Data quality vendor map

Written By:
Content Copyright © 2009 Bloor. All Rights Reserved.
Also posted on: Accessibility

Like most companies we are always trying new methods of visualisation to better understand and communicate our message. One technique that we have recently introduced is what we call a market map, a variant of which is a vendor map. I thought it would be useful to introduce the use of this in the context of the recent reports on data quality that we have published. So, the following diagram shows the vendor map for general-purpose data quality platforms, based on our recent research.

Data quality vendor map diagram

There are a number of points to be made about this. The first is that it includes updated information since our data quality reports were published. For example, IBM has acquired Exeros, so the latter’s capability has been folded in to those that IBM had already and, although integration will barely have been started, this explains why IBM has now the highest ranking for data profiling.

Also on the product side, we have been able to include an assessment for Pervasive’s Data Matcher and for Datanomic’s profiling capabilities and an estimate for Microsoft’s data cleansing (this is still pending Zoomix integration). Oracle remains a notable omission but its resale of Trillium, Silver Creek and Identity Systems (Informatica) means that its position can easily be assessed from this diagram, although these are not all resold in or with the same product.

In terms of the actual visualisation the x and y axes should be self-explanatory. However, the sizes of the dots representing the different vendors may need some explanation. Briefly, this combines numbers of deployments, geographical coverage, company stability, and so on as well as including suitability for enterprise deployments. This last point is important. Global IDs, for example, is neither very large as a company nor does it have a huge number of customers. However, its software is specifically targeted at (and suitable for) very complex data quality environments in what the company describes as “organisations with global reach”. As a result, Global IDs has a larger dot than might otherwise be the case. Similar considerations may also apply in reverse.

One of the nice things about vendor maps is, of course, that you can actually choose a variety of different measures to graph against. For example, instead of plotting profiling against cleansing we might have combined this along a single axis and compared it with support for data governance, for example. Or I could have stuck to the same axes but had the size of the dots represent support for governance. At some time we would like to have an interactive capability on our web site that allowed users to choose their own attributes to graph by—but that’s probably dependent on whether we can get SAS, Tableau, Advizor, Spotfire or one of those guys to sponsor the capability—and in any case, that’s for the future.

In the short term I would be interested in any feedback readers might have as to whether you like this approach, whether it makes it easier to understand the market, any improvements you might suggest, and so on.