As shopping has continued to be transformed into being online-first for a huge percentage of modern consumers, it’s also become clear that search has become absolutely central to that retail e-commerce experience. Up to 43% of visitors to retail websites immediately go to the search bar, and those consumers are 2-3x more likely to make a purchase than site visitors who don’t use the search bar.
The retailers and brands with whom we work often see numbers like that and it tends to focus their thinking very quickly. They know that consumers have many choices, and will quickly abandon an e-commerce site that doesn’t present them with what they want. In fact, the average fashion and apparel e-commerce conversion rate is now 1.4%, which is well below the overall 2.68% e-commerce average. When things are going this poorly, the understandable first instinct of many brands and retailers is to “rip and replace” their search vendor, and upgrade to a new one.
This, we and our customers have found, is the proverbial cart-before-the-horse problem. When a retailer or brand is trying to solve search problems by migrating from Shopify to Algolia or Bloomreach, or to Demandware, or vice-versa, they may very well be making an important and impactful decision to better optimize search for their users. Yet without a robust product attribution taxonomy that takes into account how real people search and shop for products in the real world, they’re missing half the problem. They have a lovely and expensive cart sitting there with no horse to pull it.
Plug-and-Play Product Attribution Data That Converts
What’s needed is not an either/or strategy but a strategy of “both”. Even the most powerful back-end search provider is only as good as the data that is fed into it. We call that the “garbage in, garbage out” problem, and it’s a real one that can mitigate many of the gains that come from migrating to a new search provider. At the end of the day, consumers still want their searches to land upon those items that they have in their heads, but which their fingers may lack the pinpoint-perfect descriptions for. Brands and retailers who take in product attribution provided to them by manufacturers and distributors – and stop there – aren’t able to push the language that customers actually use in their searches to that shiny new on-site search engine, and then wonder why conversion is still stuck at an anemic 1.4%.
We’ve seen phenomenal success across the brands and retailers that Lily AI works with once they’ve set up the rest of their e-commerce stack for success with platform-agnostic, plug-and-play product attribution data that can feed every step in the retail value chain. Their extensive product taxonomies are the key input to some very different search providers, and reflect their desire to not only have a fluid, customer-friendly search bar experience on the website, but one that actually returns the sort of results expected by their shoppers – even on some of the longest of long-tail searches.
Speaking in this language of customers, rather than that of the manufacturer or distributor, has hugely beneficial ramifications across the entire retail value chain. A product with two to three attributes can, through visual, AI-driven tagging and a carefully-designed taxonomy, immediately unlock its sales potential to become a product with 10-15 attributes.
When product attribution is done correctly in this manner, it instantly drives better site search, filters and facets, product recommendations, and demand forecasting. On-site conversion goes from that anemic fashion/apparel industry average of 1.4% to 4%, to 5% and more – and just as importantly, the ability to sell at full margins is enhanced because demand is now forecast using the language of the customer with customer-driven attributes – not generic attributes.
If your company happens to be in the market for a new search provider today, great. Let’s talk about how to really make it pop by adding exceptional depth to your product catalog at the same time.
6 Steps to E-Commerce Site Search That Converts
