Choose how to group your proposals
Pick one way to cut your proposals first. The options are by offer (audit, project, retainer), by size band, by lead source (referral, inbound, outbound) or by format (single price or tiers, live-presented or emailed). Start with the cut you can actually change. Keep it to five groups at most, or the numbers get too thin.
Build the table
Export the last 12 months of proposals from the CRM, such as HubSpot or Pipedrive, into Google Sheets. One row per proposal with these columns: date sent, type, source, value, number of options, pages, days to decision and outcome.
Use three outcomes: won, lost and no decision. Treat any proposal with no answer after 45 days as lost. Fill the gaps by hand. It takes an hour and it is the most useful hour of the exercise.
Calculate four numbers for each type
- Win rate: proposals won divided by proposals sent.
- Average value of the won ones.
- Median days to decision.
- Revenue per proposal sent: win rate multiplied by average value.
The last one is the number that ranks the types. A type with a 25 per cent win rate and large deals can beat one with 50 per cent and small deals. Here is an illustration with made-up numbers:
| Type | Sent | Win rate | Average won | Revenue per proposal | |---|---|---|---|---| | Fixed-price project | 24 | 25% | 18,000 | 4,500 | | Monthly retainer | 30 | 50% | 6,000 | 3,000 |
Check the sample size
Fewer than 10 proposals in a group is a hint, not a result. Around 20 is a reasonable signal. At those sizes, a gap of less than about 10 points between two groups is probably noise. Say so when you share the numbers, and wait another quarter for thin groups.
Look for the cause
A pattern is not yet a reason. Referral proposals might win more because of trust, not because of the format. Compare inside the same source before you conclude that the format works. Then read five lost proposals of your weakest type and look for what they share, for example a missing price option or a long scope section. Pair this with Analyse lost deal reasons.
Decide one action per type
Give every type one of three verdicts:
- Push: send more of these and qualify for them earlier.
- Fix: change one element, then test it on the next 10 proposals using Test pricing presentation formats.
- Stop: stop sending, raise the price or qualify out sooner.
Write the verdicts and the date you will recheck them. Live reading data, such as which sections prospects open, sits in How to track proposal performance.
Common mistakes
- Looking only at the overall win rate.
- Ranking by win rate alone and ignoring deal value.
- Counting stale proposals as open forever.
- Drawing conclusions from 6 proposals.
- Slicing into so many groups that each has two or three proposals.
- Changing three things at once, so you cannot tell what worked.
How you know it works
You can say in one sentence which proposal type earns the most per proposal sent, and why. Your next quarter's proposal mix has moved towards it. Revenue per proposal sent has gone up, even if your overall win rate has barely changed.
The review repeats each quarter and takes under an hour, because the sheet already exists. On Monday morning, export the proposals and add the four columns.
