The value of high-quality data cannot be lost on any marketer. It’s our job after all, to know the market – what customers need, what they expect – fully and completely. Top-quality data enables that knowledge. As a marketer armed with perfect data, I can engage large numbers of prospects with pinpoint accuracy. No prospect is left behind. When people respond, my error-free information is instantly relayed to Sales so they can immediately act on the opportunity. When they reply, I get timely and structured feedback – the impact can be measured immediately. Look ma, no hands!
As tempting as it may be to pursue “perfection” in data quality, the practical limitations make it an impossible dream. Vicki Raeburn describes the pitfalls in her blog post, Understanding the Cost-Benefit Quality Curve. Every “just-noticeable improvement” in data quality involves an exponentially higher cost. Given the unavoidable tradeoffs of cost and quality, Raeburn recommends that marketers seek to yield “useful knowledge, not perfect data.”
Let’s explore this through a familiar example from B2B marketing. A company sells its wares to other companies operating in a variety of different industries. In their database, only 50% of contacts are tagged to an Industry. How much would it be worth to tag more of these contacts by Industry, say 60%? The improvement yields a noticeable increase in list size, providing better market coverage for their vertical campaigns. What does it really cost to realize such an increase? They discover they can tag more contacts by simply rearranging the Industry data they already have on file, using some clever techniques that run on a fully automatic basis. Presto, they say, let’s get it done!
Having a taste for this initial success, they take it up a notch. They set a new goal to tag even more contacts by Industry, from 60% to 70%. This gain might represent an equally noticeable improvement: bigger list size, better market coverage.
To reach this new goal, they now need to start thinking about how to consistently capture or derive Industry for every contact in their database. Every web form used for lead generation must capture Industry from now on. They scrutinize every upload to the database, to make sure the Industry values are included. This is not enough to get them to their goal, though. To get their existing contacts updated need to append missing Industry values from an outside vendor, such as Hoover’s. The numbers are getting closer now, but the costs are adding up fast too. Is the improvement worth it? Given the exponential effort level, they might decide the project will take a backseat to other priorities.
For the perfectionists out there, “good enough” is no cop-out. It’s the best way to rationalize data quality projects. The economic benefits of better data may be hard to quantify, but projects aimed at improving data quality must be both feasible and worthwhile to be successful. The excellence comes in executing well against goals that have been carefully considered for their impact on the business.
Reflect on this last idea for a moment. Data quality begins well before any change is made to your software. Does everyone in your company — marketing managers, sales managers, and senior managers alike — agree on the priorities for data quality? Are these priorities feasible? What are they doing to help you with these priorities (not just cheering you on from the sidelines, but specifically)?
One way to set worthwhile and feasible goals for data quality is to start by hosting a workshop for your data end-users. I personally like to use a survey to structure the feedback. When done diligently, a survey fosters an amazing discussion that lets you quickly spot the pain points and set goals that will be relevant and impactful.
Try it! Take the quiz below.
Please rate your level of agreement with the following statements
(strongly disagree, disagree, agree, strongly agree, etc.)
1. When I use our house list for my campaigns, I feel I am getting good coverage of my entire target market(s).
2. I can use our house list to accurately target our known customers.
3. I can use our house list to accurately target our target accounts.
4. Our house list allows us to target contacts who are involved in active sales opportunities.
5. Our house list allows us to target contacts who were involved in discontinued (lost/no-decision) sales opportunities.
6. When I add in my targeting criteria (Industry, Product, Job Function, etc.), the list gets so small that it is practically unusable.
7. My campaigns generate complaints from unintended recipients, but I have no way to suppress these types of contacts.
8. I tend to rent lists (or use an outside agency for my list) because our house list does not provide enough coverage for my specific needs.
9. If I had to choose ONLY three fields to segment our customers and prospects for my campaigns, these would be:
How did you fare? Which force-ranked priorities did you list for question 9? Depending on the culture in your marketing team you may have your peers openly share and discuss the answers, or keep it anonymous.
Does your company formally prioritize its data quality projects, or is it more reactive/ad-hoc?