Skip to content
Book a demo
For Constructor teams

Get more from your Constructor. No second migration.

Constructor optimizes for revenue with behavioral ranking. searchHub adds independent, auditable query understanding and attribution on top, so you can prove and steer relevance without adopting another full platform.

Book a demo See how it works
How Constructor works, and where it needs help

Revenue ranking, in one box.

Constructor ranks for revenue using behavioral signals, but adopting it is a full search replacement, and the same system that ranks your results also reports on how well it did.

The complement
What searchHub adds on top of Constructor

searchHub stays engine-agnostic and gives you independent, auditable query understanding and revenue attribution. It is a complement that keeps you in control, not a second migration.

It optimizes Constructor. It does not replace it.
What searchHub adds to Constructor

Three capabilities added in front of Constructor, not instead of it.

Intent shaping as-you-type

Suggest orders autocomplete by what converts, typo-free and duplicate-free.

Query rewriting

Every spelling and typo is rewritten to one proven query, so Constructor retrieves on clean input.

Prioritized insights

Insights ranked by conversion potential tell you which journeys to fix first, so you act on evidence, not vanity charts.

Engine-agnostic by design — the one constant in your discovery stack, through every engine change and re-platforming. Why searchHub →
Topics right now

What Constructor teams are asking right now

Drawn from the current Constructor news cycle — and what each topic means for your search KPIs.

Personalization needs dense signals

Clickstream learning starves on long-tail queries. Clustering pools variant traffic so ranking learns from whole intents, not fragments.

An A/B-test culture

Constructor teams prove everything on live traffic. searchHub is proven the same way — against your own baseline, in weeks.

Attractiveness needs clear intent

Ranking by shopper preference works once intent is unambiguous. Master queries remove the ambiguity first.

weeks
to a live A/B test, not a 12-month migration
0
changes to your Constructor setup

What would you rather add: independent proof, or another platform to migrate to?

Book a demo