Four Steps to Relevant Communications

I’ve never met a marketer who didn’t think that relevance is essential in marketing communications. It’s a no brainer. Yet the overwhelming mass of direct marketing campaigns is what we call “spray and pray.” Everyone gets the same message, the same offer. Where’s the relevance in that?

Are these marketers merely giving lip service to the idea of relevance because it’s fashionable? I don’t think so. I believe they sincerely want relevance. So why all the spray and pray?

Maybe they don’t fully understand how much relevance raises response rates. It has a huge effect, often generating lifts of 100%, even 400%. If they knew those numbers, they’d work harder to deliver relevant campaigns.

Or maybe they don’t know how to do it.

At first glance, delivering relevant campaigns can seem impossibly difficult. If you’re sending email to a million customers (or even 10,000 or 1,000 customers), how do you make each one uniquely different, uniquely relevant? No company has the resources or the personnel to manually create that many different messages.

Full disclosure moment: Until a few years ago I didn’t think it was possible either. Then the obvious smacked many of us in the face: It’s not the message that drives uniqueness, it’s the product offerings. Your company may have only a few basic messages, but you have thousands of SKUs. You can create individualized direct marketing communications by offering each of your customers exactly the set of products you’ve determined are of interest to them.

Creating those unique-to-each-customer product offerings isn’t simple, but neither is it beyond the reach of most companies. Here we offer a four-step path to help yours. The first step is psychologically hard for many. The second is a technical challenge. But the last two are straightforward housekeeping.

The psychological challenge

The first step is simultaneously the easiest and the hardest. It’s easy because it’s only a change in your point of view. There is no programming or technical work involved. But it’s psychologically difficult because you’re asked to change the way you’ve been marketing for many years. You need to switch from product-centric marketing to customer-centric marketing.

Instead of starting with a set of products to sell and finding customers to buy them, start by determining the interests of each customer and find products in your inventory to match those interests. Instead of measuring success by determining which product offers were most often selected, measure success by revenue per customer. Once you’ve made this mind shift, the path to relevance is clear.

The technical challenge

The second step is to build a recommendation engine. This is the technical challenge, because you need to do it at the individual customer level. The good news is that you already own the necessary data. Recommendation engines are algorithms or business rules that transform transaction data into purchase propensities for each customer. Transaction data is the most accurate predictor of future behavior, and this goldmine is sitting there in your enterprise resource planning and point-of-sale systems. You have it for every customer, so predictions are available for each one, assuming enough transactions.

You may decide that deciphering the patterns of clicks on your website or digging through the transaction data are not the best use of your resources.  If so, use a search engine to see the range of software available for purchase, from low cost to extremely dear.  Avoid those based on the “wisdom of crowds” and find one that calculates purchase probabilities for a customer based on that customer’s transactions.

The last mile

The final two steps are easier. First, collect your digital assets, the graphics that depict your products, and make them digitally accessible. Usually a printer or ESP with variable data capabilities has a system in place to handle this for you. As your campaign material is being produced, whether it’s a postcard or an email, the vendor needs to mate the images with the recommended products as each postcard goes through the printing press or as each email is delivered.

Finally, you need to create a file that instructs your printer or ESP how to produce the collateral.  There will be one record for each customer, showing the address, any messaging, which products will appear in template slots, and where to find the digital asset corresponding to that recommended product.  Typically, this involves some database programming that can take up to several hours. When the process is automated, it can be done in minutes.

Producing a million unique emails is much easier than it looks. The path outlined here can pay for itself many times over in higher response rates, more revenue, and better customer retention.