← Use cases
Conversion

Searches for milk are up. Orders for milk are down. What happened?

Bicycle watches search conversion keyword by keyword and location by location, and names why a drop happened: availability, price or the catalog.

Book a demo Start for free

No credit card required

Before

An overall funnel dip says something is wrong but not where. The cause usually sits in a few keywords in a few locations, where the products people want are missing or priced out.

With Bicycle

Drops are detected at the keyword and location level against their own baseline, with the products and stores where they sit and the reason.

What happens when a KPI moves.

  1. 01

    Detect

    Conversion for a keyword drops in one location against its baseline.

  2. 02

    Explain

    The cause is named: the top products for that keyword are out of stock, priced above competitors, or switched off in the catalog.

  3. 03

    Act

    Bicycle recommends the next step: replenishment restocks, category reactivates a product or adjusts price.

  4. 04

    Learn

    Keywords that keep dropping expose assortment gaps or uncompetitive prices.

80%+

of conversions on each keyword come from its top 10 products

Seen at: a leading online grocer.

Everyone can ask. Data & Analytics controls what gets trusted, published, and acted on.

Runs on
  • Search logs
  • Add-to-basket events
  • Availability
  • Prices
  • Catalog
Used by
  • Category
  • Merchandising
  • Product
  • Replenishment

Common questions

How do you find the cause of a drop in search conversion?
For each affected keyword, Bicycle identifies the reference products that drive most add-to-basket events, then checks each one against availability, catalog status, price movement and the competitive price index, and assortment. Bicycle also tests for a search volume spike the catalog was not ready for. The output is a ranked attribution that names the dominant cause, not a list of metrics to investigate.
How do you tell a real conversion drop from normal variation?
Bicycle runs statistical anomaly detection against each keyword's historical baseline, with the threshold tuned to that keyword's volume tier. Bicycle looks for two kinds of drop: over time, such as today against last week, and across places, such as one city against the national average or one store against its city. Your search funnel and location and category hierarchies are already modeled.
Who acts on a search conversion drop, and who sees what?
Bicycle sends each alert to the team responsible for the dominant cause: replenishment for availability, category for catalog and merchandising for pricing. Merchandising reviews the competitive price index and decides whether to match. Access is scoped by role, so head-office teams see the full national view and regional teams see their own region on the same data model.

More on conversion

See it on your own KPI.

Bring one revenue-critical KPI. We'll show how Bicycle detects the movement, explains the cause, and recommends the next step.

Book a demo