-
1
Scope the quarter and pull a representative sample of 20 to 50 won deals
Start by defining the cohort cleanly. Pull every closed-won opportunity from the last full quarter, then filter down to a representative sample of 20 to 50 deals. Below 20 the themes are anecdotal and will not survive challenge from a skeptical exec. Above 50 the marginal deal rarely changes the ranked theme list and you burn analyst time. Stratify the sample so it reflects the shape of the quarter rather than the loudest wins. Pull across segment (SMB, mid-market, enterprise), across product line, across new-logo versus expansion, and across your top three to five named competitors. If the quarter was skewed heavily toward one segment, mirror that skew in the sample rather than forcing an even split.
- Export every closed-won opportunity from the quarter with amount, segment, product, source, and primary competitor fields.
- Target 20 to 50 deals with a stratified cut across segment, product line, and new-logo versus expansion.
- Oversample strategic wins (named accounts, flagship logos) but tag them separately so they do not swamp the aggregate.
- Freeze the sample list and share it with Product Marketing and Enablement before coding begins, so nobody can cherry-pick later.
Tip: If the sample looks nothing like the quarter (for example 80 percent enterprise when the quarter was 60 percent mid-market), rebuild it. A report that misrepresents the cohort will get torn apart in the readout.
-
2
Blend sources so no single artifact drives the narrative
Any single source lies in a predictable way. CRM close-reason fields are rep-authored and over-index on features and price. Call recordings capture buyer language but miss the executive moments that happen offline. Deal retros capture the rep's narrative but not the buyer's. The report earns its credibility by triangulating across at least three sources per deal: the CRM record and close reason, call transcripts or recordings covering discovery and the technical evaluation, and a short 15-minute debrief with the rep or the AE pair. For strategic deals, add a 20-minute call with the economic buyer or champion. Mixing sources also lets you catch contradictions, which are themselves high-signal moments.
- Pull the CRM record, close-won reason, and competitor field for every sampled deal.
- Pull discovery and demo call transcripts from Gong, Chorus, or your call-recording tool.
- Run a 15-minute structured debrief with the AE or AE pair on every sampled deal using the same four questions.
- For the top five strategic deals in the sample, add a 20-minute champion or economic-buyer conversation.
Tip: When the rep's story and the call transcript disagree, trust the transcript and flag the gap. Those contradictions often expose coaching opportunities worth more than the theme report itself.
-
3
Code every deal against a shared taxonomy of 8 to 15 themes
Themes compound only when the taxonomy stays stable across quarters. Agree a shared code list of 8 to 15 themes with Product Marketing and Enablement before coding starts. A workable baseline covers pain-point themes (process friction, data fragmentation, compliance pressure, headcount constraint, revenue leakage), winning-capability themes (specific feature wins, integration coverage, implementation speed, pricing model fit, services depth), and context themes (incumbent displacement, greenfield build, consolidation play, trigger event). Code each deal against three to six themes maximum. If you need more than six to describe a deal, your taxonomy is too narrow or the deal is a genuine outlier that deserves a sidebar in the report.
- Lock the taxonomy with Product Marketing and Enablement before coding. Add codes only at quarter boundaries.
- Code every sampled deal against three to six themes drawn from the shared list.
- Record a verbatim buyer quote (anonymized if needed) for each theme coded against a deal. Quotes carry the report.
- Flag any deal that does not fit the taxonomy as a candidate outlier and resist the urge to force-fit it.
Tip: If the taxonomy grows past 15 codes you lose the ability to compare quarters. Collapse near-duplicates (process friction and workflow pain are the same theme) before you commit the code list.
-
4
Segment the themes by deal size, segment, and competitor
An aggregate ranked list hides most of the useful signal. The same theme can be the top driver in enterprise and barely register in SMB, or it can show up only in deals against one named competitor. Cut the themes at least three ways: by deal size band (for example sub-$25k ACV, $25k to $100k, over $100k ACV, keeping specific thresholds in the appendix rather than the headline), by segment or ICP tier, and by primary competitor. Look specifically for themes that invert across cuts, where a top-three driver for enterprise is bottom-three for SMB. Those inversions are where positioning and packaging decisions actually get made.
- Build a theme-by-cut matrix with rows for themes and columns for each cut (segment, size band, competitor).
- Rank themes within each cut and highlight the top three per cut.
- Flag themes that invert between cuts for a dedicated section in the report.
- Call out themes that show up in only one segment but represent a disproportionate share of ACV in that segment.
Tip: If a theme is top-ranked in the aggregate but not top-three in any single cut, it is probably a coding artifact rather than a real driver. Pressure-test it before you headline it.
