All products

Analyse customer journey drop-off with AI

Compare aggregate entry and completion counts across a funnel. Review the number lost, conversion rate and volume-weighted priorities before choosing an experiment.

Analyse customer journey drop-off with AI when a product team wants to decide which part of a funnel deserves investigation. The customer-journey-gap-analyzer skill reads ordered stage counts and calculates conversion, drop-off and a volume-weighted impact score. It also compares the first stage’s entries with completions at the last stage.

A large percentage loss from a tiny starting group may matter less than a smaller loss affecting hundreds of attempts. Check that the rows represent the same population and time period. The suggested experiments are templates based on stage names, not discoveries about an individual customer’s behaviour.

What it does

  • Compare entries and completions at each stage.
  • Keep absolute loss alongside percentage drop-off.
  • Rank up to five gaps for further investigation.
  • Review the end-to-end conversion for the ordered stages.
  • Prepare experiment ideas to validate with customer research.

Install and use

  1. Install the skill and provide Python 3 and an authorised CSV containing each stage, entered count and completed count in journey order. Individual customer identifiers are unnecessary.
  2. Ask, “Rank drop-offs from registration to checkout. Show both number and percentage, and label suggested experiments as hypotheses.”
  3. Check for impossible counts and inconsistent cohorts before using the priorities in a planning meeting.

FAQ

Source

Made by sisodiabhumca. Original skill: customer-journey-gap-analyzer, published on GitHub under the MIT licence.

View on GitHub