We’ve seen studies, a few released very recently, that suggest that personalization of direct marketing materials, including print and e-mail, boosts response significantly – in some cases as much as 3-400% of the blind “A” side of the test. But there was an interesting study among them that perhaps showed the risky downside to this type of approach. Apparently, personalization CAN be overdone. But you’d be surprised how far you have to go . . .
The study we read told of a DM control package that had been declining from fatigue after nearly 6 months of heavy mailing. The commissioners of the test decided to test personalization, but in a new way. They personalized the letter in multiple places, and progressively added incidents of personalization as the letter progressed, per segment of the test. Segment One was mailed to a random list select of the house list, and contained a personalized greeting. Segment Two was mailed to a similar random list pull but the letter had a greeting and another incident further into the body of the letter that was personalized. Segment Three was mailed to a similar list and contained three incidents of personalization, and so on. All data was composed of elements contained in the address block of a five-line address containing job title and company name.
When the results were read and analyzed, the staff was astonished to see response rates continue to rise through segment 14! In a two-page marketing letter there were fourteen instances of personalization before the results started to flag. The package didn’t drop back below profitability until Segment 22! Clearly, people like reading about themselves, and as a result, feel you know them and are safe to buy from! The rise of the rate was roughly linear from one incident to 14, and tailed off sharply from 14 to 22 and dropped off less sharply after that. Seems it takes a lot to overdo personalization, at least as used in this study.
I get the feeling that most mail package probably wouldn’t get to 14 before dropping off, based on how much imagination is usually put behind this portion of the marketing effort. Just dropping the name in the copy 14 is not going to do it! To test this, simply read the letter out loud to yourself.
Have you ever been engaged with a retail sales person who had only gotten partially through their “Training” but got to the part where they were told to use the customer’s name whenever possible to help “connect” with the customer on a personal level? The overuse of your name in the conversation in unnatural places becomes annoying, then grating, then off-putting to the point where you want to stop the interaction and walk away.
With that experience in mind, read your letter out loud and see at what point your tolerance for the use of your name and other info seems to peak. Now, subtract one incidence, and that’s likely the sweet spot for that particular letter or package. Now you have a baseline, and can test above and below that number and see how accurate your initial read was.
Keep in mind, too that there is a gold mine of information in that address block, if you’re willing to make a few leaps. If you have accurate salutation and prefix information, you likely know marital status, gender, and to some extent, age. Ms. Brittany Jacobs isn’t likely to be 75, or married, or likely to buy orthopedic shoe inserts or support hose. You get some of this at least in a reinforcing capacity, from your list segmentation selections. And you can use all parts of the address block beyond the address.
There are data overlays that can be appended based on zip code that can give you a read on income, age, and other data down to the block level. Combine that with the job title info, and you have a pretty good picture of your prospect. Use street names, city and state info, job title, whatever you can to build a box of credibility around your offer. Now you just have to find a creative way to work those elements into the conversation so that they appear to be accurate but are really vague. If you’re having trouble with this, try imagining you’re setting up a fortune teller booth at a carnival – they use this same technique to read cues from you to weave a story that sounds believable. It’s like they KNOW you!
This may seem deceptive, or underhanded, sneaky, etc. but ethically all you’re doing is making some educated guesses, and feeding the information you have that is freely available back to the audience on a specific basis – nothing sneaky about that! Let your creativity run wild, build a conversation with as much credible personalization as you can – you’ll find it gets harder once you get above 7-8 incidents. And now that you know that the average person’s tolerance level is higher than that, go for it!
Test personalization on your next package, and see how much it boosts your response – in today’s digital world, there’s no reason not to . . .
If you found this information valuable, and would like more of it, be sure to pick up your copy of “The Marketing Doctor’s Survival Notes” . . .
