Run one when growth slows and the leadership team disagrees about why. It also helps before a large change, such as a new sales leader, a new market or a new system.
The example plan below assumes one product line and up to about 30 people in customer-facing roles. Treat the two weeks and the team size as planning assumptions. Allow more time for larger teams, extra regions or slow access to reports. You need a sponsor from the leadership team, access to your reports and time with 8 to 12 people.
If you already know the problem sits in one handover, a narrower review is faster. The guide Fix the process before you change the software covers that approach.
Agree the question and the scope
Write down one question the diagnostic must answer. For example: "Why has new revenue fallen over the last two quarters?"
Agree the period to review, such as the last two to four quarters. Agree which teams are in scope and who receives the findings.
Ask the sponsor to tell the teams what is happening and why. Explain that the review looks at how work moves between teams, and that findings will be reported by role and step.
In a small team, a role can identify a person. Tell each person this before the interview, and agree how their comments will be reported.
Pull the funnel numbers
Build one table that shows each stage from first enquiry to renewal. Show the count at each stage and the conversion rate from the stage before.
Run these as summary reports inside your CRM or reporting tool. Avoid exporting customer records to personal files or devices.
Use cohort-based conversion rates
A cohort is a group of leads or opportunities that entered a stage in the same period. A cohort rate follows that group through the later stages.
Counts from the same quarter can relate to different leads. Some opportunities created in Quarter 2 came from meetings held in Quarter 1. Dividing one quarter's count by another gives only a rough ratio. Where your reports allow it, measure each rate on a cohort.
All figures in this table are illustrative. Each rate assumes the row follows the same cohort as the row above.
| Stage | Quarter 1 | Quarter 2 | Change |
|---|---|---|---|
| New leads | 1,200 | 1,250 | Up 4% |
| First meetings held | 240 (20%) | 250 (20%) | Steady |
| Qualified opportunities | 96 (40%) | 75 (30%) | Down 10 points |
| Deals won | 24 (25%) | 18 (24%) | Down 1 point |
| Customer renewal rate, of customers due to renew | 90% | 82% | Down 8 points |
In this example, lead volume rose slightly and the win rate held close to its earlier level. Qualification and renewal rates fell. Check deal values, timing and customer revenue before you estimate their effect on revenue.
To explain the fall in revenue, add average deal value, the timing of won deals and revenue by renewing customer.
Separate renewal from revenue retention
Customer renewal rate counts customers. Revenue retention measures the revenue you keep from existing customers. Report these separately:
- Renewal rate: customers who renewed, divided by customers due to renew.
- Contraction: revenue lost when renewing customers reduce their spend.
- Expansion: revenue gained when existing customers increase their spend.
- Revenue retention: revenue from existing customers at the end of the period, divided by their revenue at the start.
A business can renew 90% of its customers and still lose revenue. This happens when the customers who leave or reduce spend are the larger ones.
Interview the people who do the work
Book 45-minute interviews with 8 to 12 people. Cover each role in scope, such as marketing, sales development, account managers, customer success, sales operations and finance.
Ask everyone the same core questions, so you can compare the answers. Add one or two questions about the stages where the funnel drops.
| Question | What it shows |
|---|---|
| Walk me through the last deal or campaign you worked on. | How work moves in practice |
| Where does your work wait for someone else? | Handovers that slow things down |
| Which report do you trust least, and why? | Data that people have stopped using |
| What do you track outside the main system? | Workarounds and side lists |
| What changed in the last six months? | Possible causes of a drop in the funnel |
| If you could change one thing, what would it be? | What the team sees as the priority |
Keep notes in one shared folder with restricted access. Leave out customer names. Report comments as themes where a role could identify the speaker. If you want to record an interview, ask the person first.
Review real deals and accounts
The funnel highlights changes to investigate. Record reviews and interviews help you test possible causes.
Choose a deliberate sample of about 20 records from the stages where the funnel dropped. For example, pick five won deals and five that were lost or stalled. Add five customers who renewed and five who left or reduced their spend.
This is a qualitative sample, chosen to show reasons and patterns. It cannot estimate win rates or churn across the business. Use the funnel reports for rates.
Read the notes, emails and activity history for each one inside your system. Note the lead source, the time spent in each stage and the reason it was won or lost. If you review recorded sales calls, use only calls where the customer was told about the recording.
Score and rank the findings
Link each finding to its evidence. Then score it, so the leadership team can agree what to fix first.
Copy this table. Score impact, confidence and effort from 1 (low) to 5 (high). Note any finding that depends on another. The rows are examples.
| Finding | Evidence | Impact | Confidence | Effort | Depends on | Owner |
|---|---|---|---|---|---|---|
| Fewer first meetings become qualified opportunities | Cohort funnel report, 6 of 10 interviews | 5 | 4 | 2 | Agreed qualification criteria | Sales manager |
| Renewal conversations start in the final month of the contract | 4 of 5 lost customers reviewed | 4 | 3 | 1 | None | Customer success lead |
| Lead source is blank on many records | Summary report, 3 interviews | 2 | 4 | 2 | None | Sales operations |
Use the four measures together:
- Impact: how much of the drop in the funnel the finding could explain.
- Confidence: how strong the evidence is. A finding that only one person raised needs more evidence before you act on it.
- Dependencies: whether other changes rely on it. For example, agree qualification criteria before you change how leads are scored.
- Effort: the time and cost to make the change.
Start with findings that score high on impact and confidence. Move a finding earlier if others depend on it. Use effort to choose between findings that score about the same. Choose the top three to act on, and keep the rest on a list for the next review.
Follow an example schedule
Adjust this schedule to the number of interviews and how quickly you can get reports.
- Days 1 and 2: Agree the question, scope and sponsor, and send the interview invitations. Build the funnel table and mark the largest drops.
- Days 3 to 5: Run the interviews, about five each day. Review the sample of deals and accounts.
- Days 6 and 7: Write up and score the findings. Book short follow-up calls to fill any gaps.
- Day 8: Test the top findings with each team lead. Ask what you have missed.
- Days 9 and 10: Give each of the three changes an owner, a measure and a target date. Present the plan, and set a date to measure the funnel again, for example after one sales cycle.
Avoid these common mistakes
- Starting with a solution in mind. If the sponsor already wants a new system or a new hire, the review turns into a case for it. Agree the question before anyone proposes an answer.
- Ending with a long report. Forty recommendations give the team nowhere to start. End with three changes, each with an owner.