Two findings sit beside each other in the Federal Reserve’s 2026 Report on Employer Firms, and they cannot both be read as good news. Reaching customers and growing sales was the most commonly reported operational challenge. Rising costs of goods, services and wages was the most common financial challenge, and 77 percent of firms reported cost pressure from rising costs, tariffs, or both. (fedsmallbusiness.org)
Read together, those two sentences make a small business’s revenue line hard to interpret. Forty-eight percent of firms said they source at least some inputs from outside the United States, and among those firms, 76 percent passed at least some of their input cost increases on to customers. (fedsmallbusiness.org) For any business that raised prices last year, a rising revenue number is partly a record of that price increase. It is not, by itself, evidence that more people bought anything.
That is the practical problem with picking revenue growth as the number you steer by, and it is the reason this playbook starts somewhere other than the usual advice to choose a headline metric and cascade supporting indicators beneath it. Before a growth metric is worth steering by, it has to clear two much duller tests: it has to measure the thing you think it measures, and you have to be able to count it without buying anything.
What revenue growth measured last year
A metric is contaminated when two different changes move it in the same direction and you cannot tell them apart afterwards. Revenue is the most contaminated number on a small business dashboard right now, because in the current cost environment it moves when you sell more and it moves when you charge more.
Those are opposite situations. Selling more usually means hiring, buying stock, or adding hours. Charging more usually means protecting a margin that costs already ate. An owner who cannot separate them will read a good month as demand and staff up against a price rise, or read a flat month as failure when the customer count actually grew. If you have never written down what your costs actually are, that separation is guesswork, which is the case for a documented cost baseline before any of this.
The fix is not a more sophisticated metric. It is a second number placed beside the first.
The correction that costs nothing
Count the unit that cannot be inflated by a price change. Every business has one, and in most cases it is already written down somewhere you are not looking:
- Trades and home services: completed jobs per week. Not quotes sent, not calls taken. Jobs finished and invoiced.
- Retail and food: transactions per week, which your point of sale already counts, plus items per transaction if you want a second line.
- Professional services: active paying clients this month, counted as a headcount, not as billings.
- Subscription or online products: paying accounts at month end, which is a stock rather than a flow and moves far more slowly than revenue does.
Put that count next to revenue for the same period and the pair tells you something neither number tells you alone. Revenue up and units flat means you repriced. Units up and revenue flat means you grew and gave the gain away, usually through discounting. Both up means demand. Both down is the only one that needs no interpretation.
None of these counts requires new software. That matters more than it sounds, for a reason that shows up on the vendors’ own pricing pages.
What it costs to count something fancier
The advice to instrument a custom metric is usually given without a price attached. It has one, and it recurs every month.
QuickBooks publishes a free tier that gives you one connected bank, two invoices a month and a profit and loss report. Tracking custom performance metrics is not in it. That feature first appears on the Essentials plan at a list price of 85 dollars a month, alongside integrating third-party data into KPIs and dashboards, and customizable dashboards with unlimited custom KPIs sit on the Advanced plan at a list price of 340 dollars a month. Promotional pricing was running at 90 percent off for three months when this was checked. (quickbooks.intuit.com)
Xero prices its plans at 25 dollars a month for Early, 55 for Growing and 90 for Established, each currently offered at a heavy introductory rate for the first six months, and the site states plainly that subscription prices are increasing from October 1, 2026. (xero.com) Wave keeps a genuinely free Starter tier covering unlimited invoices, estimates, bills and bookkeeping records, with its Pro plan at 19 dollars a month or 190 dollars a year. (waveapps.com)
So the honest ladder looks like this. Counting what your books already contain is free or close to it. Counting something custom starts around 85 dollars a month at list price and runs to 340. That is between 1,020 and 4,080 dollars a year, before anyone spends an hour maintaining it.
The point is not that those tools are overpriced. It is that a custom metric is a recurring purchase, and it should be argued for like one. If you cannot name the decision it will change, you are paying a subscription to feel informed.
The part most metric advice gets backwards
Here is the argument this piece rests on, offered as reasoning rather than as a finding, because no survey measures it: for a small business, how often you look at a number matters more than how good the number is.
