The issue starts with identification. In fact, our experience shows that many brands can only identify about 20% of their site visitors at best. That means that most personalization programs ignore four out of every five visitors. Identification doesn’t have to be black or white. It’s not limited to paying customers and loyalty program members. Brands must start to build identities for every visitor using relevant bits of insight that can inform an ongoing campaign.

An Old Problem That Hasn’t Gone Away

I recently came across a piece of research from VB Insights from 2015, which names identity as a major problem for personalization. The report stated that 80% of marketers surveyed don’t understand their customers beyond basic demographics and purchase history. Many marketers struggle with these same issues four years later.

There are three marketing norms that are hindering progress.

The first is that personalization programs, and in fact many content programs overall, are set up as message testing shops.

The second is that many marketing programs are not continuous, but are instead based on individual actions or specific campaigns, and so data isn’t collected in any way that’s usable for ongoing personalization.

The third is that many marketers are not collecting bits of insight about new visitors that could be valuable as part of a personalization strategy on the second, third or fourth pages of a session, or on a second visit, even if they aren’t an identified customer.

These problems limit the ability for marketers to learn more about the four out of five people visiting their site that they don’t know, and it stops them from becoming more relevant to the one that they do. Luckily, marketers can change their approach and see tremendous improvement.

Solve the Identity Problem Backwards

Marketers face a lot of uphill battles in their quest to personalize. They have data that’s too old or too hard to access. Creative and content resources are limited. Their marketing platforms aren’t integrated. They risk alienating customers if they make a misstep. As a result, marketers often find themselves in an A/B testing ‘Groundhog Day’ scenario, where they set up a campaign and create a test to pick which message works best for most consumers, and then do that all over again.

To break out of this, marketers fear they have to wait for big data projects, CRM integrations and creative resource miracles to occur, but that’s not the case. It all comes down to flipping the approach in reverse. Marketers should start with the customer, not the campaign.

An online university did this when they created lifecycle campaigns that followed the journey that their students take from enrollment to studying for tests to engaging in the alumni community. At each stage, instead of different predetermined content, they set up campaigns that would first gather information about an individual that could then be used to personalize content later in the journey.

Find the “Glue” To Connect Campaigns

The online university now had insights they could use for the next interaction, and there they could gather more insights, which made every subsequent interaction more relevant. They could then analyze campaign performance based on which data was most valuable to collect and in turn, which type of personalization was then most relevant. This approach works for rich personalization on a web site, in email, and across channels.

But what about campaigns for people who haven’t become paying customers yet? Retailers can focus on picking up clues to inform personalization in a more subtle approach. For example, using search behaviors, browsing history and third party data.

Trusting Personalization To Do Its Job

This approach does require marketers to start thinking about personalization as something that needs to be running in the background, so that the system can learn and improve interaction by interaction, page by page, email by email. The goal is not to replace the work of creatives, and the outcome is not that of a runaway AI-driven machine talking to customers with no human intervention. Rather, “always on” personalization can help marketers gather insights, focus their creative approach and test more messages, more often.

Asking a marketer to embrace the cyclical, ongoing nature of personalization can feel a bit like asking a runner to jump on a moving treadmill. It’s hard to understand where to start and how to do it without getting hurt. But it’s really not like that. Marketers can start with a few small questions, such as “what do we know about the four out of five people we can’t identify, and what is valuable about those insights?”  These questions can lead marketers down the right path, to identifying more people, and personalizing more often. After four years of starting the identification problem in the face, it’s time for marketers to take the first step towards a better approach.