The Five Numbers a Small Business Should Actually Track

Marketing dashboards have a way of growing. Someone adds impressions because it was available. Then reach, then follower count, then sessions, then bounce rate, then engagement rate. Eighteen months later there is a monthly report with forty numbers in it and nobody in the room can say whether things are going well.

More measurement is not better measurement. Most of those numbers describe activity rather than outcome, and a business owner reading a report full of activity metrics is being told what happened, not what it was worth.

Five numbers cover it for almost any small business. Here they are, why they matter, and what they tell you when they move.

1. Leads, by source

How many new inquiries arrived this month, and where each one came from.

Not sessions. Not impressions. A lead is a person who took a real action: called, filled out a form, sent a text, booked, or walked in.

Why it is first: everything downstream depends on it, and almost nobody counts it accurately. Businesses track form submissions because those arrive as email, and lose phone calls entirely because nobody logs them.

The hard part is attribution on calls. If you cannot tell which channel produced a call, you are making budget decisions on partial information. Call tracking solves this, and even a simple version beats guessing.

What it tells you when it moves: if leads drop while nothing else changed, look at visibility first. Rankings, ad delivery, a Business Profile issue, a broken form. If leads rise and revenue does not, the problem is downstream and you will see it in the next number.

2. Conversion rate, from lead to customer

Of the people who contacted you, what percentage became paying customers.

Why it matters more than lead volume: this is the number that determines whether more marketing spend makes sense. A business converting a high share of its inquiries can profitably buy more of them. A business converting a low share is buying the right to lose more people.

It is also the cheapest number to improve. Moving conversion from a quarter to a third of inquiries has the same revenue effect as increasing lead volume by a third, and it usually costs a fraction as much because the fix is process rather than spend.

Track it separately by job type where the work differs meaningfully. Emergency service and planned installation convert at completely different rates, and averaging them together produces a number that describes neither.

What it tells you when it moves: a sudden drop usually means an intake problem rather than a lead quality problem. Someone left, someone stopped following up, a phone line stopped being answered. Check that before concluding the leads got worse.

3. Average customer value

What a customer is worth. For transactional businesses, the average sale. For anything with repeat purchase or recurring work, what a customer is worth over the relationship.

Why it belongs here: without it, nothing else can be interpreted. A lead is not worth a fixed amount. It is worth something only in relation to what a customer produces.

This is also the number that changes strategy most often when a business finally calculates it. Owners frequently discover that a customer segment they had been treating as secondary is worth several times another, and that they have been spending equally to acquire both.

What it tells you when it moves: a falling average with steady volume usually means your marketing has drifted toward a lower-value job type. That can be deliberate or accidental, and it is worth knowing which.

4. Cost per customer acquired

Total marketing spend divided by new customers. Spend means everything: ad budget, agency fees, software, and the value of time spent on it internally.

Why not cost per lead: cost per lead is the most commonly quoted number in marketing and one of the least useful on its own. A cheap lead that rarely converts is expensive. An expensive lead that converts half the time is cheap. Only the customer-level number tells you which you have.

Compare it against number three. That comparison is the entire question. If a customer is worth substantially more than it costs to acquire one, spend more. If the gap is thin, fix conversion or average value before increasing spend.

Track it by channel once you have enough volume, because channels rarely perform equally and the differences are usually larger than anyone expects.

5. Repeat and referral rate

What share of your business comes from existing customers returning or sending someone.

Why it is the most neglected number on the list: it costs nothing to acquire these customers, they tend to convert faster, and they are usually worth more. Yet nearly all marketing attention goes to the top of the funnel.

A business where a healthy share of revenue is repeat and referral has a fundamentally different economic position than one running entirely on new acquisition, even at identical revenue. The second one has to buy its whole revenue base again every year.

What it tells you when it moves: a falling referral rate is an early warning that shows up before revenue does. Something about the experience has changed, or you have stopped staying in contact after the job. Both are fixable and both get expensive if ignored.

What to stop reporting

Being direct, because these consume attention.

Impressions and reach. They measure how many times something was displayed. They do not measure whether anyone cared.

Follower count. Interesting. Not a business metric. Businesses with modest followings routinely outperform businesses with large ones.

Rankings alone. Rankings are a means. If rankings improved and leads did not, the ranking is for a term nobody valuable is searching. Report rankings as context alongside leads, never instead of them.

Bounce rate, in isolation. A visitor who found your phone number and called has bounced. That is a success recorded as a failure.

Email open rates. Increasingly unreliable because of privacy features that pre-load images. Look at clicks and replies instead.

Posts published and blogs written. Activity, not outcome. This is the metric agencies report when the outcomes are not there, and it is worth noticing when a report leans on it.

How to actually get the numbers

The obstacle is rarely analytics. It is that leads arrive through five channels into four places and nobody assembles them.

A minimum viable setup: every inquiry lands in one system with a source attached, call tracking so phone leads are attributable, a pipeline with stages so conversion is visible, and job value recorded against the customer record. That is what a CRM does for a small business, and it is most of why one is worth having.

If that is further than you want to go today, a spreadsheet with five rows updated monthly still beats a forty-metric dashboard nobody reads. The value is in the consistency, not the sophistication.

The one that matters most

If you can only maintain one, track leads by source with conversion attached. Everything else can be reconstructed later. That one cannot, and without it every marketing decision you make is a preference rather than a judgment.

We build reporting around these numbers rather than around whatever a platform happens to export, and the accounts and the data belong to you. Get in touch if you want help working out what you can currently measure and what you cannot.