-
5
Identify the capabilities that closed the deal versus the capabilities that were merely present
Buyers mention many features during an evaluation. Only a few actually tip the decision. Separate the two by looking at the critical moments in each deal: the moment the shortlist got cut, the moment the champion committed, the moment procurement greenlit the price. Trace back which capability or proof point the buyer referenced at each moment. That is the closing capability. Capabilities that showed up in demos but never got mentioned at a decision moment are the present-but-not-closing set. The report should headline the closing capabilities, because that is what Product Marketing turns into messaging and Enablement turns into demo flow, and should relegate the present-but-not-closing set to an appendix so Product does not over-invest.
- For each deal, identify the two or three moments where direction shifted (shortlist cut, champion commit, procurement signoff).
- Note which capability or proof point the buyer cited at each moment, in their own words.
- Aggregate across the sample to produce a ranked list of closing capabilities.
- Separately list capabilities that showed up in evaluation but did not drive decisions, as context for Product.
Tip: If the closing capabilities list reads identical to your current feature marketing, you are either doing a great job or you are coding what you expect to see. Have Enablement spot-check 10 deals to catch confirmation bias.
-
6
Map the competitors beaten and extract the plays that worked against each
Competitive wins are the highest-value cohort inside the report because they tell Sales and Enablement exactly how to run the next deal. For every sampled deal where a named competitor made the shortlist, record which competitor, how far they got in the evaluation, and what specifically moved the buyer off them. Group the findings by competitor so each named rival has its own one-pager inside the report: three to five recurring plays that worked, two to three objections the competitor raised that reps need a response for, and two to three verbatim buyer quotes about the competitor from the sampled calls. This section feeds the next battlecard refresh directly.
- Tag every deal with the competitors that made the shortlist and how far each got.
- Record the specific moment or capability that moved the buyer off each competitor.
- Build a one-pager per competitor with winning plays, their objections, and anonymized buyer quotes.
- Hand the competitor pages to Enablement as the direct input for the next battlecard refresh.
Tip: Resist writing competitor pages for every rival mentioned. Build pages only for competitors that appeared in five or more sampled deals, so each page sits on real signal rather than one anecdote.
-
7
Write the report in a layered format and keep the headline to one page
The audience for the report is busy. Write it in layers so each reader can go as deep as they need. Layer one is a one-page executive headline with the top five themes, the top three closing capabilities, the top two competitor stories, and three recommended actions with named owners. Layer two is a 6 to 10 page body with each theme covered on its own page, including the cut matrix, two or three anonymized buyer quotes, and the specific sample deals that support it. Layer three is an appendix with the full sample list, the taxonomy, the methodology, and the raw cut matrix. Keep the one-page headline readable in 90 seconds. The exec team will read nothing else, and that is the page that actually drives action.
- Draft the one-page executive headline first. If it does not fit on one page the themes are not ranked tightly enough.
- Expand each top theme into its own body page with cut matrix, quotes, and supporting deals.
- Put the sample list, taxonomy, and methodology in an appendix so the body stays readable.
- Share a draft with Product Marketing and Enablement before the readout so the recommendations land without surprise.
Tip: Open the body of each theme page with a single anonymized buyer quote in the buyer's own words. Nothing drives action faster than the exec team reading the pain in a real customer sentence.
-
8
Land the findings in positioning, enablement, and roadmap with named owners
A report that nobody acts on is wasted budget. Close the loop by assigning every recommended action to a single owner with a date. Product Marketing owns messaging updates, website refreshes, and launch narrative changes driven by the themes. Enablement owns the next battlecard refresh, the discovery-question updates, and the demo-flow changes based on the closing capabilities. Product owns the capability-gap list that falls out of the deals where you won despite rather than because of a feature. Open every quarterly readout with a 10-minute status check on last quarter's commitments. After three to four cycles the win themes report becomes the single most trusted source of what is actually working in the market, and the taxonomy is stable enough to show real quarter-over-quarter trend lines.
- Translate the top five themes into three to five concrete actions with named owners in PMM, Enablement, and Product.
- Set a target date for the first visible change per action (new battlecard, revised demo flow, updated homepage headline).
- Open the next quarterly readout with a status check on the prior commitments before you present the new report.
- Publish the one-page headline to the whole go-to-market team, not just leadership, so reps see the signal that drove the changes they feel.
Tip: Pair the report readout with a 10-minute Enablement session where a rep walks through one of the sampled wins live. Hearing a real deal story alongside the aggregated themes anchors the data in a way that charts alone never do.
