I will never forget hearing the song “Celebration” by one of my favorite bands Kool & the Gang immediately after the Atlanta Braves went from worst to first in 1991 and won the NCLS (yeah, I know, they lost the World Series that year). For you to celebrate in 2014, you need to understand how to calibrate the results of your marketing automation initiatives.
About 20 years ago, I was head of new customer acquisition for one of the country’s most successful direct marketers, Bear Creek Corporation (Harry & David, Jackson and Perkins, and others). My staff included three PhD. statisticians who did nothing but analyze results and make recommendations about who should receive catalogs and other mails from our brands.
Our marketers were aggressive about testing new offers, pricing, and it wasn’t unusual for a 1,000,000 piece mailing to include as many as 10 tests. Each test required a minimum universe of 5,000. One of the tests was always something we called a “null sample.” This group was sent the same offer as the control group, and generally the null sample performed about the same as the control. This assured us the other test samples were reliable and any differences in the test results were valid.
Applying this approach to B2B companies today, and more specifically to marketing automation, there is a strong case for calibrating results. A simple way to do this is to set up a null sample test by taking a percentage (10%) of random prospects and keeping them out of marketing automation. (Yes, this means you should actually pick up the phone and talk to this group instead of waiting for them to provide you with their digital body language.)
Here are three reasons to try this:
- Not every senior executive wants to be treated like the human equivalent of a pinball, capturing your attention only when they hit the right bumpers and score enough points. What this means is that marketing automation tends to drive smaller deals with lower level decision makers. By actually reaching out and talking to people you will find your deal size increases. A former client who abandoned us for nine months out of a 10 year run saw their pipeline dry up and their deal size fall to about a third of what it had been when they were working with us. They came back, deal size went up and they were sold for a healthy multiple to a big company in 2013.
- We recommend that you keep previously captured data (such as any visits captured in marketing automation) when you pull the “null sample.” In addition to providing fodder to keep the conversation relevant to the prospect, it will allow you to evaluate, validate or calibrate the scoring that you are using in the marketing automation scoring. For example, you might find that a whitepaper download skews to lower level decision makers and that the score for this activity should be lower than it currently is. Conversely, you may find that video testimonials are viewed by more senior executives and should be scored higher.
- By actually talking to prospects, you will glean considerably more insight into their pains, their perceptions about solutions and their actual plans than you ever will via scores from marketing automation alone. I’m not suggesting you ditch marketing automation, just add the human touch back into the game and score some real home runs. Now, that’s something to celebrate!