Finding the better in the different

By Ryan Dahlstrom Staff Manager, Interactive Windstream Communications


  Everyone is unique. We all think, act, and respond differently to situations than the person sitting next to us in the office, on the bus, or at home on the couch. So, assuming this to be true in each and every situation in which we find ourselves, why don’t more websites take this into account when serving up content for users to consume? Why do we continue to be faced with one-size-fits-all (or at least one-size-fits-enough-to-remain-profitable) experiences?

The answer, as I see it, has a lot to do with Mr. Ives’ thought. It is so easy to simply serve content differently; it’s an entirely different matter when it comes to doing so and achieving better results. Why spend the extra effort to concept, create, and stand up something that you don’t know will work? Most business owners won’t. So then, how do we go about figuring out what different works?

My suggestion: segmentation testing. “Huh? What’s that,” you may be asking. Simply, it’s finding the better in the different by trying different things with different users and measuring the results. As a business owner, you know the ins and outs of your audience. You (should) know who is most profitable, who complains the most, who are your greatest advocates. You (should) know who prefers Product A to Product B. Use this knowledge as your starting point when figuring out what to test and where.

Three things to keep in mind when choosing segments to target & test:

  • You must be able to easily identify the segments that you want to test. It does no good to conceptualize a test that targets 50+ men when you have no way to actually identify this group once they hit your site.
  • You must be able to do something once you learn something…and be able to do it over and over and over again.  (Big bonus points if you are able to do something differently in other channels with what you’ve learned)
  • The segment you choose should be large enough to be profitable. There’s no sense in dedicating design and programming hours against a test (no matter how successful from a data standpoint) when that test won’t yield enough positive lift to cover the development work.

Ok, so you’re sold on the idea, but are having trouble figuring out this whole segment identification piece. Here’s some ideas of things that you could try:

  • Geography – try selling ‘Pop’ to a New Yorker or “Sneakers” to someone in the mid-west. They won’t know what you’re talking about, let alone purchase from you. Testing something as simple as product references could go a long way.
  • Life Cycle – what is the sense in testing ‘Offer A’ vs. ‘Offer B’ to someone his or her first time back after purchase? They just purchased! How about trying a “Thank You?” Even an upsell test wouldn’t be so bad…but don’t try to sell them what they’ve just bought!
  • Traffic Source – someone clicking on a flashy banner ad is likely in a much different sales funnel location than the person organically navigating to your URL.
  • Time of Day – we do different things at different points during the day, which means that our frames of mind are very different throughout our waking hours. Use testing to appeal to these variations…flash sales or “pressure tactics” during lunch hours when people don’t have a lot of time to think about what they are doing or move from one thing to another very quickly.
  •  Weather – if you knew that the “storm of the century” was coming early next week would you buy a new jacket today if it was on sale with free two-day shipping? The likelihood is a lot greater than purchasing new swim trunks at the same offer. Unless, of course, you’re headed to Aruba to beat the snow…

So then comes the age-old question, “What do I test?” Honestly, it doesn’t matter. Test something. Test anything…

  • Headlines
  • Button Colors/Shapes/Sizes/Font size/CTAs
  • Imagery & Artwork
  • Body copy – especially useful when “echoing” search terms to capitalize on intent

…just get away from thinking that one-test-fits-all. Please. So, go. Get out there. Start looking for the better in the different.

Save $100 when you register

This post was originally published in

Image credit: Zoha.Nve via Flickr


About the Author

Dahlstrom-vRyan Dahlstrom leads digital customer acquisition, support, and engagement at Windstream Communications. His experience includes overseeing digital strategy for a Fortune 500 company, managerial and leadership skills in e-Commerce, SEM, SEO, digital media, A/B Testing, web optimization, and customer support.

Leave a Reply

Your email address will not be published. Required fields are marked *