Marketing ROI: the three numbers that matter

Most marketing reports measure activity - impressions, clicks, followers. Three numbers decide whether the money is working, and you can get all of them from the booking software you already pay for.

1. Cost per booked client

Not cost per click. Not cost per lead. Cost per person who actually sat in the chair and paid. Divide what you spent on one channel in a month by the number of new clients that channel produced.

A salon spending 300 on local ads that produced 12 first-time clients has a cost per booked client of 25. The same salon spending 150 on printed flyers that produced two clients is at 75 - five times worse, and you would not guess it from the flyer. The number only works if you count bookings, not enquiries - a lead that never books is a cost with no revenue.

2. First-90-day value

A new client is not worth one appointment. Take the average ticket, multiply by how many visits a new client typically makes in their first three months, and you have the number you are actually buying.

If the average visit is 60 and new clients come 2.4 times in 90 days, each new client is worth about 144 in the first quarter. Against a cost per booked client of 25, that is a payback ratio of roughly 5.8 to 1. That ratio - not the raw spend - is what tells you whether to increase the budget or fix the offer.

Ninety days is deliberate: long enough to include a second visit, short enough that the data is measured rather than forecast, and it stops you justifying bad spend with a hypothetical lifetime value.

3. Share of revenue from repeat clients

Split last month's revenue into clients who had been before and clients whose first visit was last month. The repeat share is the health check on everything else. Below about 40 percent, you are running a bucket with a hole in it and acquisition spend is mostly refilling it. Above 65 percent, growth depends on how well you reactivate people who drifted, not on how many new faces you can buy.

It is also the number that makes retention work pay for itself: a rebooking prompt at the end of the appointment costs nothing per client and moves repeat share faster than most ad budgets.

Where to get the data

All three numbers come from two exports most booking tools already produce: an appointments report with a first-visit date, and a revenue report for the same period.

Cost per booked client: spend per channel divided by new clients attributed to that channel.
First-90-day value: for every client whose first visit was in a given month, sum their revenue for the following 90 days, then average across that cohort.
Repeat share: revenue from clients with a first visit before the period, divided by total revenue for the period.

If your tool will not export it, tag every new client with the channel in the client note on day one. The export question disappears.

Attribution without a tracking degree

Pixel attribution breaks constantly: iOS privacy settings, cookie blockers, people booking from a partner's phone. The cheapest reliable source is the question you already ask - how did you hear about us - placed in the intake form and answerable in one tap.

Pair it with the channel's own report: Google Business Profile shows calls, direction requests and website clicks from search and maps; Google Analytics 4 shows which pages produce booking-page visits. Neither tells you who paid, so the intake answer stays the source of truth.

Count walk-ins as their own channel, and count referrals separately from social - merged, your best channel looks like it is failing.

What the numbers change

Ratio under 1.5: stop the spend. Either the offer is wrong or the traffic is. Do not add budget to find out which.
Ratio 1.5 to 3: fix the follow-up before scaling. Usually the gap is a second visit that never happens - rebooking and a reminder sequence close most of it.
Ratio above 3: scale until the ratio drops toward 3. That point, not a feeling, is your budget ceiling.
Repeat share falling while acquisition rises: you are buying clients faster than you keep them. Spend one month on reactivation before adding another channel.

The 20-minute monthly routine

First Monday of the month: export appointments and revenue for the previous month, fill three cells - spend, new clients, repeat revenue - and write one sentence about what you will change. One decision per month is enough. Ten dashboards you never read are not measurement, they are decoration.

Keep the last 12 rows in one sheet. Trends beat snapshots: a cost per booked client creeping from 25 to 40 over four months is a warning a single report hides.

The mistake that wastes the whole exercise

Measuring channels you cannot control the volume of. If referrals and walk-ins are 70 percent of new clients, your ad spend is a small lever; the bigger one is making referrals easier - a review request after the first visit, and a reason for the client to mention you. Measure the small lever anyway, so you know when it stops paying, but do not mistake it for the growth engine.

Frequently asked

What is a good marketing payback ratio for a small service business?

Around 3 to 1 on first-90-day revenue. Below 1.5 to 1 the channel is losing money once you account for the cost of delivering the service, not just the ad spend.

Should I use lifetime value instead?

No. Lifetime value is a forecast, and forecasts get optimistic when you are deciding whether to keep spending. Use measured 90-day value and revisit it every quarter.

How do I track marketing ROI if I take cash only?

The same way. Cost per booked client and repeat share do not depend on card payments - the intake question gives you attribution, and a daily cash total in the booking tool gives you revenue.