Strategies for Successful Testing: Segmenting Your Way to Great Results (Part 2)

By Adam Figueira

October 27, 2011

As a follow up to my Monday post about understanding the difference between testing and targeting (and why tests without targets often yield meaningless results), let's take a look at three examples of great Who/What (i.e., target/test) combinations involving the following business scenarios:

  • iPad users who experience different site functionality than computer users.
  • A retailer with a high rate of return on certain product types.
  • Promotional banners on the homepage.

1. (Who: iPad Users; What: Modified Functionality and Site Layout): Technographic information about a visitor reveals not only browser, screen size, and screen resolution—but the type of device being used as well. A key difference between traditional computers and iPads is that the latter do not support actions via mouseover/hover. But marketers can circumvent this problem with two simple tactics:

  • Provide a visual cue (such as a finger tap) to execute the action.
  • Automatically execute the action so that no tap is required at all.

Next, to account for the iPad's 9.7" screen, consider:

  • Increasing the size of navigation buttons.
  • Shrinking homepage heroes so that less top-to-bottom scrolling is needed.

2. (Who: Users viewing a particular Product Type; What: In-Store Pickup Options): Apparel retailers may find that a particular product type (e.g., footwear) is more likely to be returned than any other. (Misjudging size is one of the top reasons that online clothing purchases are returned.) Although many websites offer free returns to drive visitors to make an otherwise risky purchase, a better option may be to automatically detect the visitor's location and encourage in-stock pickup at the closest brick-and-mortar store. (Most websites still require the user to input a city or ZIP code, but this creates unnecessary work for the consumer.)

Concerned about the potential loss of site revenue? You have at least four options:

  • Tell the visitor if the price of the product online is less than it is in-store.
  • Highlight tax-free or free shipping offers (shoppers may return to complete the purchase online).
  • Treat clicks on "View In-Store Availability" links as a conversion event and measure performance.
  • Track the change in your company's monthly return rate and how much money is saved as a result.

3 (Who: A Specific Segment; What: Targeted Homepage Offer): This past Sunday, Wider Funnel offered a case for removing promotional banners from the homepage. To be sure, organizational politics often get in the way of effective site designs, resulting in homepages that leave everyone equally unhappy--including your customers. But the problem lies less in the offer banners themselves (or the politics that create them) and more in the lack of targeting. Offers must be tailored to the audience seeing them—particularly on the homepage. If you're a national home goods retailer, why should you expect to sell snow blowers in Miami, even at 80% off?

Although the most effective ways to segment homepage traffic will vary from business to business, in my experience the following are business-type agnostic:

  • New vs. Returning
  • Geography
  • Traffic Source

Below is an example of how one retailer geotargets (northern vs. southern hemisphere) to set up different homepage experiences:

Next week, we'll look at the third "W" in developing a great test campaign: When. I'll walk you through the best practices for using dayparts to deliver the right offer to the right customer at the right time.

Adam Figueira is a former product marketing director at Monetate.

Experience the future of ecommerce personalization

Book your personalized demo today.