Amadesa

Segmentation is no doubt a vital next step in your optimization effort, if you are not doing it already. By far, this can be one of your best strategies, by helping you focus on visitors that are more important to your business and creating content that is more directly relevant to them. Everybody is recommending it, and it makes sense – dividing your visitors into groups by interest and creating and delivering the information to best suit their needs. But damn, why does it have to be so hard? Did all the people recommending it and talking about it ever think of what it takes to segment traffic on a site?

We are talking about hundreds of traffic sources, a multitude of keywords used, tens of content categories visitors went through, some of them made a purchase while others didn’t, some put an item in the cart but others did not, and so on. I could probably come up with a few dozen more data points that make it even more complex. And what if the site is an international one? Now I have to consider different time zones, interests and language…Try to take that into consideration, all you segmentation experts.

And the cost for all this effort? It’s the time it takes to configure all the segments you want to track and the time you need to review their performance. But wait, you also have to track it all over time, consider trends and seasonality, etc. Sounds like a full time job…as if anybody had any time to spare.

We have all experienced it before. We get tons of data from the analytics package but it doesn’t necessarily help us in getting better results and it’s not always cost effective. “Oh, don’t get me wrong, having more insight into what’s happening on my site is crucial, I cannot do without it, but I need help in sorting all that information and making sense of it.” It’s almost a problem on every site we optimize. Sorting the enormous amounts of data, whether related to simple segmentation or going the next step towards behavioral segmentation, proves to be a very difficult task for the site’s marketing team.

For me, as an optimization manager, it’s not any different from what the site experiences – trying to figure out how to segment traffic is a very challenging task on every new project and it’s costing the site lots of analysis and testing hours. So what we’ve done is taken a deeper look into this problem and in a different manner, by developing a personalization engine to help streamline the segmentation tasks on a web site:

• Sort through the segments that drive more conversion and revenue on each page or funnel we optimize.
• Automatically display the most relevant option for each visitor.
• Expose the major groups of visitors that drive the site’s business, and the ones that are most neglected or overlooked.
• Help the product and marketing teams create offers that will drive more response from visitors.

There is still a long way to go. So far our personalization engine is engaged in presenting the most relevant offerings on a site’s home page and has shown considerable lift in click through rates on displayed banners. We are also in the process of implementing these abilities on the checkout flow for presenting each visitor with an optimal version that drives more conversion. Next steps will include implementing it with our recommendations efforts and all other optimization tools included in our product suite. Stay tuned.