If you’re thinking American Litho has gone into the business of fortune-telling, let me set you straight. We’re still working hard to be your #1 resource for omnichannel and direct marketing solutions.

But when it comes to predicting who, among the millions of potential customers out there, might be ready to buy your products and services — then yes, we’re definitely looking to get inside the heads of your customers.

As marketing evolves in the digital age, we’re learning a lot from online partners who’ve developed powerful big-data approaches to understanding customer attributes and preferences.

We’re interested in these techniques because we know that direct mail campaigns delivering the most appealing offers to the right people at the right time get up to 4 times as many responses as those sending out static messages to the same customers over and over again.

In competitive times, relevancy is crucial – and we’re pursuing every means of helping top brands achieve it.

The mirror image of your best customers

Direct marketers are having great success with a predictive technique known as look-alike modeling, which digital marketers have used for years to identify potential new customers.

The concept behind look-alike modeling is simple. The people who most closely match the qualities, preferences and behaviors of your current customers are likeliest to become your new customers.

The more data points used to compare current and potential customers, the more precise we can be in putting together a mailing list filled with people who can’t wait to learn more about your products and services.

Recently, we used this technique to help an online clothing marketer reach millions of new buyers. The results showcase all the things we love so much about predictive marketing.

Good, better, best: a look-alike success story

Our client was working with a customer list of 1.5 million names with only a few points of demographic data available for each. Five mailings sent in 2016 garnered a respectable response rate, but the marketing team knew the brand needed more sales in order to thrive.

In 2017, they entrusted their entire database to American Litho’s in-house data team, who ran the records against a national database featuring more than 300 attributes for each individual. From the rich data obtained, the team was able to construct a look-alike model that would identify potential customers that closely matched current customers based on hundreds of data points.

The resulting prospect list was broken into deciles from 1-10, with Decile 1 being individuals who most closely resembled the brand’s existing customers. In January we ran a test mailing to 1.1 million prospects in Deciles 1 through 5. Our client was thrilled when the campaign achieved a 17% lift over their 2016 efforts.

Tighter focus, bigger results

Two months later we focused solely on individuals from Decile 1, capturing all names available in the national database. More than 2.1 million mail pieces went to prospects on this second list.

We expected to see results similar to the January test — but it turns out we were wrong.

Instead, this highly targeted mailing showed an astonishing 82% improvement over the previous year’s campaigns.

Cost per acquisition also dropped by nearly 75% from the test campaign to the second campaign. And with the average first-time purchase for new customers at about $200, total potential sales equaled $2.2 million from the second mailing alone.

What can we achieve for you?

American Litho customers in retail, financial services and many other fields are getting amazing results with look-alike modeling. Ask us for a free review of your current campaign to see how we can improve the precision and cost-effectiveness of your mailing efforts.

In our next post: How high-quality personalization teams with accurate mailing data for the ultimate boost in direct mail ROI.


Mike Gruper

Mike Gruper is a print and omnichannel expert working hard to deliver superior quality for American Litho clients. Reach him through our Contact Us page.

Facebooktwittergoogle_plusredditpinterestlinkedinmail