• by Carlos Carvajal , April 12, 2011

The vision of personalized Web experiences has been around since the consumer Web was born. But for most Web sites and other digital channels, the promise is far from reality because it's often riddled with errors, requires extensive manual analysis and does not adapt quickly enough to the audiences' changing interests. Here are the most common mistakes made in Web personalization today, and four simple recommendations to help your company get the most from its own strategy.

1: Over-relying on profiles -- instead of actual customer intent.

If you're marketing effectively, a large proportion of your Web traffic will be new visitors; people you have no data on. But that doesn't mean you know nothing about them. In fact, if you're looking, you can pick up dozens of clues from each session. We call these weak signals intent clues because they tell you what the visitor is trying to do -- and they're available to you whether or not you know the name, address and gender of the visitor. Once you're tuned in to the visitor's intent, adding in profile data can be quite valuable. But without any insight about intent, profile data alone has limited personalization power.

advertisement

advertisement

Recommendation: Think about the intent clues your Web visitors are sending out as they click through your site -- things like scrolling, dwell time and on-site search behavior. Now think about how you can leverage these even if you don't yet know the visitor.

2: Depending only on expert rules -- instead of being guided by customer insight.

Personalization rules ("If they do this, give them that.") are commonly created by expert analysts who need to understand changing customer dynamics and the ins and outs of your business. They're rare and overloaded because of the effort required to configure and maintain effective targeting rules. This leads to simplifying the approach (e.g., show most popular products) which may not be the best strategy for engaging consumers online.

Recommendation: Deploy recommendation engines and other technologies that will automatically adapt to changing customer insights to generate the best content for each interaction based on intent clues. This will minimize the need for manual rules, and will improve effectiveness of personalization.

3: Relying on the wrong signals -- instead of responding to micro-behaviors.

Most personalization solutions are based on watching the obvious behaviors like purchases or clicks. The problem with focusing on clicks is that there are far too many of them and they don't necessarily tell you about the visitors' true intent. Alternatively, while purchases are a great indicator of intent, they are few and far between.

Recommendation: When you only pay attention to purchases and clicks, your personalization efforts are too little, too late. You can learn an awful lot from what happens between the big behaviors. Responding to micro-behaviors like mouse hovers, frequency of return visits, or time spent on certain pages while the session is still live is where the real uplift comes from.

4: Under-valuing search behavior -- and all you can learn from it.

Search is one of the most powerful intent signals any user can transmit. But most Web sites (and many personalization platforms) ignore it completely. It's a simple concept: the search phrase that brings a visitor to your site should be a primary driver of the content they see the moment they arrive. For example, if you know a visitor comes to a sight searching for one item such as a ski jacket, then automatically adapt the landing page to show ski jackets that will likely drive deeper engagement on the site.

Recommendation: Whether it's natural search, pay-per-click or onsite, the search term is a big, bright, neon sign that says: "I'm looking for X." Especially once the visitor is on your site, you should be able to personalize the overall Web site experience based on their original intent.

So take a closer look at your personalization strategies. You may find that you are guilty of one or two -- or all of these -- mistakes to some extent. By recognizing and fixing them, I guarantee you will see greater engagement and overall brand loyalty because your customers will have a more satisfying online experience that will keep them coming back for more.