Illustrative example: cutting zero-result searches on a 1C-Bitrix home-goods shop

Illustrative example. This is a composite scenario written to show how AACSearch can be used. It does not describe a named customer, and the figures are illustrative, not measured results.

The problem

The shop sells plumbing, tools and homeware on 1C-Bitrix, with roughly 40,000 SKUs and heavy trade-offer structures — colours, sizes and kits that the standard Bitrix search handled poorly. Shoppers typed the way they speak («смеситель для кухни», with typos and Latin/Cyrillic mix-ups), and the built-in search answered with either nothing or an unfiltered wall of results.

Zero-result pages sat at 9.4% of searches — each one a shopper with purchase intent who found no product.

The move to AACSearch

The team installed the AACSearch module for 1C-Bitrix, mapped infoblocks and trade offers to index fields, and ran the first full index in a day. The module's agents now keep the index in sync, so deactivated items leave the results within minutes of a catalogue change.

The rescue ladder — keyboard-layout translation, transliteration, then progressively broader matching — took over the typos that used to produce empty pages. Facets (brand, price, availability, room) come straight from trade-offer properties, so they always match the actual stock.

What could change (illustrative figures)

• Zero-result rate: 9.4% → 1.8% (illustrative). The remaining zero-result queries surface in analytics, and the merchandiser answers them with synonyms that autopilot proposes with evidence attached.
• Search conversion: +12% (illustrative) over a quarter, measured as orders per search session against the pre-migration baseline.
• Seasonal campaigns run without developers. The merchandising team schedules boosts for heating and garden ranges; an A/B experiment checks whether the winter boost lifts revenue per search before it goes live for everyone.

Why it matters

The module kept Bitrix search as an automatic fallback, so the cutover carried no availability risk. The goal: the index stops being a project and becomes plumbing — nobody thinks about it, it's just current.