Client zero: why we build Storymachine on ourselves first

In every product decision we make, we hold to one rule: we use it ourselves before a client does. Every feature runs first on our own content production. Every workflow proves itself first on our own agency work. Internally we call that 'client zero'. It isn't a marketing slogan — it's an operational discipline that decides what does and doesn't make it into the product.

Why this works

Products not used by their makers evolve toward features that sound good in a demo. They pick up dashboards that look great in screenshots, integrations that add checkmarks on a landing page, and workflows that look sensible on paper but nobody really uses in practice.

Products used daily by their makers evolve differently. Friction becomes visible before a customer complains. Unnecessary buttons die because we get tired of clicking. Edge cases surface in real work, not in a QA report. The product takes the shape of something you actually use, not something you sell.

• →We feel friction before a customer feels it.
• →We discover edge cases in our own work, not in support tickets.
• →We dare to cut what we ourselves don't find useful.
• →We know exactly what we promise, because we've done it ourselves.

What we've learnt so far

There are three lessons we could only learn by being our own customer. They sound obvious in hindsight. We couldn't have thought of them upfront.

Lesson 1: A brand brain only becomes valuable after weeks

In the first week the brand brain produces mediocrity. It doesn't know enough yet. Output is better than standalone ChatGPT, but not spectacularly so. Any customer passing final judgement on day three would conclude it doesn't work.

From week four on that changes. Enough corrections, enough examples, enough context. The first draft suddenly reads like something we could have written ourselves — faster than we could have written it. From week eight it becomes almost unfair: the brain notices when we've used a claim before, knows our favourite examples, knows which terms we avoid.

We only have that timeline because we're our own customer. Clients who used to complain about day three now understand the curve looks the way it does — because we've experienced it that way ourselves.

Lesson 2: Editorial work is the real work

When we designed the product we thought the generation step was the core. Build a great generation engine and the rest follows. In practice we noticed something different: the time we save typing shifts to editing. It's not that there's less work — it's different work, at a different level.

That insight changed the product. We put much more thought into the editing interface than into generation settings. Comparing variants. Quickly accepting changes. Feeding back what we changed and why. That's where people spend time, so that's where most of the product love belongs.

"We don't sell a tool. We sell what the tool has become after three months."

Lesson 3: Our own mistakes are the best roadmap

Every time we get internally frustrated about something in the product, that's a feature idea. Every time we take a detour, that's a workflow improvement. Every time we have to explain to a colleague how something works, that's a UX problem. We don't need external user research to know what's wrong — we bump into it daily ourselves.

That doesn't mean we don't talk to customers. It means we know what's broken earlier than customers can articulate. Customers come with symptoms ('it's too slow'). We already know the cause ('the retrieval step is inefficient on long documents'), because we ran into it yesterday.

The downside

Client zero isn't risk-free. You risk your product looking too much like you as a customer and too little like the broader market. We're a small team of seniors. A ten-junior marketing team at a corporate has different needs. We don't see that when we only build on ourselves.

So we compensate on purpose. In parallel we work with ten external clients who actively provide input. We observe how they use it, we listen to their frustrations, and we test features on their work before generalising. Client zero is the first filter, not the last.

Why we write about this

Because we often get asked: 'Does this really work?' The answer isn't 'yes because we have good engineering'. The answer is: 'yes, because we've been running it on ourselves for two years'. Every piece of content we publish — including this one — comes out of our own brand brain. Every workflow we deliver to clients, we've made fail and work on ourselves hundreds of times.

Client zero isn't an original strategy. Amazon did it with AWS. Basecamp does it with Basecamp. Figma built Figma in Figma. It's an old principle. We apply it in agency form — and it's why we dare to promise what we promise.