Welcome to this edition of “E-Commerce Tip of the Month,” a blog series where we provide retailers like you with actionable advice to increase sales and improve the customer journey in your online shop. This month: semantic search.

We recently had a webinar I really enjoyed where we did live shop audits of top European retailers, and the audience got to vote which online shop would be audited (if you haven’t seen it yet, definitely check it out here). There was a list of 5 retailers to choose from but we weren’t able to get to all of them in the live event, so I figured this would be a good chance to fit in one of those that got left out: L’Occitane. For this month’s tip, I’ll be taking a closer look at a core feature of any online shop: search, especially related to semantic search.

Nowadays, people expect search to understand them. We’re all familiar with and are daily users of Google, where we type in whatever comes to mind and expect to receive relevant answers immediately. This has carried on over to our digital shopping habits, where we expect a shop’s search function to perform with the likes of Google – and unfortunately many retailers fail here. You’ve surely heard of the importance and difficulty of catering to long-tail search queries, which generally have less traffic overall but higher conversion rates.

Semantic Search Audit: L’Occitane

So let’s see how L’Occitane, a leading beauty retailer, performs here. We’ve reached the end of a long winter and one persistent problem both my husband and I are having is dry, cracked hands. I’ve bought various lotions and creams but my husband doesn’t appreciate using my super floral, pink tubes that make him smell a little too pretty for his taste. So let’s search for some more manly varieties.

If I type “hand cream” in L’Occitane’s search bar, I’m shown some relevant results:

Search results for L'Occitane when searching for 'hand cream.'
L’Occitane’s search results show relevant products for basic search terms.

So we’re on the right track, but I can’t imagine my husband going for about half of these. What if I put in something more specific? Let’s try “hand cream for men,” a long-tail search query that needs to take semantic context in order to provide relevant results for a fairly basic intent. L’Occitane provides these results:

Irrelevant search results at L'Occitane when searching for "hand cream for men."
Once the query becomes a little more complicated, the quality of search results deteriorates.

Above the fold on my relatively large screen, I see two rows of results showing products that seem to be targeted at men – but what’s shocking is that not a single one of them is for hand cream. There’s after-shave, deodorant and even cologne targeted at the word “men” but the context of “hand cream” has completely disappeared. If I didn’t know any better, I’d think that somehow they didn’t have anything for me. And this is forcing customers to do one of two things:

1) try again with a different search, which causes frustration;

2) leave the site and go to another.

Basic search is no longer enough – online shops need an intelligent semantic search tool that understands user intent and provides the most relevant results no matter the length or complexity of the query.

Semantic Search Audit: Douglas

Let’s try this same search in another online shop for beauty. And I’m really going to put this shop to the test: not only will I check their long-tail and semantic search abilities, but I’m also going to test their language-independence by searching their German online channel in English. The shop we’re going to look at here is a leading retailer in Germany called Douglas, one of more than 1,800 online shops powered by FACT-Finder.  For the following screenshots, my browser will automatically translate the text to English.

So, I’m going to do the same search: “hand cream for men” (in English). The search results show me this:

Relevant search results at Douglas when searching for "hand cream for men" in English in their German channel.
Douglas uses language-independent, semantic search to match user intent.

The first 3 results show me hand cream, with no pink or flowers in sight. In order to do this, the search had to not only identify English versus German (a cinch thanks to our patented Worldmatch® algorithm) but also understand the context behind my long-tail search in order to provide products that would actually match my intent. Impressive, right? The 4th result isn’t a hand cream, but it is a shaving cream for men – which matches the query better than ‘after-shave.’

Don’t let your online customers get lost in the semantics – provide them with an experience that understands what they’re looking for. Now if you’ll excuse me, I need to go apply some more lotion before my hands start cracking again.

Subscribe to the FACT-Finder newsletter