§ Blog 2 July 2026

Planneropause

A glass installer with Harleys beat a six-year enterprise AI roadmap in one afternoon. Twenty euros a month, paid back in thirty minutes. The gap is not tools. It is planneropause.

build log
enterprise
agentic ai

A friend of mine runs a glass business in a smaller city. Big personality. Harleys in the yard. The kind of operator who measures twice because the customer is standing there, not because a steering committee asked for a methodology.

We talked AI for a couple of weeks. Then we sat down for maybe two hours.

By the end he had automated and illustrated the gains in a way I rarely see from people who attend quarterly transformation forums for a living. Same tools many enterprises already license. Different metabolism.

What he actually did

He took an AI-assisted workflow into his measurement and pricing stack. The boring part. The part that decides whether you win the quote or lose margin on a Tuesday.

He challenged his price list against the market, line by line, with AI doing the tedious comparison work he used to postpone until Sunday night. He is supposed to be the expensive option in town. He intends to stay there. The question was not can we go cheaper. The question was can we stay premium without bleeding hours into admin.

He found hours. He found clarity. He found a workflow he could show to his own team without a forty-slide preamble about digital ambition.

That last bit matters. Illustration is not decoration. When a hands-on owner can point at before and after, the conversation moves. Enterprises sometimes forget that proof can be a screenshot and a stopwatch, not a programme office.

Planneropause

I want to name the disease on the other side of the table.

Planneropause (n.): the condition in which an organisation stops acting on AI until every technical detail, every cost driver, and every plausible evolution of the tooling over the next six years has been modelled to the satisfaction of people who will not operate the workflow on day one.

Symptoms include the following.

A pilot that waits for a TCO model before anyone measures a single window frame.
A request to understand token economics, infrastructure depreciation, and vendor roadmap variance before spending twenty euros on a subscription.
A workshop that produces alignment and a second workshop that produces alignment about the alignment.
Enthusiasm, genuinely held, trapped behind the need to be certain about step seventeen before trying step one.

Planneropause is not stupidity. It is risk management stretched until it becomes a hobby. Large corporations have good reasons to move carefully. Regulated environments exist. Procurement exists. The audit committee did not wake up evil.

The funny part is timing. While enterprises schedule the next planning cycle, the Harley guy is already compounding.

Twenty euros, thirty minutes

His subscription costs roughly twenty euros a month. Call it twenty dollars if you prefer the exchange rate of storytelling.

He paid it back in less than thirty minutes of hands-on use. Not projected payback. Not a business case with a three-year horizon and a sensitivity tab. Actual Tuesday afternoon.

Since then he reports many more hours saved every month. Pricing checks that used to eat evenings. Measurement follow-ups that used to sit in a mental queue behind installations and supplier calls. The work is still his. The assistant is still an assistant, not an autonomous agent army. That is the punchline for people waiting for perfect agentic maturity. You do not need the finale to get value from the opening act.

Agents will add more. He knows that. He is not blocking on a six-year agent roadmap. He is using what works now and leaving room for what comes next. That is not recklessness. That is how small operators stay on top when larger competitors move like continental drift.

Doing beats waiting

The strict difference is temperament.

He is hands-on, enthusiastic, and open-minded in the practical sense. Try it Tuesday. Adjust Wednesday. Show the team Thursday. He does not need to understand every cost driver of every hypothetical step in 2032 before he touches a tool that saves him an hour tonight.

Enterprises often invert the order. Understand everything, then act. Model everything, then pilot. Align everyone, then align the alignment owners. The result is not always failure. Sometimes it is a superb production platform two years later. Sometimes it is also a smaller city glass installer eating your lunch because he stopped waiting for certainty.

I have sat in rooms where very smart people debated whether AI spend should be OpEx or CapEx while a line manager somewhere was already three workflows ahead with a personal subscription and a spreadsheet they are not supposed to use. Planneropause does not stop shadow IT. It feeds it.

What enterprises could steal

You do not need Harleys. You need permission structure.

Run a thirty-minute ROI test on one real workflow before you build the six-year model. If twenty euros cannot earn thirty minutes back, you have learned something cheap.
Let operators illustrate gains the way this friend did. Visual proof travels faster than architecture diagrams in businesses where the owner still answers the phone.
Separate explore from commit (I keep saying this because it keeps being ignored). A hands-on trial is not enterprise production. It is evidence that enterprise production might be worth the alignment tax.
Treat planneropause as a risk, not a virtue. Planning is necessary. Planning as a substitute for contact with reality is expensive comedy.
Assume agents are a sequel, not a prerequisite. AI-assisted work today funds the patience for governed agent loops tomorrow.

