I recently sat on the Master Data Management panel for the DAMA South Africa Vendor Day.

I was surprised that, of the +- 100 people in the audience, only a handful responded that they were busy with a MDM implementation.

Data Governance and Data Quality

The discussion began to take shape as the audience, for the most part, agreed that Data Governance and Data Quality are significant contributors to successful MDM. It is interesting how this reflects the comments from the recent international TDWI MDM conference – as discussed by my Trillium DQ colleague, Jim Orr.

Duplicate data

Another thread that developed was the concern raised by an audience member, who has implemented a stack solution for MDM, that they were left with significant levels of duplicate records once they had implemented – although they had understood that the MDM stack would address this.

We know that best-of-breed data quality tools are typically far better suited to the complex problem of standardising, deduplicating and merging (or commonising) records than most MDM stacks base functionality – precisely why most MDM tools will integrate with these tools.

Blind allegiance is a risk

The message, summarised very nicely by Evan Levy, is that blind allegiance to a vendor (we will buy XXX MDM because we are an XXX shop) is a leading cause of project failure. Customers should rather make decisions based on Needs, Priorities and Cost Metrics. As differing solutions have differing strengths, a proper analysis may indicate that your preferred vendor’s product simply does not match your needs – whether for data quality and automated matching and commonising, or in a range of other areas.

To quote Evan, “Working with a reputable vendor is smart. Gathering requirements, reviewing product features, and determining the best match creates the opportunity for developing a client/vendor partnership. So why would anyone throw all of that out and just decide to pick a vendor? “

If Data Quality is a significant contributor to MDM success, surely it makes sense to evaluate the Data Quality solution provided with your stack against best-of-breed alternatives (all of which will integrate), and make sure which will best suit your needs?