The human leg
What your people publish
Each quarter, your published story — site, posts, materials — read against the canonical one. Drift named with receipts, and the fixes small because they are early.
Narrative Governance
Stories do not break. They drift.
Six months after the rewrite, the old words creep back — a “we” here, a feature list there, a machine answer quietly going stale. Narrative Governance is the quarterly rhythm that keeps your story true.
CHF 750 per quarter.
Your first quarter is included with the Charter — because the point is not a report, it is the habit.
Drift is free on the day it starts. By the time someone notices, the old words are back in the deck and in what the machines say about you.
Book a story conversation
It holds while one person is watching
Early on, the story lives in one head. The founder writes the deck, answers the difficult emails, rewrites the homepage on a Sunday. Everything sounds like the same company because it is the same person.
Then the company grows. A sales hire starts posting. An agency takes over the newsletter. Support answers four hundred tickets a month. Someone connects a model to the CRM and it starts drafting replies. Nobody decides to change the story. Each of them fills the gaps with their own habits — and the newest writer on your team has the strongest habits of all, because it learned them from everybody else’s copy.
If your story only holds while its author checks every draft, you do not have a governed story. You have a bottleneck with a job title.
Two audiences drift, not one
A story is a living thing told by many mouths: your pages, your team’s posts, your deck-of-the-week — and a second audience of machines retelling you to buyers you will never see. None of them drifts on purpose. All of them drift.
The machine leg
What the machines say back
Your Second Audience reading, re-run on the same questions and the same engines, scored the same way — so you can see the machines’ version of you move, in ranges, with dates, honestly.
A story only stays true if the people who never attended the workshop can tell when something is off.
Your team already has the test for that: the drift rubric, the checkpoints and the decision rights they built in the Narrative Charter. That is what stops drift at the draft. Governance is how you find out whether it is holding.
What drift looks like, and what we do with it
Three lines from one quarter at one company. Each is reasonable on its own. Together they describe three different companies.
An invented example, built to show the mechanics — not a client. A twelve-person company selling payroll software to Swiss SMEs.
Three rules from its Charter
- The customer is the operations lead who runs payroll on top of her real job. Every piece opens with her month-end, never with our product.
- Payroll is the one day a month every employee checks whether the company can be trusted. A piece that sells only speed or features is off-story.
- We never frighten anyone with fines or penalties.
LinkedIn — a sales hire, week three
What went out
🚀 Excited to share our new AI-powered payroll engine: 40+ integrations, real-time sync and automated compliance!
Why it is off-story It opens with us and sells features alone. Rules 1 and 2.
Back in the story
Last month-end, an operations lead ran payroll for 84 people between two meetings. Nobody came to her desk afterwards with a question about their payslip. That is the job.
Same product, same week. The fix keeps the writer’s voice. It changes who the story is about.
Support — the reply after salaries went out a day late
What went out
We apologise for any inconvenience. Due to a technical issue, processing was delayed. Our team is working to resolve the matter.
Why it is off-story Correct, polite, and off-story: it treats the one thing the company promises as an inconvenience.
Back in the story
Your team was paid a day late, and that is the one thing we promise will not happen. Here is what broke, what we have already changed, and a note you can forward to your staff this morning.
Stories break first under pressure — apologies, price changes, outages. So the quarterly reading always includes them, not only the marketing.
An AI assistant, asked for payroll software for a small Swiss company
What went out
“A low-cost payroll tool with many integrations and a per-employee price.”
Why it is off-story The machines have filed the company in the price-and-features category its Charter refuses.
Back in the story
Nobody can edit a machine’s answer. We trace where it learned it — here, a pricing page that opens with the per-employee figure, and an integrations page with no customer on it — and those two pages go on the quarter’s list.
The machine leg is read in ranges, with dates. One answer is an anecdote; the same answer across engines and quarters is a finding.
What arrives, each quarter
Three on a rhythm, and one that does not wait for it.
