I assumed this had been reported widely outside of Massachusetts, but a web search suggests otherwise.
The story is this: Cognos sold 20,000 seats of software to Massachusetts for $13 million. There were technical violations of purchase procedures, and other aspects of the deal that didn’t pass the smell test. After IBM bought Cognos, the deal was rescinded, and is being rebid.
As per the Boston Globe, highlights of the mess include:
- The Cognos salesman on the deal was well connected politically, specifically with the House Speaker, who seems to have been a prime mover behind there being a BI purchase.
- A much higher fraction of the workforce got seats than is generally regarded as necessary in state governments (25% overall). Ohio, by way of comparison, bought BI seats for 10% of the workforce.Ohio, by way of comparison, bought BI seats for 10% of the workforce.
- The overall price of the deal is much higher than the two alternative bids (SAS and Oracle — no Business Objects or Information Builders, for some reason).
- Prior studies as to what was or wasn’t needed in the BI purchase were ignored. For example, they recommended seats only for 1-10% of its workforce.
IDG has some good coverage too, the highlight being:
For one, the state’s Information Technology Division did not widely advertise the fact it was looking for performance management software, according to the report.
Instead, “a staff member at ITD simply consulted a chart of leaders in performance management developed by the analytical firm Gartner Group and e-mailed the Request for Quotes to four companies identified as ‘leaders.’ “
Three vendors — Cognos, Oracle and SAS — responded to the e-mail, according to the report. ITD staffers developed a scoring sheet containing 104 criteria. The ITD team in charge never finished evaluating the vendors with the sheet, but at the time they stopped Cognos had the high score, with 69.39 points, followed by SAS with 57.38 and Oracle with 27.49, the report states.
The IG’s investigation found that due to a typographical error in the spreadsheet’s formula, the scores for all three vendors were flawed, with many points going uncounted.
Please supply your own punchline.
What has not been reported, insofar as I recall or can find via quick searches on Google or Boston.com, is whether the Cognos software included a lot more capabilities than the other offerings. The obvious candidate would be budgeting/planning. That’s something that governments like and Cognos tries to bundle into large enterprises deals, but which is not a strength of (for example) rival bidder SAS.