CRM Data Quality for Sales and Marketing

This Summer I’m presenting CRM Data Quality for Sales and Marketing at the MIT Information Quality Symposium. I’ll post more on that in the future. As I was preparing my slides for the presentation, I remembered an exchange that led to the following blog quote. Thanks to Marci Reynolds, a former colleague, for consulting me when writing an article about Sales Data Quality. Marci and I collaborated on several projects and sales programs over the last few years and data quality was always a hot topic.

From Marci’s Sales Operations Blog

Mark Goloboy, Director of Global Data Governance for Monster.com recently shared, “ Poor data quality is a symptom of other issues. Organizations must focus on upstream data quality and correct business process and systems issues to succeed. Inconsistent information is the reason why many sales initiatives fail. One bad lead can cause a rep to ignore a whole program.”

Data quality needs to be a primary consideration in any effort, and should be discussed throughout. No dataset is perfect. Every lead has a probability to be bad, and every constituent is looking for an excuse if their portion of the project goes wrong. Manage data quality expectations and communicate issues to invalidate the excuses. A former boss would say, “take those bullets out of their guns.” That’s the best way to get results.

There is no perfect answer when confronted with data quality questions, so planning is key. Here are some common excuses to prepare for:

These leads stink. They are ALL DUPS.

No lead list is perfect. You should expect 5-20% of the leads in any list to be bad. Work with your vendors to have them help reduce these numbers. Test smaller subsets, and negotiate prices based on the percentage that are unusable. Arrange for “net” agreements based on your existing lists. Also, duplicate leads are often caused by poor data quality in the source system. When buying or gathering new leads be sure you have a strategy for matching to your CRM system, but also a plan to pilot usage of the leads to get feedback early and course-correct.

Your numbers must be wrong. They don’t match mine.

If during a program one report doesn’t match another then further research is needed. For what purpose was each report built? What system or reporting tool did the data come from? These questions could expose underlying issues and incorrect data usage. One caution… don’t assume your report is “correct.” You could miss a larger problem.

It will take weeks to review that data before it is usable.

This is a symptom of poor strategy. Matching and data cleansing needs to be automatically processed. You’re doing it wrong if you have resources manually reviewing large sets of data. You either need to mature internal processes or work with vendors who can help you do the matching externally. This needs to be built into the program cost, and should justify the program from an ROI perspective.