Benchmarking Data Governance

In talking recently to several companies that have data governance initiatives in place, it is clear that such initiatives costs a lot of money (an average of eleven full time people according to or latest survey) and that there is pressure on to identify whether such initiatives are delivering value. One difficulty here is in comparing your data governance to other – after all there is no “benchmark” available. We aim to change that situation, and have today launched a data governance benchmarking initiative, in conjunction with the Data Governance Institute.

If you have a data governance program, up and running, and want to see what you are doing compared to others, then please join this:

In return for a completed survey (it is lengthier than our usual surveys, but you “get what you pay for” with surveys i.e. the more information you provide, the greater range of analysis we can do) you will receive a free summary report of the findings. There will also be an opportunity to purchase an in-depth, customised study of your organization’s performance relative to others.

The Wisdom of Oracles

I came across an intriguing poll today. A small company called Software Advice , who provide advice on evaluation of ERP and MRP systems, has been running an on-line poll about which software company Oracle will acquire next. The results of their poll so far (which is still open) is here.

What interested me most was not the actual poll results as such, but the way in which this poll has managed to attract a very high response rate (well over 1,000 responses). It could be viewed as an example of crowdsourcing, which is an idea explored by James Surowiecki in his book “The Wisdom of Crowds”. This is a fascinating idea, reasoning that in some cases a large sample of answers will actually generate a more accurate response than a supposed expert. Of course this method does not always work (Gary Kasparov beat thousands of players voting in a poll in a chess match against “the world” in 1999) but it seems like an idea that is worth further testing.

This is the first time I have seen this notion being applied to the world of enterprise software. There are plenty of other questions that could be tackled by such an approach, and I hope that this company or others extend this poll to other areas.