Thursday, March 26, 2009

Put Data Quality in Those Requirements, Already!

Business Analysts look at all kinds of requirements.

1. Networking interface;

2. System design;

3. Database design;

4. Business needs;

5. and, much more...

How often will a business analyst look at data quality requirements?

It is a missing. Period!

Rarely will you see a heading in your Business and System Requirements documentation for Data Quality. The business has such a hard time understanding what Data Quality actually means that they ignore the subject. This is not the fault of the business analyst, but often an oversight from the beginning of project conception. Quite often the reason for the oversight is the business' inability to quantify the benefits of data quality and the actions needed to prevent bad data.

I don't want to harp on the business, but lets face it, it is the rare mandate in any IT project to ensure data quality, or even to provide data quality action items. It is often left to the support analysts at the end of the day to step in and engage the business to address these issues. Unless, the IT project has a data steward to engage, then it is very likely that data quality will become an important issue.

Not to far ago in the past, I did a search on a well known career search engine to find out how often data quality is seen as a requirement or function of the Business Analyst role. My results were disappointing. Only 19 positions with the term Business Analyst in the position name held the term 'data quality' in its' content. That translates into less than 5% of the postings that contained the term business analyst. Now I didn't go and look for data integrity, or MDM. It is what it is, a search done out of curiousity with not too much in depth data, I had kids to put to bed!

Out of curiousty, closer to the day, a free form search on 'data quality' and 'business analyst'. My results: 25% of business analyst type postings contained data quality in the body of their content.
Needless to say data quality is still a bit off the radar for many companies when it comes to defining the role of a business analyst.

With that said the aspect of including data quality in the role of a business analyst is there, but to what degree remains relatively unanswered. So to those business analysts and business analysts want to be's...remember Data Quality in your requirements, it's just good business.

Data Quality must be inseparable from the data. Good data quality at the requirements stage will positively impact:

1.Project success;

2. Data Integrity;

3. Efficient and effective support;

4. Sound decision making;

5. reduce business costs from errors and corrective actions; and

6. project longevity (something any good BA wants).

So if the business is not interested in Data Quality, make it your interest. It won't hurt, and the business will love you for it!


  1. Excellent post Daniel. I couldn't agree more that quality objectives should be captured up front as part of any business or IT project.

    Ultimately, the quality of the information is the determinant of the overall success of the solution. If you have a fast process producing crud, you are just producing crud faster.

    While it is nice to see "Data Quality" being included in job listings for Business Analysts, it is of little use to the business if there are little/no tools or supports or backing for the necessary actions to ensure quality is built in.

    But, and this is my personal experience, by starting from the ground up and, where necessary, looking outside the organisation for affirmation and support (such as to organisations like the IAIDQ and the network of peers there) you can do a pretty good 'Magyver' approach that gets you some quick wins and will (eventually) win you proper supports.

    [disclosure: I'm Director of Publicity of the IAIDQ, but got to that role by joining up to find a community of like minded people because I was building a Data Quality initiative from the ground up.]

  2. Excellent points, I have to agree, without the tools, or network support within the organization data quality will be a difficult company objective to establish. The Magyver approach will be the best option many times over. A grassroots initiatve in a way.