Your Google Ads cost per booking should fall over time, not sit still

Here is the mental shift I want you to make. A cost per booking that has not moved in six months is not stable, it is neglected. When an account is genuinely being worked, the line on the chart trends down. New campaigns are noisy and a little expensive while they learn. Mature campaigns, the ones that have been pruned and tested and fed good data, settle into a lower number inside the healthy band. If yours has been frozen at one figure since the day it launched, nobody is touching the levers.

I have written before about why cost per booking varies from one clinic to the next, and a lot of that is structural. Your catchment, your competition, your booking page. But there is a second axis that has nothing to do with structure and everything to do with maintenance. Two clinics with near identical setups can end up paying very different amounts, simply because one account is being worked on a tight cadence and the other is on autopilot. I unpack the structural side in why cost per booking varies.

The reason this matters financially is that the cost per booking is only half the equation. It is meaningful when you pair it with what each new patient is worth to you. A falling cost per booking on a stable patient value means your return is quietly compounding every month. That is the whole game.

Let me set the expectation clearly so you have something to measure against. A healthy physio or podiatry account in Australia sits around $80 to $100 per new patient booking. A new account tends to land near the top of that band, around $100, once it is past the first noisy fortnight. That is a normal starting point, not a problem. It is the price of a campaign that is still learning which searches convert and which ones waste money.

A well worked account, one that has been through months of disciplined pruning and testing, trends down toward the lower end of that band, closer to $80, with a tight catchment and a clean booking page. The accounts I worry about are the neglected ones running well past $100, sometimes north of $200, because the campaign is chasing the wrong words. The gap between a $100 account and an $80 one is not luck. It is the direct output of the work below, applied week after week.

If you want the full picture of how an account behaves in its early life, I have laid that out in the first 90 days of a Google Ads account. The short version. Do not judge a campaign on its opening number. Judge it on its trajectory.

The single biggest leak in most clinic accounts is spend going to search terms that were never going to book. Someone searches for a free service, a job, a competitor, or a condition you do not treat, and Google charges you for the click. Left alone, this waste accumulates quietly and props your cost per booking up by twenty or thirty dollars without you ever seeing why.

My rule is a negative keyword review every 72 hours on an active account. Not monthly, not when someone remembers. Every three days I am looking at the actual search terms that triggered ads and cutting the ones that have no chance of becoming a patient. The reason for the tight cadence is simple. The earlier you kill a wasteful term, the less you pay for it. A bad search term caught on day two costs you a couple of clicks. The same term caught a month later has quietly burned a chunk of your budget.

This is the unglamorous, repetitive work that separates a maintained account from a neglected one. It is also why doing your own marketing is rarely free. The review is not hard, but it has to actually happen, and it has to happen forever.

The second lever is ad testing, and the discipline here is about how you decide, not just that you test. Most people run two ads, glance at them after a few days, and switch off the one that looks worse. That is gut, not data. Early numbers swing wildly, and a losing ad on day three is often the winner by day twelve.

I judge split tests on a fixed window of around 14 days. The ads run, I do not touch them, and at the end of the window I let the data make the call. A fixed window stops two failure modes at once. Killing a good ad too early on a bad few days, and letting a genuinely weak ad limp along for months because nobody set a deadline. The window forces a decision, and the decision is based on enough volume to mean something.

Every test that resolves teaches the account something durable about the language your patients respond to. This is closely tied to the difference between a conversion and a conversion action. If you are testing ads against the wrong event, you will optimise toward the wrong winner. Make sure the thing you are counting is an actual booking before you let it decide which ad survives.

The third part is the one a single clinic running its own ads can never replicate, and it is why the cost per booking keeps drifting down long after the obvious wins are gone. A solo account learns only from its own spend. An account that sits inside a larger pool of comparable clinics gets to learn from far more data than its own budget could ever buy.

When a symptom led search term proves itself across a group of physio accounts, that pattern can be put to work faster everywhere else, because it has already shown it has legs. A negative keyword that wastes money in one account can be pre empted in the next. I have written about why that pooled data beats a solo account or a generalist in the cross fleet data network effect, so I will not relitigate the whole argument here.

The point for this article is narrower. It is one more reason the number should keep falling. The first two levers are things a disciplined owner can run themselves. This one is the difference between optimising in the dark and optimising with the lights on, and it is the part I am honest about not being able to hand a single clinic to do alone.

Put the three together and the trajectory takes care of itself. The 72 hour review stops the leaks before they cost you. The 14 day test window keeps your ad language improving on evidence instead of impulse. And the account is not guessing in isolation. None of these is clever on its own. The compounding is what makes the cost per booking fall through its band.

If you only take one thing from this, let it be the expectation. Stop accepting a flat number. Ask whoever runs your account what changed this week, what got cut, what is being tested, and what the trend line is doing. If the honest answer is 'nothing, it just runs', that is your sign. A well run account is never finished, and it should be a little cheaper next month than it is today.

If your account has been sitting at the same number for a while and nobody can tell you why, that is exactly what an audit is for.