The enterprises that win will not be the ones that skip governance. They will be the ones that govern without freezing. Planneropause feels safe. It is also how you discover, too late, that the market learned by doing while you learned by scheduling.

The personality premium

One more observation, because it matters and because it is funny.

My friend is not a technologist. He is a character. People buy from him partly because he is reliable, partly because he is memorable. AI did not replace that. It gave him back time to be more of what customers already pay for.

Enterprises sometimes treat personality as irrelevant to transformation. Then they wonder why adoption stalls. Adoption is not only SSO and policy. It is whether the person on the ground feels the tool helps them stay on top of the game.

He does. That showed in two hours.

I have seen enterprise programmes spend more than twenty euros on the catering for a single alignment session and still not produce a illustrated before-and-after a regional installer would trust.

That is planneropause. The treatment is embarrassingly simple. Try it. Measure it. Show it. Then, if you must, build the six-year slide.

Field note from the build-in-public log. NDA-safe, no client names, friend anonymised. If you recognise the pattern, get in touch.

§ Below the line

Podčárník · sideways glance, not a summary

On the man who priced a skylight before procurement finished the agenda

The workshop had forty-five minutes left and we were still on slide four.

Slide four was a cost-driver matrix for a six-year horizon. I am not making fun of the matrix. Someone worked hard on the matrix. The matrix had colours. The matrix had footnotes. The matrix had a row for agent maturity uncertainty, which is a polite way of saying we would like to predict the sequel before we watch the film.

In the chat, someone pasted a link to a glass company three hundred kilometres away. Not a competitor in the corporate sense. A local installer with a reputation and, apparently, a yard full of motorcycles. A client had mentioned him. He had sent a before-and-after pricing comparison the same afternoon he tried a twenty-euro subscription. Illustrated. Numbered. No programme office. No kicker slide.

The facilitator asked whether we should park that example as anecdotal.

We parked it. We returned to the matrix. I wrote planneropause in the margin of my notebook, which is not a corporate term and will never survive export to PowerPoint. That is probably why it is accurate.

The man with the Harleys did not wait for step seventeen. He measured a skylight, challenged his own price list, and stayed expensive on purpose. The enterprise did not fail to care. It cared so thoroughly that action needed a fever to break.

I left with coffee and a suspicion I cannot shake. The subscription pays for itself in thirty minutes if your hands are already on the work. The rest of the afternoon is for people who need a matrix to believe Tuesday exists.

Česky original podčárník

O muži, co ocenil střešní okno dřív, než procurement dořešil bod čtyři

Na workshopu zbývalo čtyřicet pět minut a pořád jsme byli na slidu čtyři.

Slide čtyři byl cost-driver matrix na šestiletý horizont. Matrixu se směju ne. Někdo na matrixu dřel. Matrix měl barvy. Matrix měl poznámky pod čarou. Matrix měl řádek agent maturity uncertainty, což je zdvořilý způsob, jak říct chceme znát pokračování, než pustíme film.

V chatu někdo poslal odkaz na sklenářskou firmu tři sta kilometrů odsud. Ne korporátní konkurent. Místní instalatér s pověstí a údajně s dvorkem plným motorek. Klient ho zmínil. Poslal srovnání cen před a po tentýž den, co vyzkoušel dvacetieurové předplatné. Ilustrované. Očíslované. Bez programme office. Bez kicker slidu.

Facilitátor se ptal, jestli to máme odložit jako anekdotu.

Odložili jsme. Vrátili se k matrixu. Do zápisníku jsem si napsal planneropauzu, což není firemní termín a v PowerPointu to nikdy nepřežije. Proto je to asi přesné.

Muž s harleyi nečekal na krok sedmnáct. Změřil střešní okno, zpochybnil vlastní ceník a záměrně zůstal drahý. Podnik nezanedbává péči. Pečuje tak důkladně, že akce potřebuje zlomit horečku.

Odešel jsem s kávou a podezřením, které si neodnáším. Předplatné se zaplatí za třicet minut, když už máte ruce v práci. Zbytek odpoledne je pro lidi, kteří potřebují matrix, aby uvěřili, že úterý existuje.