'Independent' report used MS-sourced data to trash OSS
'Microsoft gave us this stuff, but we haven't checked it'
When Microsoft announced victory  in its battle with open source for the hearts, minds and wallets of Newham council last month, the press naturally asked to see the report Newham's decision was allegedly based on. Er, yes, the report would be forthcoming, we were told. So we gave it a few days, and asked again. And again. And again. But mysteriously, although Microsoft executives are prepared to shout from the rafters about what's in the report, the report itself never seems to arrive.
Fortunately, there are ways to get hold of documents that have been presented, as the report was, to public bodies, so if Microsoft is still unable to find it, then The Register is now in a position to help them out. And an enlightening read it is, too.
But we'll parse the spun versions first, and remind you of The Facts as we go along. According to Microsoft's press release of last month, "Newham's decision to partner with Microsoft follows an extensive platform evaluation at the Borough, in which the merits of, and overall value offered by Microsoft's software were rigorously assessed by Capgemini. Capgemini were selected to conduct the evaluation by Newham, and ran the evaluations as an independent third party."
Fact One: None of this is untrue, as such, but a study by Capgemini, funded by Microsoft, was Microsoft's offer during pricing negotiations with Newham last autumn. Note that Cap Gemini's brief was to assess value offered by Microsoft software, and that it ran its evaluation as an independent third party. The Beast chooses its words carefully here.
It continues: "The Capgemini report predicts the solution could allow the council's ICT team to maintain its current level of service while reducing support costs by 13.5 per cent, a saving of £3.2m over five years."
Fact Two: Note that we're still on Microsoft software here, so the claimed £3.2m saving is derived from a migration from the current Windows infrastructure to Windows XP, Office XP and Windows Server 2003. The numbers themselves, the Cap Gemini report reveals, are produced by running Newham's stats through a "cost model based on the Gartner Total Cost of Ownership (TCO) Analyst 5.0 tool and the Gartner TCO Chart of Accounts." One wonders why they didn't just hire Gartner.
But next, we get to the tricky bit: "In addition, the Capgemini study found that the Microsoft solution could double the cost savings associated with improvements in productivity compared to an Open Source solution. It offers lower administration costs, improved security and increased productivity for users. What's more, the commercial solution will also provide a predictable, sustainable and long term IT funding model for transformation, which will protect the Borough's investment."
This one's intriguing, because it appears to contradict the known facts about what was being studied, by whom, at Newham last autumn. Cap Gemini was doing a study of Newham's current Windows setup, while consultants netproject were engaged in a more long term exercise concerning the borough's future IT strategy. The open source one was netproject (which incidentally did not recommend an immediate rip and replace in favour of open source), and the Windows one was the Cap Gemini one. There was no study involving a direct face-off of Windows and open source, far less a level playing field one. So where the blazes did the savings over an open source system come from?
Fact Three: Er, Microsoft. According to the Cap Gemini report Microsoft seems unable to find, "Microsoft have also offered figures for a (current configuration) to Linux migration that they have generated from the same Gartner model using input data validated with the ICT group. We have not independently validated these figures. However, we feel it is right to include them in this report so that Newham is aware of them and can choose whether or not o [sic] consider them in any judgement they may make about their future strategy.
"By using Gartner's methodology in this way, Microsoft believe that a desktop migration to an open source solution based on a Linux Platform could potentially generate lower operating costs for Newham in the region of £350,000 per annum. This shows potential cost savings to the council over five years of £1.6m, or roughly half that available from the deployment of a Microsoft based solution." (our italics throughout)
Here as you can see it's Cap Gemini's turn to be careful about the wording; it has its good name to consider after all, and can't be seen to just shove in a bunch of data handed to it by Microsoft without pointing out that it's nothing to do with Cap Gemini. But the bottom line is that the headline 'finding' of the study is based on data submitted by Microsoft but not verified by Cap Gemini and is specifically qualified by Cap Gemini as Microsoft's 'belief.' Cap Gemini itself simply concluded that Newham could save a heap of money by switching from Microsoft to Microsoft, and said it was up to the council whether it gave any credence to Microsoft's figures or not.
The Cap Gemini report is also helpful on the subject of how the Microsoft vs Microsoft cost savings might be derived, and this is important, because the amount paid to Microsoft was of prime importance to Newham during the negotiations. As you'll recall, Newham wanted price cuts, Microsoft initially held out and offered a free study instead, and in November 2003 Newham looked at the netproject recommendations and the Cap Gemini report and provisionally decided in favour of open source, subject to further price negotiations. Then mysteriously, Microsoft and Newham loved one another again.
This becomes less mysterious with the help of the Cap Gemini report: "Looking at the long-term transformation exercise for Newham, we have consulted with Microsoft to explore how a future partnership with Newham might secure predictable, sustainable and long term funding model." [sic] What on earth could that mean, one wonders? It sounds remarkably like some kind of coded wording for 'price cuts' or 'special deals.' Which indeed it appears to be: "This goal can be supported by: A Local Government Partnership Programme that provides very close working with a select number of Local Authorities and influencers that provide detailed knowledge; Use of Microsoft's Local Government User Forum to allow a broader market contact and feedback; and Creating a solution reference programme to promote uptake of Microsoft."
Well... The "select number of Local Authorities" provides a mechanism for Microsoft UK to do special deals with organisations it wants to keep within the tent, without the price cut rot leaching out to the whole of government. We have no idea about the Local Government User Forum, but "a solution reference programme to promote uptake of Microsoft" can surely only mean that Newham can get special prices and support and installation subsidy by becoming a showcase that is pumped in case studies as a Microsoft success. We're pretty sure that talking to the IT supplier in order to move the pricing goalposts isn't part of Gartner's TCO methodology, and we have severe doubts about how this sort of approach squares with acting "as an independent third party."
