Marketing & GrowthFree
Case Study From a Customer Interview
Every claim in the draft carried back to the second of the interview it came from, so the approval round is one pass instead of five.
River reads the interview for what the customer said, then writes the case study from it and builds the register that gets it approved. One row per claim in the draft: the words as spoken, the second of the recording they were spoken at, the person who said them, whether the line is quoted or inferred, and what the claim rests on. The story and the register go out together, because the second one is what makes the first one signable.
The draft was never the bottleneck. Every prompt on this subject already says quote verbatim, never invent a number, and list what is missing. That produces a serviceable draft in twenty minutes. The six weeks that follow belong to the customer's communications lead. They have to defend each line to their own counsel from a document that offers no way to check a single one. The safest thing available to them is to cut it.
Written for product marketers who owe three references this quarter, customer marketing teams running a reference programme, and founders writing the first one themselves. Pair it with an article from a transcript when the same recording holds more than a story, and with the messaging from customer language pack when the verbatims should be feeding the positioning too. It runs in the marketing workspace, so every reference you have ever published sits in one place with its register and its recheck date beside it.
What the endorsement rules require of a customer story
A published customer story is an endorsement, and the FTC's guides are unusually specific about the punctuation. Section 255.1 says an advertisement need not present the endorser's message in their exact words. That holds only until the advertisement represents that it is doing so, such as through the use of quotation marks. The same section bars rewording that distorts the endorser's experience. Quotation marks are the representation, and tidying a hedge out of a sentence inside them is the exact move the rule names.
Then the headline number. A story about a central attribute of the product will be read as representing what customers generally achieve. Section 255.2 requires substantiation for that. Where you do not have it, the page needs a clear and conspicuous disclosure of the generally expected performance instead. The guides work an example in which a footnote reading "results not typical" is held insufficient. So the question a register answers is not whether the customer said the number. It is whether your other customers get it.
Two more columns earn their place from the same source. An endorser has to be a bona fide user when the endorsement is given, and the advertisement runs only while you have good reason to believe they still are, which is why an approved story carries a recheck date. And a discount, free product or early access given to the customer is a material connection that has to be disclosed, which means somebody has to know the commercial terms before the story publishes.
How it works
Paste the interview
Any transcript export, with timestamps or without, plus the account data behind the numbers.
River sources each claim
Every line traced to who said it, when, and whether it is quotation or inference.
Send the register
Their reviewer approves row by row rather than redlining prose they cannot check.
Publish and diarise
Ship the approved version and hold the recheck date so the story stays true.
What you get
- The case study drafted from the interview, with every quotation reproduced exactly as spoken
- A claim register: one row per claim, its timestamp, its speaker and its evidence
- Quoted claims held apart from inferred ones, so nothing of yours sits inside their quotation marks
- Every figure checked against the usage data you supply, with mismatches named rather than split
- A recheck date on the published story, because an endorsement runs only while it stays true
- The one-page sales version and the pull quotes, cut from approved claims only
Common questions
Does this run the customer interview for me?
It works from the recording of one you already ran. Bring any transcript export, with timestamps or without. Where the interview never got to a number, the register says so with the exact follow-up question to send, rather than reaching for the nearest figure in your own dashboard and attributing it to them. Where the recording is research rather than a reference, start with the interview transcript workup.
What if the customer gave no numbers at all?
Then the story is built on what they did say, and the register is honest that the results section rests on your data rather than theirs. That distinction matters to their reviewer, who will approve a page describing your measurement of their account far faster than a page putting your measurement in their mouth.
Their comms lead wants to rewrite their own quote. Now what?
Take it. A quote the endorser rewrites is still their honest statement, which is the standard that matters. The register logs the replacement alongside the original timestamp, so six months later you can see which lines came from the recording and which came from their approval round, and nobody relitigates it.
How is this different from a case study prompt?
A prompt gives you a draft with verbatim quotes and a list of missing metrics, which is genuinely useful and takes twenty minutes. It does nothing about the approval cycle, which takes six weeks. The register is the artifact their reviewer works from, and it converts a redline argument into a row-by-row sign-off.
Is this overkill for a two-line testimonial?
A two-line testimonial on your homepage is an endorsement on exactly the same terms as a four-page story, and it usually carries the boldest number on the site. The register for one is four rows and takes a minute. Run it, particularly where the line quotes a figure or names a competitor.
What happens when the customer churns?
That is what the recheck date is for. A published story asserts a current relationship, so the register carries their renewal date and a review date ahead of it. When a reference lapses you pull the page or you rewrite it in the past tense, and the register tells you which pages are affected the same day.
Case Study From a Customer Interview
Fill in the form and your workspace opens with the work already underway.