogma. And these are the ones which the audit office says deliverd on time, on budget with clear VfM benefits.
There are many non-commercial specialists involved in MoD procurement, most are intelligent and capable within their specialist areas, but a little bit of legal and commercial training followed up with some real experience would serve them and the Country well. Sadly, it does not always happen: hence a contract is viewed as adversarial; development contract methodologies are largely unknown, and industry is still occasionally allowed to run rings around bright, professional people operating well outside of their competences.
Well that's one view. my own experience is that the "commercial specialists" usually cause the problems - insisting on KPIs being allocated to every line item in a 500 page contract with each being measured and sentenced every month, requiring large staffs of people solely to collect data for sentencing when it serves no other purpose. I've seen these people insist that it's "mandatory" to spend £500k/yr for five years measuring and reporting performance stats for an item whose worst-case delivery failure would cost the front-line command an extra 4 maintenance man-hours per squadron per year. I think such commercial expertise is something we can all do without, actually!
PDR