Searchandising
Searchandising is the practice of tuning on-site search to surface the right products through synonyms, typo tolerance, and business-rule boosts — a high-leverage lever since site searchers convert 2-4× the average visitor.
Searchandising
Tuning on-site search to surface the right products using synonyms, typo tolerance, and merchandising rule boosts.
Searchandising is the merger of site search and merchandising: shaping what shoppers see when they type a query in the search bar. It combines linguistic tuning (synonyms, stemming, typo tolerance, query understanding) with commercial rules (boost in-stock SKUs, promote high-margin lines, demote sold-out colours, pin seasonal campaigns).
The lever matters because on-site searchers are already past the discovery phase — they know roughly what they want. They convert at 2-4× the site-wide rate, so a single bad result page disproportionately costs revenue. Searchandising treats the search results page as a merchandised PLP, not a database query.
A standard search engine returns what matches the query. A searchandised engine returns what matches the query AND what the business wants to sell — weighted by stock level, margin, recency, and conversion rate. The two layers run together: linguistic relevance first, commercial re-ranking second.
On Shopify and Magento, default search behaviour is rough: "trainers" won't match "sneakers", "t shirt" won't match "tshirt", and an out-of-stock hero SKU can sit at position one for weeks. Searchandising fixes that with a synonyms dictionary, a typo distance threshold (usually 1-2 edits), and a boost-and-bury ruleset on top of the relevance score.
Search Revenue Uplift = Search Sessions × (CR_after − CR_before) × AOV
Search Sessions
Search sessions
Sessions that include at least one on-site search query in the period.
CR_before
Baseline conversion rate
Order conversion rate of search sessions before searchandising changes.
CR_after
Post-tuning conversion rate
Order conversion rate of search sessions after synonym, typo, and boost rules are live.
AOV
Average order value
Average order value for search-driven orders.
A Shopify apparel store gets 80,000 search sessions a quarter. Baseline search-session CR is 6.0%. After adding 120 synonyms (trainers↔sneakers, jumper↔sweater), fixing typo tolerance, and boosting in-stock items, search CR rises to 7.5%. Search AOV is €78.
Search sessions: 80,000
CR before: 6.0%
CR after: 7.5%
AOV: €78
→ €93,600 incremental quarterly revenue
A 1.5-point lift on a high-intent traffic slice compounds fast. The same percentage gain on the broader site would need a much larger absolute traffic base to match it.
Searchandising sits inside the wider PLP Optimization discipline — the search results page is functionally a dynamically generated category page, and the same merchandising heuristics apply. The difference is intent signal: a search query tells you the shopper's vocabulary, so you can also feed those queries back into category naming, SEO landing pages, and ad copy.
Typical search-session conversion lift by vertical after searchandising tuning
| Vertical | Site-wide CR | Search-session CR (baseline) | Search CR after tuning | Revenue share from search |
|---|---|---|---|---|
| Apparel & footwear | 1.8% | 5.5% | 7.2% | 20-30% |
| Beauty & personal care | 2.4% | 6.8% | 8.5% | 25-35% |
| Home & furniture | 1.2% | 3.8% | 5.0% | 15-25% |
| Electronics | 1.5% | 5.2% | 6.8% | 30-45% |
| Food & supplements | 2.8% | 7.0% | 8.4% | 20-30% |
The single highest-ROI starting move is reviewing your zero-result queries report weekly. Every zero-result query is a shopper who told you exactly what they wanted and got nothing — a synonym, a redirect, or a content card usually fixes it in minutes. Pair that with a no-results fallback that shows popular SKUs instead of an empty page.
Searchandising FAQ
Regular search ranks results by linguistic relevance alone. Searchandising layers commercial rules on top — boosting in-stock items, demoting low-margin SKUs, pinning campaign products — so results serve both shopper intent and business goals.
A search results page is a dynamically generated PLP, so the same optimization principles apply: sort order, faceting, in-stock prioritisation, and visual merchandising. Searchandising is the search-driven branch of PLP Optimization.
Searchers have explicit intent — they're past browsing and are looking for a specific product, brand, or attribute. That intent signal is the strongest you'll see short of an add-to-cart event, which is why search-session conversion rates are structurally higher.
A zero-result query is a search that returned no products. Each one is a lost-revenue event from a high-intent shopper. Reviewing the zero-results report weekly and adding synonyms, redirects, or content cards usually recovers a meaningful share of those sessions.
Shopify's built-in search handles basic relevance and limited synonyms on higher plans, but it lacks granular boost rules, typo tolerance tuning, and merchandising controls. Most stores past €1M revenue use a dedicated search app like Algolia, Klevu, or Searchspring.
Start with 50-150 high-volume mappings drawn from your top 500 search queries. Focus on regional vocabulary (trainers/sneakers, jumper/sweater), abbreviations (tshirt/t-shirt), and brand-to-category mappings. Expand monthly based on zero-result reports.
In-stock comes first — there's no point ranking a sold-out SKU at position one. Once stock-aware ranking is live, layer margin and conversion-rate boosts. Avoid aggressive margin boosting that overrides relevance; shoppers notice and bounce.
Track search-session conversion rate, search-driven revenue share, zero-result rate, click-through on first three results, and search-to-add-to-cart rate. Segment by query type (brand, category, attribute, long-tail) to see where tuning is paying off.
An edit distance of 1 is safe for queries of 4+ characters; distance 2 helps on longer queries (8+ chars). Disable typo tolerance on very short queries and on SKU codes to avoid false matches. Most search platforms expose this as a per-query-length setting.
Synonym and zero-result fixes show up in conversion data within 7-14 days for stores with meaningful search volume. Boost and re-ranking rules need a full traffic cycle (2-4 weeks) and ideally an A/B test on the ranking algorithm before you trust the lift.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.