Improving consultation booking conversion
Booking a consultation is the journey that matters most on the platform. A user passes through several stages before completing one, so every unclear decision point is somewhere they can quietly drop out.
- Product
- TalkndHeal
- Role
- Product Manager
- Problem
- Booking friction
- Outcome
- +18% conversion
Where are users losing confidence or motivation to complete their booking?
The question that framed the whole piece of workThe journey, stage by stage
Select a stage to see what the user is trying to do there, and what gets in the way.
What I did
- Identified user problems and points of friction across the booking journey.
- Analysed user feedback, journey behaviour and internal stakeholder insight to separate noise from pattern.
- Converted those observations into prioritised product opportunities and written requirements.
- Worked with design and engineering from definition through to launch.
- Supported post-launch measurement of the resulting experience.
Prioritising what was worth building
I scored each opportunity on user impact, business impact and development effort. Hover or focus a row to locate it on the plot.
| Opportunity | User impact | Effort | Priority |
|---|---|---|---|
| Simplify key booking steps | High | Medium | P0 |
| Improve information clarity | High | Low | P0 |
| Improve decision-making flow | High | Medium | P0 |
| Additional engagement enhancements | Medium | Medium | P1 |
Three opportunities cleared the bar on impact. The engagement work was real, but it would not have moved the conversion problem on its own.
How I read that number
The three P0 changes shipped as one release. That makes the +18% the effect of the bundle, not of any single change, and I would not defend a per-change attribution from it.
It is a before-and-after comparison against the pre-launch baseline, not a controlled test. Seasonality and the acquisition activity running at the same time are not separated out. I am confident the direction is real and clear that the precision is softer than the number looks.
Given the volume the platform had, I would now stage the releases so each change gets its own read, and hold back a portion of traffic where the sample supports it.
What I took from it
- Clarity drives decisions. People need the right information at the moment they are deciding, not earlier and not buried.
- Friction compounds. Each extra step is another chance for someone to leave, and the cost is multiplicative rather than additive.
- Hierarchy is a product decision. How information is ordered changes how fast someone understands what to do next.
- The journey has to feel like one thing. Discovery, evaluation and booking should not read as three separate products.
Conversion problems are not always solved by adding features. Often the biggest gain comes from removing friction and making the next action obvious.
What I would do differently
Shipping the three changes together made the outcome easy to communicate and impossible to decompose. Sequencing them would have cost a few weeks and told me which one actually mattered.
I would agree the drop-off definition for each stage before the work starts rather than during measurement. Settling what counts as an abandoned booking up front removes an argument later, when everyone has a stake in the answer.
I would also keep a written record of the opportunities we chose not to build, with the reason attached to each. Six months on, that record is more useful than the roadmap.