IdeasSingle-case observation
Leave While It Is Still Working
The decision that reads as agency was taken roughly two years before it had to be.
Why this matters
Most decisions in the archive answer a constraint that has already arrived. This one answers a constraint that had not — which is a materially harder thing to do, and worth marking even though only one case shows it.
Case evidence
2 documented claims
Each entry shows the record first and our reading second. The quoted claim is read directly from the case — we do not restate it here, so it cannot be sharpened in the retelling.
01The counter he switched off.
The Constraint
In 2019 Almaty rolled out the Onay cashless transit system. In his words: once cash payments disappeared, the device would not make much sense any more.
The constraint is dated and external, and at the time of the decision it had not yet reached his market. It was read as a forecast rather than experienced as a problem.
02The counter he switched off.
The Decision
Abandon a working, revenue-positive product ahead of the shift, and carry forward only the insight underneath it — that most offline businesses operate without data.
Shutting down a revenue-positive product is costly precisely because nothing is forcing it. The cost is what makes the decision legible as a choice rather than a reaction.
Not yet a pattern
One case cannot establish a pattern, and this reading rests on one. It is published as an observation so that it can be tested against the archive as the archive grows.
one case is a note, not a findingWhere this stops holding
One case is not a pattern, and this one is not presented as one. We would need at least two more independently documented instances before claiming anything general — and anticipation is unusually hard to evidence, because a correct forecast leaves no visible crisis behind it.
What to do with it
Treat this as a question rather than a lesson: what are you running that is working now and will not be working in two years?
Sources
Everything under this idea.
An idea is only as good as the cases beneath it, and a case is only as good as its sources. These are all of them.
Entrepreneurs from Tashkent created AI cameras for stores
digitalbusiness.kz · 2025-11-03
ArticleIndependent reporting
TASS Vision raises €1.3M to bring privacy-first analytics to physical retail
Tech.eu · 2025-07-14
ArticleIndependent reporting
Success story of TASS Vision: from bus sensors to AI-powered retail analytics
IT Park Uzbekistan
ArticlePublished by an involved organisation