I’ve touched on the importance of data quality before - even made it one of my three recommended New Year’s resolutions - but I think it’s worth returning to a subject that in the opinion of Bernice Grossman of the DMRS Group is so important that it ought to represent an asset line on your balance sheet.
Bernice contributed to a recent article on the subject by David Kilpatrick of Marketing Sherpa, and I think that it’s worth sharing some of his conclusions - and offering some of my own. The Marketing Sherpa article focused on the need for data hygiene, but I think the issue is deeper than that - it embraces the whole concept of data quality.
Fact: Data Goes Out of Date
The simple fact about any prospect or customer data is that it goes out of date - in some situations faster than others, but the clock is ticking constantly. You might be surprised how quickly the rot sets in. Scott Holden of Salesforce.com believes that as much as 70% of a marketing database can go stale in a year.
The absolute rate of decline might be arguable, but the trend is inevitable - and relentless. Without regular data maintenance, a large and growing percentage of your marketing effort will be misdirected. You’ll be reaching the wrong people with irrelevant messages - and failing to reach the right ones.
What Data Do You Really Need?
Any data quality initiative needs to start with a clear sense of what data you need to know, and how you plan to use it. There’s little point in capturing soon-to-be-redundant data “for data’s sake” - but it’s equally true that just collecting demographic data means that you’ll be missing some invaluable insights.
Take a step back, and determine what data is going to be most valuable to you. If you’ve completed an Ideal Prospect Profiling exercise, you’ll probably have recognised that certain structural, behavioural or environmental data points have tremendous predictive value when it comes to identifying your most valuable prospects and understanding what issues are most likely to stimulate a response.
Then prioritise the information. What “must have” data do you need about each organisation, and each contact? Which roles are most important? And what about the “nice to have” information - and how much might it be worth for you to collect it? These decisions can help you make intelligent decisions about how to prioritise your data quality investments.
Eliminate Islands of Information
Duplicated, inconsistent data benefits no-one. Your goal must be to establish one version of the truth, shared and accessible to all. If your applications force you down a multiple database route, then you must define a master version and establish effective mechanisms for sharing information and synchronising changes.
You’ll almost certainly want to find a way of associating behavioural information with the raw data - who clicked on which link, visited which webpage, and downloaded this or that piece of collateral. Applications are getting much smarter about this - and if your current platforms don’t allow you to collect, collate and share this information, it may be time to change them.
Fill In the Data Gaps
Make everyone feel responsible for contributing quality data into the system, but make one person responsible for monitoring and reporting on data quality, and for driving data quality initiatives. Key metrics include the completeness of data vs. your “must have” and “nice to have” goals - as well as how recently potentially volatile data was validated.
With the right attitudes and processes, checking the data and filling in the blanks is a less intimidating task than it might at first appear. Of course, you can call upon specialist data cleaning and appending services, but my recommendation is to make it a team sport. Make everybody feel involved. Encourage data sharing. Reward your best contributors. Consider a “crowdsourced” solution like Jigsaw from Salesforce as a potential external resource.
Take Action Today
If you don’t have a formal data quality initiative, can I suggest that you start one today? Bernice Grossman was right. Your data ought to be one of your most valuable valuable assets. But failing to actively manage it will turn it into a liability.
How have you addressed the data quality challenge? Have you any top tips to share with readers? I'd like to hear your experiences...