Story mining is a structured search for the narratives your company already contains. Almost no business has a story shortage; it has a retrieval problem. The material exists in support tickets, sales-call recordings, win/loss notes, onboarding threads and the anecdote one rep tells in every meeting — it just is not written down, indexed, or recognised as an asset. What follows is the session we run: 45 minutes with the four or five people who touch customers, an explicit list of the places stories hide, and a filing convention that makes them findable in nine months. The output is a five-story portfolio — one origin story, one customer transformation, one objection story, one failure-and-fix, one employee story.
- "We don't have any stories" is almost always false. Nobody's job is to notice them, so they evaporate.
- The highest-value story in most companies is the anecdote a rep always tells — already tested on real buyers, never written down.
- Score every candidate on one criterion above the others: would the subject repeat it on camera unprompted?
- A one-line index beats a folder of beautiful case-study PDFs nobody can find.
- Film two of the five. Keep the rest written — we'd rather say that than sell you five films.
You don't have a story problem
When a client tells us they have nothing to work with, we ask their support lead one question — "who was furious last quarter and is happy now?" — and the answer arrives in about four seconds, with a name.
What is actually true is narrower and more fixable: nobody's job is to notice stories, so they evaporate. A support agent resolves a hard ticket and closes it. A rep wins a competitive deal and moves to the next one. An onboarding lead hears a customer say something startling in week two and repeats it at lunch, where it dies. Every one of those was an asset for about ninety seconds. Story mining is simply the practice of assigning someone to notice, and giving them somewhere to put it.
The five places stories hide
Search these before you book any session. Twenty minutes in each will give you more raw material than a workshop can.
Support tickets — the angry-then-resolved thread
Filter for tickets that opened hot and closed with thanks. That arc is a complete story with the tension already documented, and the customer has demonstrated they will speak plainly. Search for "still not working", then read what happened next.
Sales-call recordings — the moment a prospect says "we tried that"
Search transcripts for "we tried", "we used to" and "the problem was". Those phrases mark the before-state in someone's own words, which is the beat marketing can never write convincingly. You are mining for language as much as for events.
Onboarding threads — the surprise
Week-two messages contain the most honest sentences a customer will ever send you: what they expected, what actually happened, what nobody warned them about. Search for "didn't realise" and "turns out".
Win/loss and churn notes — the most honest documents your company owns
Nobody writes these to be read externally, which is exactly why they are useful. Losses tell you which doubts to build objection stories against. Churn notes tell you what your best story must not overclaim.
The anecdote a rep always tells — the highest-value one
Every sales team has a story someone tells in every pitch. It has been tested on dozens of real buyers and refined by their reactions, and it is almost never documented anywhere. Ask "what story do you always end up telling?" and write down the answer verbatim. Then check it — battle-tested does not mean accurate, and a story that has been retold fifty times has usually drifted.
The 45-minute story-mining session
Timed, deliberately short, and run once a quarter. The failure mode is well known: a session that turns into a debate about messaging produces nothing. Guard the clock.
Support, sales, delivery or onboarding, and whoever has been there longest. Marketing facilitates and does not answer first. A session composed only of marketers produces positioning, not stories.
No polishing, and no deciding what is good yet. Both rules exist to stop the room editing before it has collected anything. Say explicitly that half-remembered fragments are welcome.
"A customer who was sceptical and changed their mind." · "The moment you knew this was working." · "A time we got it wrong and fixed it." Run them in that order and do not move on early — the second and third answers in each round are better than the first, because people stop performing.
Three fields only: who · what changed · the number if there is one. One line per story. Paragraphs at this stage are how a session runs out of time with four stories instead of twenty.
Three columns, one to five: specificity (is there a named person, moment or number), verifiability (could a stranger check it), and the one that matters most — would the subject repeat this on camera unprompted? If it only exists because someone in this room suggested the phrasing, it is not a story you own.
