How to Run a Weekly CRO Review in 30 Minutes
Conversion optimisation works as a habit, not a project. Use this 30-minute weekly agenda to check your funnel, watch the right sessions, pick one fix, and verify last week's change.
Short answer: spend five minutes on numbers, ten on recordings, five on errors and speed, five choosing one fix, and five checking whether last week's fix worked. One small, evidence-based improvement every week adds up faster than a quarterly redesign.
Table of Contents
- Why Weekly
- Before the Meeting
- The 30-Minute Agenda
- The One-Page Log
- Who Should Attend
- How to Keep It Going
- Key Takeaways
Why Weekly
- Problems are found sooner. A broken coupon field costs a week of sales, not a quarter.
- Fixes are small. Small changes ship quickly and are easy to attribute.
- Learning compounds. Fifty small improvements a year is realistic.
- It survives busy periods. Thirty minutes is hard to cancel.
Before the Meeting
One person spends fifteen minutes preparing:
- Last week's funnel numbers by device
- A saved filter of sessions that dropped at the weakest step
- The top new JavaScript error groups
- The status of last week's fix
Do not prepare slides. Open the tool on the screen.
The 30-Minute Agenda
Minutes 0 to 5: The numbers
Look at one funnel for your primary goal.
- Step conversion this week against last week and the four-week average
- Split by device
- Any step that moved more than normal variation
Decide which step to look at. Only one.
Minutes 5 to 15: The sessions
Open the sessions that stopped at that step. In DeepSync, clicking the step's drop-off number takes you straight there.
- Watch five or six together, using Highlights to jump to the moments that matter
- Say out loud what you see, not what you think it means
- Note every repeated behavior
This is the part that changes minds. A team that watches a customer fail to check out stops debating priorities.
Minutes 15 to 20: Errors and speed
- New JavaScript error groups on conversion pages, ranked by affected sessions
- Any change in real-user loading or responsiveness on those pages
- Rage clicks on conversion pages
Minutes 20 to 25: Choose one fix
Score candidates quickly:
| Candidate | Seen in | On conversion path? | Effort |
|---|---|---|---|
| Submit hidden by keyboard on mobile | 4 of 6 sessions | Yes | Small |
| Pricing table confusing | 2 of 6 | Yes | Medium |
| Slow hero image | Performance data | Partly | Small |
Pick the one with the most evidence and the least effort. Assign an owner and a date. Bugs do not need a test. Fix them.
Minutes 25 to 30: Verify last week
- Did the fix ship?
- Did the step's conversion change for the affected segment?
- Watch two sessions through the fixed flow to confirm the behavior is gone
On Business and Enterprise plans, Pulse keeps watching the metric behind a finding after you mark it fixed and reports whether it moved.
One fix per week
Choosing three means shipping none. The discipline of choosing one is what makes the habit work.
The One-Page Log
Keep a running document. One row per week.
| Week | Step examined | Finding | Evidence | Fix | Owner | Result |
|---|---|---|---|---|---|---|
| 1 | Cart → checkout, mobile | Shipping cost appears late | 5 of 6 sessions | Show estimate on cart | Name | Step up for mobile |
| 2 | Signup form | Phone format rejected | 4 of 6 sessions | Accept spaces | Name | Pending |
After three months this log is the best record of what you have learned about your customers.
Who Should Attend
Keep it to four or five people.
- Someone who owns the product or site
- A designer
- An engineer who can say what a fix costs
- A marketer who knows what traffic is arriving
- Occasionally, someone from support
Rotate who prepares. It spreads the skill.
How to Keep It Going
- Same time every week, early in the week.
- Never skip two in a row.
- Stop at thirty minutes.
- No slides.
- Celebrate verified results, not shipped changes.
- When numbers are flat, look at a different flow rather than cancelling.
Key Takeaways
- Five parts: numbers, sessions, errors, one fix, verification.
- Examine one funnel step per week.
- Watch sessions together and describe what you see.
- Keep a one-page log of findings and results.
Conclusion
Conversion optimisation rarely fails for lack of ideas. It fails for lack of rhythm. Thirty minutes a week is enough to build one.
Frequently Asked Questions
Related articles
Stay in the loop
Get the latest insights on product analytics and user behavior delivered to your inbox.



