Health & Beauty Glow-up: Matching Search Intent with Attractive Results

Most ecommerce search engines optimize product results for relevance, assuming keyword matches drive conversion. Our original research into the drivers of search revenue shows that relevance alone doesn’t convert. Results must also be attractive to the individual shopper.

After drilling into our data at the industry-segment level, we uncovered that 25% of visitors to Health & Beauty websites use site search, they’re 2.16x more likely to add items to cart and close to 3x more likely to convert than those who only browse.

And searchers are far more likely to convert when search results match the context of their query and the individual preferences you can detect about them (intent). Let’s decode how intelligent search matches attractive products to search intent.

Beauty Shoppers Who Search Click and Convert Higher Than Average

In addition to the search-vs-browse stats shared above, our original research found the average add-to-cart rate for visitors that search is 33.4%, and their conversion rate is 13.4% across all industries. 

But the beauty segment shows even stronger metrics: searchers have a 48.5% add-to-cart rate and 17% conversion rate.

Percent of Searchers vs Non-searchers on Health & Beauty websites

Searchers: 25.2%, Non-searchers: 74.8%

Percent of Searchers vs Non-searchers that add-to-cart on Health & Beauty websites

Searchers: 48.5%, Non-searchers: 22.5%

Percent of Searchers vs Non-searchers that convert to sale on Health & Beauty websites

Searchers: 17%, Non-searchers: 5.8%

The data demonstrates that beauty searchers have higher purchase intent. But not all searchers are at the same stage of the buying journey. The search queries they submit provide hints at their stage, and search systems need to pick up on these cues.

Search Terms and the Buying Journey

Health & Beauty is a category where a shopper may have strong intent to buy, regardless of whether they're at an early or late stage in their discovery. 

It’s easy to identify the pre-sold shoppers that arrive with a specific product in mind. They go straight to the search box (if they didn’t land directly on a product page in the first place), and they are detailed with their search query (i.e. they’ll use “long-tail” terms).

But visitors using broader “head” terms like “moisturizer,” “protein powder,” or “contour” may be just as eager to convert. They just need a bit more help choosing the right product.

Broad terms really need personalization

Naturally, high-level search terms match more products. And the more products, the harder it is to decide what to click. Here’s where relevance matching falls short: 100 products may all have “moisturizer” in their titles and descriptions, but they’re certainly not all best matches for every given shopper.

To best serve high-intent buyers early in their journey, the search system needs to tailor experiences to each user, and:

• Suggest more specific keyword phrases to steer them towards more relevant results
• Surface the most attractive* items first, by applying intelligent algorithms that understand the individual user’s context
• Present helpful filters and facets (and prioritize them by usefulness)
• Display relevant attribute badges (when available) to help them identify key specs
• Boost the brands and price points they’re most likely to prefer

*Constructor uses a variety of signals to determine attractiveness scores for each product in light of each user’s individual search case, including clickstream data, purchase history, first-party data, device, geolocation context, and product factors. Rather than pure keyword-matching or tie-breaking with collective trend data, each product grid position is personally ranked for the user.

Long-tail queries also need personalization

When users do get specific (for example, looking to repurchase La Roche-Posay Toleriane Hydrating Gentle Cleanser), they shouldn’t need to type every. single. character. Great search experiences pick up on cues as they’re typing through autosuggest to help them bypass search results entirely. 

When users ignore autosuggestions and submit their product-name-search, they often truncate their terms to include just the major keywords that matter (or the ones they recall). In these cases, search must meet users with high relevance to fill the gaps.

But not all long-tail searches are intended to recall a single, specific product. Queries like “digestive enzymes with pancreatin” still need a blend of keyword relevance and attractiveness, and terms like “best mascara for short lashes” (a little fuzzier) need semantic** treatment where keywords may not exist in titles or descriptions.

Sometimes, longer search terms contain keywords that don’t match enough products. Smart search knows when to expand results to include semantically relevant, near-match products that are also attractive to the individual user.

**Constructor determines semantic relevance (the intended meaning of a search query) by mapping shopper language to structured product attributes and historical engagement patterns. So, results reflect what people actually mean, whether the product data includes exact keyword matches or not.

Health & Beauty vs Everything Else

Compared to other product categories, health and beauty shoppers skew towards specific behavior and contextual trends. Use this cheat sheet for a quick summary of these characteristics: 

Putting it together

• Health and beauty buyers often search for specific brands and ingredient specifications (as in, they do their homework and often come pre-influenced)
• This category is highly personal, and shoppers are typically buying for themselves. They care about product performance and social proof
• Many buyers will use search to quickly locate products to repurchase (or stock up on)
• Most are shopping through mobile devices. This is where autosuggest can help them hone in on long-tail searches faster
• Many searchers include attributes in their query. It’s important to use semantic matching and include behavioral data to make results attractive

Implementing Attractiveness Optimization

Getting attractiveness right starts with a foundation of smart search technology. Your search application must continually learn from customer behaviors — at the global, segment, and individual levels — and tie these learnings back to product factors like inventory, seasonality, price, customer review sentiment, sales and returns data, brand popularity, and more.

Smart search puts all the pieces together in real-time to rank products by how likely they’ll get clicked, added to cart, and purchased by any given visitor, in context to the intent signals they display during their session.

Merchants who adopt this capability can expect to see an average of 86% more clicks on attractive results than simple relevance matching where results may be unattractive to the user.

What Happens When You Optimize for Attractiveness?

If your search results reflect not just what the user typed, but what they’ll ultimately choose based on clickstream behavior, past purchases, inferred keyword intent, and collective data, you:

• Increase conversions by surfacing products that match important attributes
• Build trust through great user experience and guided selling
• Reduce drop-off from zero-result or uninspiring pages
• Build emotional equity with shoppers who feel understood by your site

Ultimately, attractive search results drive more clicks and conversions. And healthy search performance is a beautiful, revenue-driving thing. 

See where your Health & Beauty search experience may fall short and leak revenue with our complimentary search experience audit.