Business confidence just hit a five-month high. That is a survey, not your order book.

City A.M. reports that business confidence has climbed to its highest level since March. The figures come from the Lloyds Business Barometer, published by Lloyds Banking Group: overall confidence up four points to 53%, economic optimism up seven points to 49%, and the trading outlook up two points to 58%, a three-month high. Sixty-four per cent of bosses described themselves as optimistic against 15% pessimistic, and two-thirds expect higher output over the coming year.

Amanda Murphy, chief executive of Lloyds' commercial arm, said businesses are reporting stronger customer demand, greater optimism about the wider economy and growing confidence in their own trading outlook. The report attributes the improvement to consumers who have carried on spending despite higher energy prices, and to an unusually quiet run-up to the Budget — Chancellor John Healey has kept the October fiscal event under wraps rather than letting it leak through the summer, which is a break from recent years.

It is a genuinely good set of numbers. It is also, and this matters more than the coverage suggests, a set of numbers about what people expect rather than what they have done.

What a confidence index actually measures

The Barometer asks business leaders how they feel about their prospects. That is a legitimate and useful thing to measure — expectations drive hiring and investment decisions, and hiring and investment decisions drive the economy. Sentiment is not fluff.

But it is a leading indicator built from opinions, and opinions move faster than order books. A confidence reading rising four points tells you that the mood among the several hundred firms surveyed improved in August. It does not tell you that anybody's revenue improved in August, and it is silent on whether yours will.

A survey tells you what a room full of other people expect. Your own three-month enquiry count tells you what is actually happening to you. Only one of those is worth acting on.

The gap matters because of how the number gets used. A confidence headline is comforting when things are hard and reassuring when things are fine, and in both cases it invites you to do nothing. The right response to a national sentiment reading is not to feel better or worse. It is to go and check the same question in your own business, where the answer is knowable.

The three numbers that answer it for you

Take an hour this week and pull three figures. None of them requires an accountant, and all three are usually sitting in a system you already pay for.

Enquiries, by month, for the last six months. Not leads-in-the-CRM-with-a-status, just the raw count of people who got in touch wanting to buy something. If that line is rising, the confidence headline is true in your market. If it is flat while the national mood improves, the improvement is happening somewhere that is not your customer base — which is worth knowing before you hire against it.

Conversion rate on quotes, same six months. Rising enquiries with falling conversion is the signature of a market where people are shopping rather than buying. It looks like growth in the top line of your pipeline and lands as wasted quoting time. It is also the earliest warning you get that your pricing has drifted out of line with what competitors are doing.

Forward-booked revenue, as a number of weeks. How many weeks of work do you have committed right now? Compare it to the same point last quarter. This is the single most honest number in a small business, because unlike a pipeline it cannot be inflated by optimism. Confidence surveys and forward bookings pointing in opposite directions is a signal, and the bookings are the one telling the truth.

The number underneath the good news

There is one figure in the Barometer that does not fit the headline, and it is the most useful thing in the report. The share of firms expecting to raise their prices fell three percentage points, to just over half.

Read that next to the rest of it. Demand is holding up. Optimism is rising. Energy costs are higher. And fewer businesses now intend to put their prices up than did a month ago.

Those things only reconcile one way: firms are choosing to absorb cost rather than pass it on, because they are worried about what happens to volume if they do not. That is a margin story dressed as a confidence story. Output can rise, optimism can rise, and profit per job can fall at the same time — and the businesses that get hurt in that combination are the ones that read the confidence number, assumed the good times had arrived, and added fixed cost.

It is the same trap that catches people every cycle. Busier is not the same as better off. A year of record turnover on a compressed margin buys you more work, more risk and more working capital tied up, in exchange for roughly the same money.

What to actually do with a good confidence reading

Three things, in order.

First, use it to test rather than to relax. Run the three numbers above. If they agree with the survey, you have external confirmation of something you already knew, which is worth having. If they disagree, you have found out early, which is worth much more.

Second, look hard at your own pricing before you join the half of the market that has decided not to move. If your costs have risen and your prices have not, you have already taken a pay cut; you have just not written it down anywhere. Working out your gross margin by job type this month is a more useful hour than reading any survey, including this one.

Third, be careful about what you commit to on the strength of sentiment. Confidence is cheap to change and staff, leases and finance agreements are not. If you are going to hire or sign on the back of an improving outlook, do it against forward-booked work rather than against a mood — yours or the country's.

The Budget on 28 October will move all of these numbers again, in one direction or another, and nobody currently knows which. That is an argument for making the decisions that do not depend on it — pricing, margin, the enquiry-to-quote process — in the eight weeks before it, rather than waiting to be told what the weather is going to do.