Aug 2015 | Data Quality | Data quality solutions
By Posted by Jason Goodwin

Being relatively new to Experian I relish every opportunity I have to meet customers and our current roundtable series offers a great opportunity to do that. We kicked off the latest season on 12th August with a discussion based on “data quality maturity” that aimed, amongst other things, to explore what that means, how you define and measure it and, more importantly, what major challenges our customers are experiencing.

We were fortunate to be well represented by customers across a variety of sectors from Financial Services and Local Government to Retail and Charities, from whom I wanted to share the insights I found most fascinating.

1. Data quality definitions are not all the same

We talk widely about data quality but in reality it means something different to every individual and every organisation. It’s important for organisations to define what it means for them, whether that’s timeliness, completeness, accuracy, latency…or indeed a combination of them all. Only once that is defined can you can set the base-line for what is expected in order to try to achieve alignment across the organisation.

2. Data quality maturity is rarely consistent across organisations

Through our discussions and a look at the results of our Data Quality Improvement Assessment it was clear that data quality excellence exists in many of the organisations that attended, be that at department, team or individual level. However, it is rarely evenly applied across the organisation due to a variety of factors that can include culture, measures, technology, and “data quality isn’t my job” syndrome! This variance makes measuring data quality maturity across a whole organisation difficult as it is often in the eye of the beholder.

3. Pareto strikes again: The 80/20 rule

A clear theme was how focusing on the quality of just 20% of your data attributes can actually drive more value than attempting to fully maintain all of your data –  it’s working out which 20% that’s the issue. To help identify your most valuable data, a workshop approach can prove useful, where attributes are weighted with a justification as to that weighting. “Value” has more than one dimension however. It may be that something is valuable as it costs or wastes more resources if it’s wrong or it may be the number of instances where an attribution is incorrect or unusable that increases its impact.

4. Driving accountability proves challenging

Ensuring that there is someone, or indeed a group of data stewards in your organisation who own the data is a consistent challenge. The group’s experience in managing this was varied and included a mix of “carrot” (e.g. share of savings to re-invest) and “stick” (e.g. name and shame the worst offenders) approach. The group agreed that accountability for quality data needs to be part of an organisation’s DNA but that takes time. Technology to make changes easier and issues more visible, along with top-down measures and messages are a couple of ways in which attendees have seen accountability evolve more rapidly.

5. Proactive data quality starts at the point of capture

There is a general perception in the non-data world that “data quality” is about fixing issues. But the reality is however, garbage in and garbage out, which means data quality must start at the point of capture.  A “get it right before it goes wrong” ethos will help ensure progress maturity from a reactive to proactive data management strategy.

In my role as Data Quality General Manager, these sorts of insights are invaluable in ensuring that our technology is meeting the needs of users and the evolving data quality market.  What’s encouraging is for many of the challenges we explored, Experian can help through its combination of visionary data management software, Experian Pandora, and its unique reference and enrichment data. Just one example is how it can support in identifying data quality priorities by calculating the cost of data inaccuracy for the business’. which can then be tracked on data quality dashboards.

As a data management platform that combines rapid time to value with incredible ease of use, what makes Experian Pandora particularly relevant is its suitability to business users at any stage in their data quality journey – from understanding your data issues through to supporting a range of data-driven initiatives such as enabling data governance, building a single customer view, supporting adherence to compliance and regulation or ensuring a successful data migration.

If you’re looking to scale up your data quality I’d recommend completing our Data Quality Improvement Assessment which can provide a view on your current level of maturity and give useful pointers on how to progress.  Alternatively Experian can help by providing a Data Quality Health Check, where the combination of software and our expertise can rapidly deployed in a 10 day exercise to provide insight and recommendations on how to make improvements.