The written reading
Drift named with receipts on the human side, and the machine deltas in ranges with dates. Short, and specific enough to act on — the record of where your story actually stands this quarter.
Half an hour together
We walk the reading. The report is the record; the call is where you decide what to do about it. A reading nobody discusses is how a retainer quietly becomes a subscription.
Your Charter, kept current
Governance maintains the Charter rather than producing another document. Each quarter, the two or three pieces that held the story best go into it as examples, with a line on why they work — so the next hire learns from your own best work, not from a rulebook. And when one channel keeps straining a rule, it gets a written translation for that channel instead of every writer improvising one.
And between quarters
An automated tripwire watches for the drops that cannot wait for the next reading. It is the exception, not the rhythm — most drift is slow, which is exactly why a quarter is the honest cadence.
A quarter, start to finish
The same four moves every time. Nothing depends on anyone remembering what was said last time.
01
We sample what went out
Per channel, the ten most recent pieces and the five that reached the most people: recency shows where the story is heading, reach shows what the market actually received. Plus the hard surfaces — support replies, apologies, anything announcing a price. And the Second Audience questions, re-run on the same engines.
02
We read it against your Charter
Every sampled piece is held against the rubric your team built, the same way your own reviewers apply it. Each flag names the rule, quotes the line and proposes the fix inside your story. A flag without a fix is a complaint, and you will not get any.
03
We talk it through
Half an hour on what moved and why. When flags cluster on one rule there are three possible causes — the rule is ambiguous, the story is weak at that point, or the channel genuinely strains it — and each has a different fix. Choosing between them is your narrative owner’s call, not ours.
04
Your Charter moves a version
Examples added, a rule sharpened, a channel translation written. The change log records why. The next quarter measures whether it worked.
What it costs
CHF 750 per quarter
Your first quarter is included with the Charter. Governance starts as part of the work that produced the story, rather than as a separate decision taken months later when the drift has already happened.
Quarterly rather than monthly, and deliberately so. Checked every month, most of what you would be looking at is noise rather than movement — a quarter is the shortest span over which a change in how your story is landing actually means something, and we would rather bill on the rhythm the evidence supports than on the one that produces more invoices. It is also cancellable: if the habit has taken and your team no longer needs us in the loop, that is the outcome, not a lost customer.
Test the guide you already have
Most companies arrive with a brand book. Two tests show, in an afternoon, whether yours can govern anything. You do not need us for either.
The adjective test
Open your brand guidelines and count two things: the adjectives describing how you sound, and the sentences a reviewer could hold a draft against and answer yes or no. “Warm, bold, approachable” describes a feeling. It does not tell a writer what to do, and it does not let anyone say no.
- Describes We are approachable.
- Governs Every email answers the reader’s question in its first two lines.
- Describes We are bold.
- Governs Any piece over fifty words states our standpoint, outright or unmistakably.
The three-writer test
Give the same brief and the same guidelines to two colleagues and one AI assistant, then put the three drafts side by side. If they read like three companies, your guide describes a voice; it does not govern a story. The machine’s draft is the most useful of the three: it shows where your copy goes by default when nobody is watching — eager, abundant, explaining.
What to compare
- Who the first line is about — your customer, or you.
- Which category each draft puts you in.
- What each one promises, and whether you could keep it.
A voice guide makes everyone sound the same. A Charter makes everyone tell the same story. If your guide fails both tests, the next step is not Governance — it is just below.
It needs a story to govern
Governance is not a way of finding your story. It is a way of keeping one true, which means there has to be one — agreed, written down, and owned by the people who publish. Sold to a company that has not done that, it becomes a rulebook enforcing a story nobody settled on.
If you do not have that yet, the honest order is the Narrative Charter first, and Governance at the end of it — the moment it actually makes sense is you have just rebuilt the story; now keep it true. If you are not sure whether your story needs rebuilding, that is what the Narrative Audit answers.
We would rather tell you that now than sell you this today.
Book a 30-minute story conversation
We will tell you whether Governance is the right step yet. No pitch. No obligation.
