Try Before You Buy: Data Sampling Strategies to Keep Your Data Vendor Honest
Last week’s post covered the different types of data vendors and the pros and cons of working with each of them. Regardless of which type you use, you’ll want to get a sample data set and put your vendor to the test before you purchase your list.
Why sample? Once you’ve identified a few suitable vendors, try to resist the temptation to buy a big list from one of them and blast the whole thing right away. Remember that the goal of sampling is to make sure a vendor has what you need … and it’s also an opportunity for you to do some pre-campaign testing — to ensure both the quality of the data and the accuracy of any vendor guarantees (such as average delivery rate). Sampling can also provide a window for you to test your messaging before your full campaign launches. But remember, in order to be statistically valid, the only variable at play should be the list itself. And it’s very important to have your vendor create your sample on the fly … even while on the phone with you, if possible. Don’t give them time to “rig” your sample!
The three “S”s of smart sampling:
Size: In general, the size of your sample lists should be about one percent of the total size of list you intend to buy: ideally at least 100 records per sample. That should be enough to give you a good idea of the data’s consistency, but don’t sweat it if your samples are smaller.
Selects: If multiple contact methods, or touch points, are required to execute your campaign (for example, email, phone and mailing address), make sure your list samples include each as a select. This can also help you uncover any “hidden” costs you may incur later, as some vendors charge differently for different selects.
Send: For the best results, test each of the contact methods you intend to use. This can be time-consuming for phone and address, but you’ll at least want to test a handful — say, 10 from each list — to make sure that direct phone numbers, for example, are indeed direct calls and an IVR (interactive voice response) menu or a receptionist. And if, when you call, you get the “she no longer works here” response, see if you can ferret out when the person left. If it was years ago, you’ll get some insight into the “age” of the data. Track the results and compare them to see if your vendors’ delivery rates are on target with their guarantees. Then determine which list performs the best against your campaign’s goal metrics.
Using these strategic sampling techniques will help you choose the best vendor and list. Next week, we’ll look at ways to negotiate pricing with your vendor to ensure that you get what you’re paying for, before and after the sale.
Also read “Becoming a better data buyer” Part II: Why You Should Consider the Source When Buying Data
Anncy Graziano is ZoomInfo’s Manager of Data Services Operations.