The decisions an owner actually makes are weekly decisions. Whether to put someone on overtime. Whether to run the promotion again. Whether to take the job that would fill next Thursday. A metric reported monthly arrives after all of those have already been made by default, which is the same failure mode as every decision you have not made running as a default. A crude count you check every Monday changes more decisions than a precise one you review at month end, because it arrives while the decision is still open.
This inverts the usual instinct, which is to spend the available effort on picking a better metric. Spend it on shortening the interval instead. Weekly job counts written on the same page every Monday will outperform a beautifully modelled customer lifetime value that nobody opens until the quarter closes.
Four moves, in order
- Write down the uncontaminated unit for your business from the list above, and where it is already recorded today. If you cannot name where it lives, that is the first thing to fix, and it is usually a report you already have rather than a tool you need.
- Put it beside revenue for the last six months. Two columns, one page. You are looking for the months where the two diverge, because those months are the ones you have been misreading.
- Name the constraint that number will strain if it goes up. More jobs strains scheduling and cash tied up in materials. More clients strains delivery hours. More accounts strains support. Write the constraint next to the metric, because a growth metric with no named constraint is how a business grows itself into a problem.
- Set the interval to weekly and pick the day. One number, one page, same day. Only add a second metric after this one has survived a full quarter of actually being checked.
Move three is the one people skip, and it is the one that keeps this from becoming a growth-at-any-cost exercise. The same logic applies to a single campaign, where counting the real cost rather than just the ad spend is what separates a profitable channel from one that only looks profitable.
Where we land on this
Everything above is either a published figure or a price you can check on the vendor’s own page. What follows is judgment.
A growth metric can be improved in two directions, and only one of them is worth celebrating. Revenue per employee, to take the most common example, goes up when the business sells more and it goes up when the business has fewer employees. A dashboard cannot tell those apart either, and an owner under pressure will find the second one much easier to execute than the first. That is the version where a business gets smaller and files it under efficiency.
So the rule worth adopting is narrow and awkward on purpose: if a metric improves when you remove a person, it is not a growth metric, whatever the spreadsheet calls it. The growth worth measuring is the kind where the same team takes on work the business currently turns down. Pick numbers that can only move that way, and the metric stops being able to argue for the outcome you did not want.
The Growth Metric Selector prompt at BusinessPrompter.com is a reasonable structure for the selection conversation itself. Read it for what it is: it sits in the Innovation and Growth category, its own summary line is “Select the right growth metrics to focus your team”, it is written for founders and product leaders rather than for owner-operators, and it is a Pro prompt behind an upgrade. Use it to generate candidates, then run those candidates through the two tests above, which it does not apply for you.
Open your books to the last six months, write the two columns, and find the month where revenue and units moved in opposite directions. That month is the one you have been telling yourself a story about.
Frequently Asked Questions
What is the best growth metric for a small business?
The one you can already count without buying software, paired with revenue for the same period. For trades and home services that is completed jobs per week, for retail it is transactions per week, and for professional services it is active paying clients. A custom metric first appears on QuickBooks Essentials at a list price of 85 dollars a month (quickbooks.intuit.com), so it should be argued for like the recurring purchase it is.
Why is revenue growth a misleading metric right now?
Because it moves both when you sell more and when you charge more. The Federal Reserve’s 2026 Report on Employer Firms found that 77 percent of firms reported cost pressure from rising costs, tariffs, or both, and that among the 48 percent of firms sourcing inputs from outside the United States, 76 percent passed at least some of those increases on to customers (fedsmallbusiness.org). For any business that raised prices, revenue growth partly records the price rise.
How many KPIs should a small business track?
Start with one uncontaminated count plus revenue, and add a second only after the first has survived a full quarter of actually being checked weekly. The constraint is not how many numbers are useful in theory, it is how many get looked at often enough to change a decision before it is made.
How often should I review my growth metric?
Weekly, on a fixed day. Small business decisions about overtime, promotions and which jobs to accept are weekly decisions, so a number reported monthly arrives after they have already been made by default. A crude weekly count changes more decisions than a precise monthly one.
