Want To Sell More Sweaters? Check The Weather—And Your Multivariate Testing Platform

A product update from e-commerce optimization firm Maxymiser lets clients see how websites perform when it's raining in their customers' cities, among other things.

Maxymiser, a British-founded firm which specializes in helping e-commerce brands display slightly different versions of their websites to different customers, came out this week with an updated version of their MaxPredict tool, which is designed to show web designers how they can appeal to different user groups on the fly. The software essentially discovers microsegments of consumers and suggests small corresponding tweaks in a website appearance, like the color of a banner or the placement of a button, that supposedly will increase sales.

Mark Simpson, Maxymiser's president, told Fast Company that the company's product “serves up different combinations of content within a page. If you're testing within a page, you can see four variants of each area, and up to 266 versions of each site. There are a lot of combinations and you can see which experiences and combinations work best for different visitor tests.” According to Maxymiser, the software update also has the capability to let clients target customers according to local weather conditions--depending on the temperature outside, customers in Boston and Miami might see different offerings on a given e-commerce site. Based on the weather in a user's ZIP code, which is determined by IP address and other data, brands can offer different products--say, cold-weather clothing or different discount packages--in prominent places on the page.

Maxymiser works in a crowded industry of website optimizers and counts large brands such as HBSC, Holiday Inn, Epson, and Wyndham Hotels as clients. Maxymiser's microsegment work uses web cookies and IP address information to discover patterns in seemingly unrelated groups of visitors. The ZIP code of a visitor, the language they use on their computer, their web browser, the time of day they access a site, and their operating system can offer insight into their habits and even their economic class. The success of companies like Maxymiser is based around the fact that, using the web, customers can easily be sorted into demographic groups. For example, an e-commerce company may discover with the new version of the software that customers in suburban ZIP codes who use Chrome to look at shopping websites in the early evening share common habits and common cues that can be used to hook them into making a purchase. Using these super-small demographics, companies can determine common links between otherwise unrelated shoppers that allow them to offer specials which will hook them into making purchases.

Simpson added that a key part of the new MaxPredict release is that clients can find and target these microsegments with one click. In personalization tests for two client brands, Epson and Harry & David, the platform produced some interesting results. For Harry & David, MaxPredict discovered six new market segments of customers that could lead to a projected 6.22% increase in purchase conversions online; Epson found 15 new visitor and consumer segments that could lead to a 56.20% projected increase in purchase conversions.

Thanks to web cookies and the location-tracking capabilities of mobile phones, marketing and retail companies (along with everyone insurance firms to the government) are fixated on the insights data can give us into consumer behavior. In aggregate, consumer behavior has clear patterns and can be tweaked by even modest changes to a website--which means products like Maxymiser's will continue to have brands waiting to try them out.

Correction: An earlier version of this article referred to Maxymiser as a British company. Though the company was founded in the United Kingdom, they have since relocated their headquarters to the United States.

[Image: Flickr user Ines Njers]

Add New Comment

3 Comments

  • A/B and multivariate testing is great for websites, but for native mobile apps a lot of the Javascript-based tools for web won't work.

    At Splitforce, we've solved this problem by bringing the best practices of A/B testing to iOS, Android and Unity mobile apps through native frameworks. Check it out: http://splitforce.com.