An operations manager is asked forty questions a day by colleagues. She answers them all, because she knows the answers and it takes two minutes each.
The documentation exists. It is complete, accurate and stored in a system with a search function that does not work well enough for anybody to trust it.
So they ask her, because asking her takes eight seconds and always works. Her calendar shows meetings and her day is actually spent as a search engine.
Myth: the answer is to write more documentation
Reality is that the documents usually exist. The problem is that finding and trusting them takes longer than asking a person, and people optimise for their own time.
Reality: findability is the whole problem
Searchable by the words colleagues actually use, returning one confident answer rather than eleven documents. This is what separates a used library from an ignored one.
Myth: people are being lazy
Reality is that they are being efficient. Asking is faster and more reliable, and until the alternative is genuinely better they will keep doing the rational thing.
Reality: the volume is concentrated
A small number of questions account for most of the interruptions, which is the shape found in mapping where repetitive enquiries actually come from.
Myth: the cost is a couple of minutes each
Reality is that the interruption costs far more than the answer, because the recovery of concentration afterwards is longer than the exchange itself.
Reality: the expert becomes a bottleneck
One person holding the answers means the organisation runs at the speed of their availability, and their holiday is a visible event across several teams.
Myth: new starters will learn and stop asking
Reality is that they learn from whoever they asked, imperfectly, and become the next source of slightly wrong answers for the person after them.
Reality: the answers must be owned and dated
An unowned document is out of date within a year and then actively harmful, which is the discipline in building a single source of truth for answers.
Myth: this is an internal problem only
Reality is that the same questions arrive from customers, so the same answers serve both, which is the screening described in receptionist screening questions.
Reality: seasonal spikes make it worse
Periods with a predictable rush concentrate the same questions into a fortnight, which is the pattern in seasonal and winter tyre enquiries.
Myth: it can be fixed by asking people not to interrupt
Reality is that this moves the cost rather than removing it. People wait, batch it up, or make a decision without the answer, and the third is the expensive outcome.
Reality: some questions should still reach a person
Exceptions, judgement calls and anything with authority attached. Protecting that category is the actual goal, rather than eliminating questions in general.
Reality: measure before and after with real numbers
Questions per day, handling time and the recovery cost, which the chatbot ROI calculator turns into a figure that survives a budget conversation.
Make the answer faster to find than a colleague
An assistant answering the documented questions instantly, in the words people actually use, beats asking a person on speed for the first time, which is the only condition under which the behaviour changes. See pricing, and the sector view sits under small business.