At least one case study, incidentally, exists already, but we've no idea how many Exchange seats it counts as. You can get it here,  and if you read through it you'll note how The Facts morph as the spin doctors get hold of them. "A multi-disciplinary team under the lead of Capgemini was brought in to evaluate the best possible solution for the borough, comparing the costs, business benefits, and risk profiles of a Microsoft solution (incorporating Microsoft Windows XP and the Microsoft Office System on the desktop, with Microsoft Windows Server 2003 on the server) with equivalent Open Source solutions... Specific benefits of an all-Microsoft solution included: A potential 13.5 per cent reduction in ICT support and operation costs; Potentially double the productivity benefits of an Open Source solution."
Yes, that right, the heavy disclaimers about the open source exercise fell off. And off here:
"Based upon the Gartner TCO tools, the study also indicated that an Open Source solution would provide approximately half of the cost savings of a comparable Microsoft solution, but would cost three times more to implement due to significantly increased migration costs.
"The estimated cost savings for the Open Source solution, based upon Linux desktop and server operating systems and either StarOffice or OpenOffice as the productivity application, were in the region of seven per cent, per annum, as compared to the 13.5% saving per annum on ICT support costs with the Microsoft solution. The Microsoft Solution had potentially double the productivity associated cost savings of an Open Source solution."
We've already seen how the study came to indicate these things. It's also worth noting at this juncture that the Microsoft suggested open source deployment is imaginary. The netproject report recommends in the short term a server infrastructure move to OSS "where this would not interfere with the running of the current environment", the conversion of applications to run under Terminal Server or to be delivered via browser, an upgrade to Exchange 2003, then a pilot deployment of OpenOffice and Mozilla, and finally the rollout of OSS desktops "once the need to have a specific desktop operating system has been removed."
netproject envisaged a gradual evolution towards server-based applications, and suggested that in many cases OSS desktops could be installed on existing hardware, whereas a switch from Win98 to XP clients would most certainly require new hardware. Its conclusion is: "Once all the differences are taken into account we believe that the high cost of the licence fees for the new Microsoft Products more than outweighs the additional costs involved in the migration to OSS. The running costs of both environments are comparable and so since the licence costs will be repeated sometime in the future whereas the OSS costs are a one time initial cost, OSS actually provides the most cost effective route."
The Microsoft case study, on the other hand, claims a 68 per cent migration cost advantage over open source, and finishes off with a presumably unintentional contradiction from Newham ICT head Richard Steel: "While using Open Source might provide some short-term cost benefits, the longer term benefits related to technology and thought leadership of standardizing on a Microsoft platform are significant. While not easily quantifiable, these benefits are nonetheless a vital component of our decision-making process. We are confident that a standard Microsoft desktop and server infrastructure will reap great dividends for the council and its constituency."
We don't know about you lot, but if our head of IT started going on about a benficial "thought leadership" that was "not easily quantified" we'd be deeply worried about his stability and our IT budget. But on the subject of pricing, the reference to "a solution reference programme" is actually repeated in the Microsoft case study. Given that the case study is intended to be part of the Microsoft Get the Facts effort to sell Windows and discourage open source, and will therefore be a darn sight easier to get hold of than the raw data, it may well be that Microsoft intends to be upfront about the use of soft subsidy of this kind.
We've already got various Windows logo programmes and PC companies getting marketing payments for recommending Windows, so it's not a huge leap to a similarly formalised subsidy system for selected customers. Kind of screws up The Facts though. Microsoft pays you money to say it's great, which improves your bottom line, so when you do the numbers you reckon it now looks pretty good value, so maybe you think it is great. Anyway, isn't it great to be paid for saying it's great? Or something.
But here's a more positive spin on this from the case study: "The appropriate licensing mechanisms, coupled with an active role for Newham in these mechanisms, demonstrated that Microsoft was willing to step up to the challenge of assisting Newham in reaching their transformation goals, by working with them as a partner, rather than purely a vendor... Other local authorities in the UK look to Newham for technology leadership. The council’s adoption of a current and more sophisticated desktop and server infrastructure will no doubt help to influence the technology decisions of its counterpart authorities."
There you go, all perfectly legit and not dodgy at all. Don't say price cuts, say partnership. Don't say backhanders from the marketing budget, say technology leader, say key influencer.
Microsoft, incidentally, has a long and inglorious history of reverse-process studies (i.e. starting with the conclusions then sourcing the data appropriate for arrival at it). There are many, many examples but we feel this blast from the past  is particularly appropriate to current circumstances. Regrettably, the link to Linux Myths no longer works, but it was sort of release 1.0 of Get the Facts.
We'd like to thank Microsoft for its help in the research for this article, but try as we may we can think of no earthly reason to do so. We will thank Eddie Bleasdale of netproject, who kindly secured Richard Steel clearance for us to see the netproject study. Apparently Richard said we might as well have it "seeing Silicon have got it already." Well, cheers to you too, Richard... ®
Newham and Microsoft sign 10-yr deal 
MS, open source, The Facts and the fit-ups 
Microsoft, Sun, IBM and the war for government desktops 
Gates to meet Brown, OGC and NHS chiefs Sun, OSS in crosshairs? 
Savage discounts from MS flush OSS desktop from London council 
How to help MS, harm OSS by not buying Microsoft