Every owner we talk to has a dashboard somewhere. Clicks, impressions, leads, cost per lead, maybe a ROAS figure that came out of an ad platform. They can read all of it and still not answer the only question that matters on a Monday morning: where does the next dollar go.
That is not a data problem. Most local businesses have plenty of data. It is a decision problem. You are looking at numbers that describe activity, and you need one number that ranks choices.
Cost Per Lead Is the Wrong Ranking
Cost per lead is seductive because it is easy to get and it is always available. It is also the number most likely to make you kill the wrong campaign.
A lead is not an outcome. It is a raw material of wildly varying quality. One channel hands you twenty cheap form fills from people comparing prices across four states. Another hands you six expensive ones from homeowners who already know what they want and have a date in mind. On a cost per lead board, the first channel wins and the second one gets cut. Then revenue drops and nobody can explain why.
Rank on cost per booked appointment instead. It is harder to calculate, which is exactly why it is worth having. The moment you do, the board reshuffles, and the channels you were about to defund start looking very different.
The Numbers Worth Tracking
You need fewer than you think. If you can answer these four, you can make every budget decision without opening an ad platform.
- Cost per booked appointment, by source. Total spend on a channel divided by appointments that actually got on the calendar from it. This is your ranking number. Everything else is diagnostic.
- Lead to appointment rate, by source. This tells you whether a channel sends buyers or browsers. Green Forever Arizona's visualizer has tracked 475+ leads at a 27.6% lead-to-appointment rate, and knowing that specific rate is what makes their spend decisions obvious instead of debatable.
- Show rate. Booked and showed are different businesses. A source with a great booking rate and a bad show rate is a confirmation and reminder problem, not a marketing problem, and no amount of budget will fix it.
- Revenue by source, eventually. The full path, from ad to lead to appointment to signed job. Green Forever did $703K in client revenue in July 2026 on a stack instrumented this way. The point is not the figure, it is that the path was connected end to end so the figure could exist at all.
Four numbers. If your reporting cannot produce them, that is the project, not a bigger ad budget.
The Channel That Looks Dead Usually Is Not
Here is the failure we see most often, and it is expensive because it feels like discipline.
An owner pulls the report, sees a channel with almost no revenue attached, and cuts it. But the revenue was there. It just lost its name somewhere between the click and the close. The customer saw the ad on Monday, thought about it, and typed the business name into Google on Thursday. Or the booking came through a tool that writes every appointment down as self-booked, or direct, or blank. The origin got stripped, and the line that was actually working looked like a line that was doing nothing.
We have watched this happen inside real accounts. A source that appeared to be producing nothing was producing the best return in the business, because most of its bookings were landing under a generic label. Before you kill a channel, go find out what the untagged bookings actually were. Match them back by name, by phone number, by address. That reconciliation is unglamorous and it regularly changes the answer.
Hold the Window Long Enough to Mean Something
The other quiet killer is the short look. Two weeks of data on a local business is mostly noise. Payroll timing, a rainy stretch, one big job closing on the wrong side of a month boundary, and your channel rankings flip completely.
Pick your window before you look at the number, not after. Compare like periods. When we review a division for a client, we work off a year to date view rather than a month, because a short window on a seasonal local business is close to a coin flip, and a coin flip dressed up in a chart is worse than no chart. TurfSpa's 34.7% returning customer rate is the kind of number that only becomes real over a long window, and it changes how you value every new customer you buy.
This Week
Take your last thirty booked appointments and write the source next to each one by hand. Not from a dashboard, from the actual records. However many you cannot source, that is your measurement gap, stated honestly. Then divide last month's spend per channel by the appointments you could confidently attribute to it, and see whether your budget matches your ranking. Most of the time it does not, and the fix costs nothing but the decision.