Businesses answer the easy questions and drop the hard ones

Everyone in this industry writes about speed. Reply faster, reply first, reply within five minutes. There is a more uncomfortable finding sitting underneath it, and it has now been measured properly in at least one industry.

Pied Piper’s 2026 Internet Lead Effectiveness study, published on 23 February 2026, submitted enquiries through 3,290 dealership websites and measured what came back over the following 24 hours. Two sentences from it matter more than everything else in this category:

“Dealers answer a web customer’s ‘typical’ inquiry within 24 hours 78% of the time on average. However, that rate drops to just 51% when customers ask more complex questions requiring thoughtful human engagement.”

Same dealers. Same day. Same inbox. The only thing that changed was how hard the question was to answer.

This is not a speed problem

A business could reply to every easy question in four minutes, post an outstanding average response time, and still be losing the enquiries that were worth the most, because the hard question is almost always the valuable one.

The hard question is the customer who has already narrowed it down and now needs one specific thing confirmed. It is the one with a complication: a part-exchange with outstanding finance, a spec that needs checking, a timeline that depends on something else. Those are not tyre-kickers. Those are people far enough along to have run into a real detail.

And they are the ones being dropped, at roughly a third the rate again of everybody else.

Why it happens, which is not laziness

An easy question can be answered in a gap: between two customers, waiting for a kettle, at a red light. A hard one cannot. It needs someone to look something up, ask a colleague, or make a judgement they might be held to.

So it gets set aside for a better moment. The better moment requires an uninterrupted ten minutes, and in a working day full of people who are physically present, an uninterrupted ten minutes does not arrive. By the end of the week it is buried, and answering it now means explaining the delay, which is its own small deterrent.

Nobody decides not to answer the hard one. It just never becomes the most urgent thing on the screen.

The same study found the mechanism, in its own measurements: evaluations requiring “human help” scored nine points lower on average, and customers asking them were twice as likely to receive no personal response at all.

Automation moves the cliff rather than removing it

Worth reading carefully if you are shopping for AI: the same study found the industry average improved to 71, up from 65 the year before, and the share of dealerships failing to respond personally fell from 19% in 2025 to 14% in 2026. It attributes much of that to AI and automation. Multi-channel responses rose from 49% to 62%.

But the complexity gap survived all of it. The study’s own phrasing on that point is the line to keep: “When AI required human intervention, customer questions were twice as likely to go unanswered.”

Which is the whole design problem in a sentence. Automating the easy questions raises the average and leaves the expensive failure untouched; arguably worse, because the easy volume now flows so smoothly that nobody notices the hard cases piling up behind it.

A second industry, and why we are not going to lean on it

There is a well-travelled property statistic that appears to show the same shape: agents answering a rental enquiry but missing the valuation request bundled inside it, the valuation being by far the more valuable half. It is quoted as roughly 70% failing to respond to the valuation part within 24 hours.

We went looking for it, and it is not solid enough to use as a headline. The figure appears in one trade article, attributed to a Rightmove spokesperson, with no Rightmove publication behind it that we could find, no stated sample for that specific sub-finding, and no publication date on the page. A separate trade piece dates the underlying mystery shop to 2014, and that piece was written by someone who runs an AI receptionist company, which does not make it wrong but does make it interested.

So: an interesting corroboration, cited as trade reporting of unclear vintage, and not a number we will put in a headline. If it is right, it says the pattern is not peculiar to car dealerships. That is worth knowing and not worth overstating.

Measure your own gap this week

You do not need anybody’s benchmark for this. Take last month’s enquiries and sort them into two piles:

1. Ones answerable from what a person already knows, or from a standard reply.
2. Ones needing someone to look something up, check with a colleague, or make a judgement.

Now calculate the response rate for each pile separately. Most businesses have never seen these two numbers side by side, and the gap between them is usually larger than anyone expects.

If your second number is much worse than your first, more speed will not fix it. What fixes it is making sure the hard question gets answered by someone, even if the answer is “I do not have that in front of me, I will find out and come back to you today”, which takes eleven seconds and is worth more than silence every time.