Computers and artificial intelligence have come along at an exponential rate over the past few decades, from being regarded as oversized adding machines to the point where they have played integral roles in some legitimately creative endeavours.
Frost & Sullivan estimates Australian businesses will spend $3.5 billion on search marketing by 2019. But what’s the point on spending so much on getting visitors to a site if they can’t find what they are looking for once they get there?
Surfacing the right products on a retailer’s Web page to quickly resolve a customer’s needs can be a tricky proposition, particularly when the retailer sells a raft of very similar products. It was this problem that drove the women’s underwear retailer Bras N Things to sign up with ecommerce specialists, SLI Systems.
According to Bras N Things’ general manager for ecommerce, Duncan Brett, his company’s Magento-based website has more than 1500 items listed, many of which would suit the description of ‘bra’. Hence a simple search on that term would yield more results than would be workable for the customer.
“That search box has to be able to surface the product on the page that is really relevant to that search term, as soon as the person clicks return, and if it’s not then that is a bad experience,” Brett says. “Our number one focus is showing every piece of inventory on the website and making sure that people can get to it quickly.”
SLI effectively takes over the search function for Bras N Things, making complex decisions about search results rankings based on previous customer behaviour and business rules, rather than the best semantic match. New products can be moved to the top of the results, incorrect spellings can be interpreted, and certain search terms can also take customers directly to specific pages.
“You want to give the users as much inventory as possible, but you don’t want to overload them and give them inventory that’s not relevant,” Brett says. “So if somebody searches for a blue bra, you want to show them every single blue bra, but you also want to make sure the blue bras you think that customer is after are at the top of the search results.”
Brett says the nature of Bras N Things products means he generally expects low sales conversion directly off the website, as the company’s customers typically like to touch and try on garments, especially when purchase for the first time. This diminishes the focus on site conversion, but Brett says SLI is able to test the effectiveness of SLI by linking search behaviour to sales figures.
“The percentage of our sales coming from those top search terms is increasing over time,” Brett says. “What that tells me is that while it is easy to capture the search terms, the challenge is to surface the relevant product according to those search terms.
“If the percentage of revenue that is attached to those top ten search terms is increasing over time, that means that SLI is really doing a better job of responding to those search terms and giving the customers the inventory that they are after.”
The company is now testing two extensions to SLI. The first, Learning Navigation, promises more effective on-site merchandising by making real-time decisions on which products should be surfaced within landing pages, based on traffic and search patterns and Bras N Things own business rules.
“That is currently quite a manual process and very subjective as well,” Brett says. “The data we feed to SLI on sales and conversion actually decides where an item goes.”
The second trial is of SLI’s Learning Recommendations service, which is similar to the recommendation engines that appear on Amazon.com. This will also replace an existing manual process, auto-populating results based on page views and cart content.
APAC regional vice-president for SLI Systems, Mark Brixton, says the company’s services use “the wisdom of crowds” to surface the best results, but can also be used to promote slower-moving or end-of-season products.
“If they have a product that is sitting out the back that is gathering dust and no one is looking for anymore they can push it to the top of the results and force it out the door,” Brixton says.
He says for another SLI client, SurfStitch, his company’s technology led to a 25 per cent increase in the time consumers spent on the site and 15 per cent higher sales.
Follow CMO on Twitter: @CMOAustralia, take part in the CMO Australia conversation on LinkedIn: CMO Australia, join us on Facebook: https://www.facebook.com/CMOAustralia, or check us out on Google+: google.com/+CmoAu