The case-study harvest: one wrap, three marketing pieces, thirty minutes

The engagement wrapped three months ago. The work was good. The client said it was good. There's a thank-you email in your inbox that's specific enough to use as a case-study quote. And yet: no case study exists. The reason is the reason almost every solo consultant has missed almost every case study they could have written. You wrapped the engagement on a Friday, started a new client on Monday, and by the time you remembered the wrap, the texture was gone.

Texture decays. The specific moment the client said the thing the prospect would have responded to, the specific decision that made the engagement turn the corner, the specific objection you handled that another prospect will have — those are warm for about two weeks after wrap and substantially colder by month two. By month three they're gone. You remember the deliverable. You don't remember the moment.

The fix is a 30-minute procedure that runs inside two weeks of wrap and produces one case study plus three derivative social pieces from the same source. Four marketing assets, single sitting, single source pull. This article is the procedure.

Why most case studies never get written (and why the ones that do are weak)

Two failure modes account for ~90% of missing case studies on solo consultants' sites.

The procrastination spiral. The consultant tells themselves they'll write the case study "soon." Three weeks pass. The next engagement is in flight. Writing the case study now requires reconstructing context that's no longer fresh. The reconstruction cost feels high enough that the consultant defers again. Six months later, the case study doesn't get written. The next prospect lands on the site and sees three case studies from 18 months ago and concludes the consultant has been quiet.

The wall-of-text trap. When the consultant does write the case study, they write it from memory with no source material, which means the case study is generic. "Helped a B2B SaaS company improve their conversion rate by working closely with their team to identify key friction points and implement targeted solutions." That sentence has zero texture. It could describe any consultant working with any client on any problem. It produces no trust, no specificity, no signal that the consultant actually did the work.

Both failure modes resolve when the case study is harvested from existing artifacts (weekly updates, meeting transcripts, the project retrospective) within two weeks of wrap. The texture is preserved. The reconstruction cost is low because the inputs already exist.

The four-piece harvest

The output of the procedure is four pieces of content from the same engagement:

1. The case study. 600-900 words. Hosted on your site. Structured around the non-numeric outcome (the specific moment the work paid off in a way that's harder to fake than a percentage). Cross-referenced for prospects who want to see what working with you looks like.

2. The long-form social post. 300-400 words. LinkedIn, your newsletter, optionally a Twitter thread. Tells one specific moment from the engagement — the moment of decision, the surprise, the thing that almost went wrong. Links to the full case study at the end.

3. The short post. 50-80 words. A single observation or insight from the engagement, presented as a standalone thought. Doesn't require the case study context. Lives on social as its own artifact.

4. The tactical post. 100-200 words. "Here's what we tried that didn't work, and what we did instead." Practitioner content for an audience of peers. Builds credibility with prospects who care about the actual work, not just the outcomes.

All four come from the same source pull. Each one has a different shape, a different audience, a different hosting surface. You harvest once and ship four times.

The procedure (30 minutes if upstream documents exist)

The procedure assumes you've kept reasonable engagement artifacts — weekly updates, meeting summaries, the project retrospective. If those exist, the harvest is 30 minutes. If they don't, the harvest is 90+ minutes of memory reconstruction and the output quality is lower.

Step 1 (5 min) — Confirm permissions. Email the client a brief note offering three anonymization levels: fully anonymous, named with industry context but no financials, or named with full specifics. Most clients pick the middle. The choice determines what you can write next.

Step 2 (5 min) — Pull source material. Open the client folder. You need: the internal kickoff doc (your original gut read), the project retrospective (the post-wrap version), the chain of weekly updates (where the engagement progressed), and any direct quotes the client gave you across the engagement (emails, Slack, meeting transcripts).

Step 3 (8 min) — Generate the case study. Run a Claude prompt with the source material as input. The prompt should enforce a specific structure: situation → diagnostic → decision → execution → outcome (non-numeric) → what's different now. Each section is two or three sentences. Total under 900 words. Output is a near-final draft; you edit 10-15% of it to ship.

Step 4 (6 min) — Generate the three derivatives. Same source material, three new prompt runs. One asks Claude to extract the single most surprising moment from the engagement and write it as a long-form social post. One asks for a 50-word standalone observation. One asks for a tactical "what didn't work / what did" post. The derivatives are short enough that the Claude output is closer to publishable than the case study.

Step 5 (4 min) — Edit and ship. Light editing pass across all four. The case study goes to your site; the social pieces go to whatever surface you publish on; the short observation can be queued or sent immediately.

Thirty minutes. Four artifacts. Inside two weeks of wrap so the texture is still warm.

The non-numeric outcome (the part that compounds)

The most under-leveraged part of a consulting case study is the non-numeric outcome. Most consultants write case studies organized around a percentage — "increased conversion by 47%." Numbers matter, but they're also the easiest part of a case study to fake or game. Prospects know this. A case study built only on a number reads as marketing.

