The end-of-year win-themes audit
Five prompts for retiring win themes that have lost discriminatory power, and promoting the ones that actually showed up in wins. A year-end ritual that takes an afternoon and pays out across the next twelve months of bids.
The win themes you used in January are not the win themes you should be using in January a year later. Some held up. Some quietly stopped differentiating you. A few were never discriminating in the first place and survived because nobody audited them.
The end-of-year audit is the fix. One afternoon, your last forty to sixty proposals open on one screen, the win/loss record open on another, and five prompts run against every theme the team used more than twice.
Why themes drift
Themes drift because the market drifts. A differentiator in a 2025 security questionnaire — “we encrypt at rest and in transit with customer-managed keys” — reads as table stakes in late 2025. Evaluators who saw it as a proof point last year will skim over it this year. If you keep leading with it, you spend the evaluator’s attention on something that does not move them.
Themes also drift because your capability set changes. You shipped features; you hired into new verticals; you lost a reference account. The company that shows up on paper in December is not the company that showed up in January. The themes should mirror that drift, or they start to sound like a press release from the previous version of the business.
PropLibrary’s swap test remains the most honest filter: if you can swap your company name with a competitor’s and the theme still reads the same, it was not a theme, it was a brochure line. We wrote about applying the swap test during drafting in the field guide. The audit applies it again, on the full year of usage, with the win/loss record as the arbiter.
The five prompts
Open a spreadsheet. Columns: theme, bids used in, wins, losses, withdrawals, and five verdict columns — one per prompt. Fill it in from the post-mortem records.
Prompt 1 — Did the theme appear in the win, or only in the draft? Some themes get into the executive summary and then vanish by the technical sections. Evaluators in those sections never saw the theme. A theme that only lives in the first two pages of a 60-page response is a theme the proposal function is telling itself, not the buyer.
Prompt 2 — Did the buyer ever echo the theme back? Debrief notes are the cleanest signal. When a buyer’s selection letter, award memo, or debrief call quotes your language — “we chose you because of your emphasis on migration risk reduction” — that theme earned its place. When no buyer, in any debrief, referenced the theme over a year of use, that theme was scaffolding inside your head, not signal inside theirs.
Prompt 3 — Did losing bids use the same theme? If the theme correlates with wins and anti-correlates with losses, keep it. If it correlates with losses, or shows up equally in both, it is not doing discriminatory work. The theme “our platform scales” ran in 14 wins and 11 losses in one team’s 2025 record. The theme “we pre-build the migration runbook before contract signature” ran in 9 wins and 1 loss. Only one of those is a theme.
Prompt 4 — Does the theme still match the product? Themes anchored in a capability that has atrophied, a partnership that has lapsed, or a reference account that has churned are liabilities, not assets. An evaluator who validates the claim and finds it half-true reads the whole response as half-true. Audit every theme against the current state of the company, not the state of the company at the time the theme was written.
Prompt 5 — Would a new proposal writer, handed the theme with no context, understand why it wins? Themes that require a paragraph of explanation before they make sense to a new team member are themes that require the same paragraph of explanation before they make sense to a new evaluator. Evaluators do not get that paragraph. Retire the theme or rewrite it into a sentence that stands alone.
Verdicts and what to do with them
For each theme, the audit produces one of four verdicts.
Promote. The theme won, buyers echoed it, losses did not use it, it still matches the product, and a new writer can understand it. Move it into the house playbook. Make it the default for the vertical where it performed. Build evidence blocks around it so the next drafter has citations ready.
Keep on probation. The theme has mixed signal. Keep using it but tag it, and at the six-month mark do the audit again with the new data. The tag is a reminder that the theme is provisional — drafters should not lean on it without checking the bid’s specifics.
Rewrite. The theme is close but generic, or was right two years ago and needs an update. Do not retire it; rewrite it into something that clears the swap test. “We have an experienced migration team” becomes “we pre-build the migration runbook in capture, before contract signature, and our last 14 migrations ran with zero unplanned downtime.”
Retire. The theme was never discriminating, or stopped being so, or is tied to a capability the company no longer leads with. Archive it in a retired-themes file so the team remembers why it went away — that file is almost as useful as the active list, because it stops somebody two years later from reviving a theme that already failed.
Where we keep the artifact
The audit is not a one-off memo. It lives in the knowledge base next to the themes themselves. Every theme has a version history: when it was introduced, what bids it ran in, what the audit said, what the rewrite looked like, when it was retired. A proposal writer opening a new response can see the theme’s record before they use it.
That history is the compounding piece. Without it, the audit is an afternoon of work that evaporates by February. With it, every audit builds on the last one, and after three years the team has a real body of evidence about which themes discriminate in which verticals against which competitors.
The preview of the compounding pillar last week argued that proposal software earns its place only when the system knows more after the tenth bid than it did after the first. The themes audit is one of the mechanisms that makes that claim true. Without the write-back, themes are just slogans the team grew tired of.
One more prompt, optional
If the team has capacity: show the winners anonymously to five SMEs who did not work on the bid. Ask them which theme made the bid land. If the SMEs pick a theme the proposal function did not think was the winner, that is a signal. The function’s theory of why a bid won is sometimes the hindsight narrative, not the thing that actually moved the evaluator. A sixty-minute round of internal blind review corrects more of those errors than a year of post-mortems that only include the authors.