About Irankiai.lt
Irankiai.lt is Lithuania's most trusted specialist for professional tools - carrying Makita, Bosch, DeWalt, Milwaukee, Metabo, Festool, Karcher, and others. A long-established, top-rated specialist retailer.
The core customer base is professional buyers - electricians, builders, mechanics, woodworkers - who search with precision and have zero patience for a search bar that can't keep up. They know exactly what they want, often down to the model number, and expect to find it on the first try.
The Challenge
100,000+ products overwhelming Magento search
Magento's native search wasn't architected for this scale. Query performance degraded, relevance deteriorated. Professional buyers got poorly ordered results or nothing at all - at which point they call support, or leave.
Professional queries failing consistently
The search patterns of professional buyers are specific and unforgiving:
- Longer multi-attribute queries: "Makita 18V brushless cordless impact driver LXT"
- Product codes and model numbers: "DDF484RTJ", "HR2631FT13"
- Partial product codes: buyer knows "DDF48" but not the full string
- Feature and spec searches: "18V battery compatible with BL1830"
- Typos in technical terms: "Festol orbital sander", "Milwaukie impact gun"
Magento's default relevance model was built for none of these. Every failed query was a potential lost sale - and a potential support call.
No analytics
There was no visibility into which queries were failing, which product codes generated demand but returned nothing, or which brands were most searched. With a catalogue of this size, operating without search analytics means problems compound invisibly over time.
The Solution
Full attribute-level indexing across 100,000+ products
Every specification, category, brand, and custom field is indexed - not just product titles. A search for "18V brushless orbital sander variable speed" resolves through attribute matching across the entire catalogue, correctly identifying products that match all of those parameters simultaneously.
Professional query handling
BRADsearch processes professional buyer queries in the way they're actually typed:
- Partial product code matching: "DDF48" resolves to DDF484 and its variants
- Typo tolerance for technical terms: misspelled model numbers and brand names still return the correct product
- Synonym matching: impact gun, impact wrench, impact driver - all correct
- Multi-attribute query processing: structured queries parsed as attribute filters, not as keyword lists
Analytics dashboard
Ongoing visibility into brand search demand, zero-result queries, and which model numbers are searched without matching products. The analytics dashboard turns search behaviour into catalogue and assortment intelligence - a continuous feedback loop for buying decisions.
The Outcome
When the primary store's professional buyers can find what they need, the case for the whole portfolio is straightforward. Additional stores in the same group are now onboarding to BRADsearch.
Professional buyers find products accurately by partial code
DDF484, HR2631, any partial string - resolves to the correct product. The search bar now behaves like a professional tool catalogue should, matching the precision of the buyers using it.
Longer technical queries handled reliably
Multi-attribute professional searches now return correctly ranked results. A buyer searching "Makita 18V brushless LXT impact driver" gets the right product family, ranked by relevance, not by keyword overlap.
Analytics drive catalogue and assortment decisions
Teams see which brands are most searched, which codes return zero results, and which product specifications generate demand that the current catalogue doesn't satisfy. This is a fundamentally different mode of catalogue management - demand-led rather than assumption-led.
CS load reduced
Professional buyers who previously called support to locate a product now find it through search. The reduction in search-related support queries is a direct operational benefit of getting search accuracy right at the point of intent.
Portfolio-wide rollout in progress
Additional stores in the same group are being onboarded to BRADsearch. The decision wasn't difficult once the primary store demonstrated that professional buyers with complex search behaviours could be served accurately at scale.
← Back to all case studies