The mid-story checklist I use so a case study one-pager stays honest

A metric nobody measured, greened because the page looked thin. Two clients' quotes mashed into one spine. A logo "cleared" in a Slack guess. Before/after rewritten until the dump no longer matches. Stale praise from a different project still on the page. That is not mid-story truth. Mid-story is where folklore becomes a fake case study — invented outcomes, mixed attributions, permission blur, and a pretty one-pager pretends the testimonial dump agreed with itself.

I am a freelancer admin helper. I turn messy testimonial dumps into a clean case study one-pager. I am not a CPA, marketer-of-record, lawyer, or PR advisor — and this is not financial, tax, legal, marketing, or contractual advice. It is the checklist I run mid-story (and again before anyone treats the page as what the client said) so the case study stays honest while quotes and outcomes are still arriving.

(If you want the full testimonial dump → case study one-pager workflow, see my earlier post: How I turn a testimonial dump into a case study one-pager. This one is the second-pass checklist so invented metrics, mixed attributions, uncleared quotes, before/after inflation, and stale praise do not quietly rewrite what the dump actually supported.)

What "honest" looks like mid-story

• Every quote is one attributable line — the same praise under two screenshots is merged, not two proof blocks
• Outcomes exist only when evidenced in the dump (or labeled ask / guess / TBD) — remembered "we doubled revenue" folklore is not the result
• One story spine — problem → approach → outcome for this client/project, not a mash of two engagements
• Permission status is labeled cleared / pending / anonymize — not assumed because the page looked ready
• Assumptions are labeled assumption — not written as settled metrics or brand clearance
• Source note + capture date exist — praise from another project or an old campaign is labeled stale, not still greening this page
• Nothing invented beyond what the dump supports — guessed logos and invented before/after numbers are not a case study

If the only “proof” you have is a confident metric and a folder titled “sales will somehow know,” mid-story honesty has already slipped.

The mid-story checklist (run it before / while you clean)

Use this against the dump (emails, Slack exports, review screenshots, Loom transcripts, survey replies, old one-pager if any) before you paste everything into a pretty template and call the page proposal-ready.

1. Inventory the testimonial dump

• [ ] Soft scope is named (this client / this project / this capture window) — out-of-scope praise stays out or flagged
• [ ] Every source is listed once: quote snippet, file, channel, capture date, permission status, notes
• [ ] Duplicate praise or the same quote under two sources (email + screenshot) is labeled duplicate — not two proof lines for one line of praise
• [ ] Superseded or out-of-scope dumps (wrong client, old campaign) are labeled stale — not resurrected as this mid-story page
• [ ] Anything clearly walked back by the client is labeled stay-off-live — do not invent it back into the one-pager

2. Lock spine, outcomes, and attribution before you trust the page

• [ ] Every outcome has evidence or is labeled ask / guess / TBD — blank metrics that look complete are not the dump
• [ ] Client/project naming is locked or ask / anonymize — remembered brand folklore is not clearance
• [ ] One short story spine is locked or mixed-project mashups stay ask — dual spines are not both green
• [ ] Quote attribution and permission tags are listed or marked ask
• [ ] Skim test: you can answer “whose praise, what changed, and what is still uncleared?” without reopening the dump

3. Kill invented metrics / mixed attributions / assumption inflation

• [ ] Same praise under two sources is one line — not two open proof items
• [ ] Missing metrics stay missing / ask — do not invent “about 2x” folklore to green the page
• [ ] Mixed-client blur stays out / excluded or ask — do not rewrite Client A’s quote as Client B’s spine
• [ ] Assumptions are labeled assumption — do not rewrite them as settled results or clearance
• [ ] Open questions stay open — do not mark them closed because the page looked tidier
• [ ] Logos or brand marks without clearance are either backed by a permission note or labeled ask / anonymize — ghost logos are not proof
• [ ] Invented before/after numbers the dump never supported stay guess / ask — not locked as the result

4. Park open asks, tidy-page lies, and superseded notes

• [ ] Invented metrics and uncleared quotes are either evidenced or labeled guess / ask / TBD — not locked because the page looked thin
• [ ] Open asks (permission pending, unclear metric, mixed attribution, stale dump vs this window) sit in one list — not scattered across threads
• [ ] Superseded or walked-back lines are archived once so they do not still look live
• [ ] Anything still incomplete stays incomplete — do not greenwash a mid-story dump into a finished case study

Soft CTA

If you want the testimonial dump cleaned into a case study one-pager for you: messy praise in, proposal-ready page out in 24 hours — $12 on Gumroad. Code ADMIN12 takes $2 off. Full catalog: shop · GitHub Pages.

Not financial, tax, legal, marketing, or contractual advice. I organize the testimonial dump; you own clearance, claims, and how you use the page.