Three named owners, three dates. Everything else goes into the index with its consent status, where it stays findable rather than being lost for the second time.
Across 31 customer-story films we have delivered, exactly one subject has asked us to change the edit (the count and its limits). Subjects agree to almost anything: suggest that an answer sounded good and you get a tidier, stronger version, offered gratefully. Which means a customer's approval tells you nothing about whether a story was theirs. The only test that survives that is the unprompted one — would they have said this without anyone shaping it? Score honestly here and the rest of the process stays clean. Score generously and you will spend money filming a sentence you wrote.
The five-story portfolio
Five stories cover almost every situation a small company faces. Build them once and maintain them, rather than commissioning something new every quarter.
| Story | Audience | Length | Where it lives |
|---|---|---|---|
| Origin | Everyone, once | 200 words | About page, first meetings, hiring |
| Customer transformation | Prospects — the workhorse | 2 min film + page | Home, proposals, sales follow-up |
| Objection story | One doubt, per deal | 40 sec spoken | Rep library, demo, objection moment |
| Failure-and-fix | Sceptics and senior buyers | 400 words | Blog, late-stage conversations |
| Employee story | Candidates | 90 sec film | Careers page, offer stage |
The failure-and-fix is the one almost nobody builds, and it is the cheapest trust you can buy: an account of something you got wrong, what it cost, and what changed as a result. It disarms the sceptic in the room because it is the one story a competitor will not copy. The customer transformation is the workhorse, used weekly — that is the one worth real production. Details of building it are in the customer story method, and deploying the objection story is in sales storytelling.
How to file them so they get used
A beautiful case-study PDF in a folder nobody opens is worth nothing. The asset that gets used is a one-line index, in the tool your reps already have open.
Five fields per row: industry · the doubt it answers · the number · where the asset lives · consent status. The doubt field is the one that changes behaviour, because reps search by objection, not by vertical — nobody has ever thought "I need a manufacturing story", they think "she doesn't believe this will survive procurement". The consent field prevents the other common failure: finding the perfect story two hours before a pitch and having no idea whether you are allowed to name the client.
Keep it to one screen. An index that scrolls is an index that gets skimmed and then abandoned.
Which of the five deserve to be filmed
We make films for a living, so treat this as an interested party arguing against its own interest — which is the only reason it is worth reading.
Film the stories where tone carries the proof and the person is willing to be named. That is usually two of the five: the customer transformation, where hesitation and the pause before an honest sentence are the evidence, and the employee story, where a candidate is judging a person rather than a claim. Both depend on something writing cannot transmit.
Keep the others written. The origin story belongs in text because it gets read, not watched, and it changes as the company changes. The failure-and-fix works better written because readers want to re-read the difficult paragraph, and because a filmed apology reads as staged even when it is not. The objection story should live as forty spoken seconds in a rep's head, not as an asset at all.
So the honest answer for most companies is two films and three documents, not a video programme. If a production company tells you all five need filming, ask which two they would keep if the budget halved.
Download: the Story Mining Worksheet
One printable page containing everything above that you need in the room: the three prompts, the capture grid, the three scoring columns, and the five-story checklist. Free, no email required, no form — we would rather it was used than tracked.
Print one per person. Bring a timer.
Sources
- SeventhSlate first-hand count: 31 customer-story films delivered to date; one subject requested a change to the edit after review. Own internal record, not an audited sample; method and limits on the storytelling statistics page.
- Green, M. C., & Brock, T. C. (2000). The Role of Transportation in the Persuasiveness of Public Narratives. Journal of Personality and Social Psychology, 79(5), 701–721. Four experiments, N=97, 69, 274, 258.
- The session structure is our own working practice, refined across client engagements. It is a method, not a research finding.
- Search the five hiding places before you book a meeting. The material is already there.
- Score on the unprompted test. Everything else is negotiable; that one is not.
- Two films, three documents, one index your reps can actually read.