The non-numeric outcome is the thing the client now does differently because of the engagement. "The product team now reviews session recordings every Monday — they didn't before." "The CEO stopped asking for weekly reports and started asking for monthly because the data quality changed." "The new hires are onboarded with the playbook the engagement produced." These outcomes are harder to fake because they require the kind of specificity that only comes from having actually done the work.

A case study built around a non-numeric outcome — with the number in there as supporting evidence rather than headline — reads as practitioner work, not marketing copy. It builds the kind of trust that turns a site visitor into a prospect. Article 10 in this series — the non-numeric outcome — covers the framing in more depth.

A worked example (same engagement, four artifacts)

Concrete shape. The engagement: six-week onboarding-funnel diagnostic for a B2B SaaS doing bookkeeping for small restaurants. Trial-to-paid conversion went from 9% back to 16%. Client agreed to be named without specific financials.

The case study (~700 words) opens with the situation ("Trial-to-paid conversion had dropped from 18% to 9% in a quarter. The product team and engineering were arguing about the cause"), walks through the diagnostic (session-recording review, funnel-stage breakdown, friction points), names the decision the engagement turned on ("the friction wasn't in onboarding copy — it was in the activation step nobody had instrumented"), describes the execution, and closes with the non-numeric outcome ("the product team now reviews session recordings weekly. They didn't before.") The numeric outcome — conversion recovery — is mentioned once, as supporting evidence.

The long-form social post (~350 words) tells the moment of decision: "Three weeks into the engagement, the engineering lead and the head of product were each certain the other was wrong about the conversion drop. We were stuck. Then we watched 40 session recordings together in one room. By minute 30, both of them said the same thing at the same time: 'oh.' That's the moment the engagement turned." Links to the full case study at the end.

The short post (~70 words) is a single observation: "The most underrated diagnostic tool in B2B SaaS is sitting two people in a room and watching 40 user-session recordings without commentary. Most internal disagreements about funnel performance survive Mixpanel and don't survive recordings. The work isn't in the watching; it's in resisting the urge to narrate while you watch."

The tactical post (~150 words) is the "what didn't work / what did" angle: "What didn't work: another funnel report. The team already had four. Another report would have been ignored. What did: scheduling a 90-minute block on the calendar for the engineering lead and the head of product to watch session recordings together. The artifact was the shared experience, not the deck."

Same source material — the weekly updates, the retrospective, the quotes — produced all four. Total authoring time: under 30 minutes. Each one ships to a different surface (site, LinkedIn, social, social), reaching different audiences from a single harvest.

Pitfalls

Harvesting from memory instead of artifacts. If you don't have weekly updates or meeting summaries from the engagement, the case study you produce will be generic — because you'll only remember the broad shape, not the specific texture. The fix isn't to write a generic case study; it's to install the weekly-update procedure on the next engagement so the artifacts exist when you go to harvest.

Skipping the permissions conversation. Sending an email asking for anonymization preference feels awkward. Skipping it produces three outcomes, all worse: a fully anonymous case study (weak), a named case study without permission (legal risk + relationship damage), or no case study at all. The permissions email is two sentences and resolves all three.

Writing one case study and never deriving the three social pieces. The math of the four-piece harvest is the math that makes it worth doing. One case study from an engagement is fine. Four pieces from the same engagement is what compounds over a year of wraps.

Letting the post-wrap window close. Inside two weeks: 30 minutes of work, sharp output. At one month: 60 minutes of work, decent output. At three months: 90+ minutes, fuzzy output. The compounding cost of delay is what makes the procedure trigger fire automatically — see the SOP version below.

Writing case studies for engagements that didn't go well. Not every wrap should be harvested. If the engagement was difficult, the relationship was rocky, or the outcome was ambiguous, skip the harvest for that one. Forced case studies on weak engagements produce weak case studies, and they pollute the rest of the catalog.

Where the procedure lives

The 30-minute harvest above works when you treat it as a procedure that fires on a trigger — engagement wraps → harvest runs within 14 days — not a discretionary task. The operational version (the four prompts, the case-study structure, the social-piece variants, the permissions email template, and the pitfalls list) is SOP 05 in The Solo Operator's SOP Bundle. Drop it into the engagement folder when the wrap is on the calendar; the procedure fires automatically on the right date.

The framing version — why the non-numeric outcome matters, what the case-study shape looks like editorially, when to break the structure — is in Chapter 7 of The Solo Consultant's AI Playbook. The Playbook chapter is the reading; the SOP is the running procedure.

The cost of not harvesting is the marketing asset that goes unbuilt. Across a year of wraps, that's four to eight case studies, sixteen to thirty-two social pieces, and the difference between a site that signals depth and one that doesn't. Thirty minutes per wrap. Most consultants are leaving the assets on the table.