An account spends two thousand pounds a month at a cost per click of two pounds. That is a thousand visits.
Twenty of them convert. Nine hundred and eighty do not, and the report treats all nine hundred and eighty identically, as traffic that was not interested.
Some genuinely were not. A substantial number could not tell whether the business covered their area, what it cost or how quickly it could happen, and left to find somebody who said.
The components of paid waste
Four of them. Wrong audience, wrong page, a page too slow to render, and a page that rendered perfectly and did not answer the visitor's question. Only the last is usually invisible.
Component 1: the audience mismatch
Clicks from people who were never going to buy. This is a targeting problem, it is well understood, and most accounts already work on it continuously.
Component 2: the message mismatch
The advertisement promises one thing and the page delivers something adjacent. Also well understood, also worked on, and the first place most teams look.
Component 3: the speed loss
Visitors who leave before anything renders. Measurable, fixable and frequently a larger share than anybody expects on a mobile connection.
Component 4: the unanswered question
The one nobody instruments. The page loaded, the message matched, the person was interested, and something they needed to know was absent, which is the pattern in bounce rate and the question nobody answered.
A worked example
Two thousand pounds, a thousand clicks, twenty conversions. The cost per conversion is a hundred pounds. Assume a conservative one in twenty of the non converting visits left over an unanswered question, which is forty nine visits.
What those forty nine are worth
If a quarter of them would have converted had somebody answered, that is twelve additional conversions from the same spend. The cost per conversion falls from a hundred pounds to sixty two.
The same arithmetic on the evening share
If a third of the traffic arrives outside working hours, roughly six hundred and sixty pounds of that budget lands when nobody is available, which is the pattern in campaign traffic that arrives outside office hours.
Why the estimate is deliberately conservative
Because the point is not the precise figure. At one in twenty the case is already obvious, and most accounts that measure it properly find the real share is higher.
Identify the specific questions first
Site search terms, the enquiries that do arrive, and what sales gets asked on every first call. This produces the list rather than a guess.
Put the answers on the page
The cheapest fix, and it works for every visitor rather than only the ones who engage. A page that answers price, coverage and timescale converts better before anything else is added.
Then answer what the page cannot
Anything specific to the individual, which is the remainder, and it is where an assistant earns its place, as set out in how to capture leads from paid traffic without landing.
Apply it to the very practical questions too
Coverage, collection and what is included are decision blockers in ordinary local services, which is the ground in garden and green waste collection questions.
Run your own numbers before deciding anything
Spend, clicks, conversions and the evening share, which the chatbot ROI calculator turns into a figure you can defend. See pricing, and the sector view sits under marketing agencies.