Case studies / Healthcare
Addiction treatment PPC · Call qualityTwenty-seven calls a week. Ten of them needed treatment.
Most agencies would report the 27. We reported both numbers every week, listened to the calls, and wrote "delete campaign" next to the ones that weren't working, including campaigns we were being paid to run.
An established facility, running on unmeasured traffic.
The Hope Center for Rehabilitation was a drug and alcohol treatment facility in Boynton Beach, South Florida, with an established offline reputation and a steady flow of patients through traditional referral channels.
The facility drew patients from across the country, but the paid search account behind it had no conversion tracking at all, so nobody could say which of that spend was actually working.
A phone call is not a patient.
Call tracking makes a campaign look wonderful. The platform counts a conversion the moment someone dials, and the report shows the number climbing. But in addiction treatment a third of those calls are people the facility cannot help: wrong coverage, wrong level of care, wrong state, or a wrong number entirely. Reporting the call count means reporting nearly three times the real figure.
Twenty-seven conversions. Everyone's happy.
It's what the platform records, it's what the dashboard displays, and it's technically accurate. It also tells the client nothing about whether their admissions team spent the week talking to people they could actually treat.
- Calls reported 27
- Cost per call $60
- Calls listened to None
- What the client learns Nothing useful
Ten qualified. Here's what the other seventeen were.
Every call reviewed and categorised on the weekly report: not qualified, referred elsewhere, coverage mismatch, no response, wrong number. It makes the headline figure smaller and it makes every decision after it correct.
- Qualified calls 10
- Cost per qualified call $161
- Campaigns cut on evidence Yes
- What the client learns Where to spend next
Same budget. Almost nothing to show for it.
The Hope Center for Rehabilitation was a drug and alcohol treatment facility in South Florida with an established offline reputation and a steady flow of patients through traditional referral channels.
The paid search account was spending at roughly the same level it would spend with us, and producing close to nothing measurable.
Four reasons, and the first explains why nobody had noticed.
-
There was no conversion tracking
No call tracking, no form tracking, no connection between spend and outcome. Without it there is no way to know which campaigns work, which keywords convert, or whether the account produces anything at all, and no way to calculate return on any of it.
-
Broad targeting in an expensive category
Loose keywords and match types in a category where clicks cost around twenty dollars. Budget reached people researching addiction, comparing facilities for someone else, or looking for something adjacent entirely.
-
Out-of-state demand is unpredictable
A South Florida facility serves patients from across the country, but willingness to travel for treatment varies enormously by state. Without call-level data there was no way to know which states were worth advertising in.
-
Volume metrics hid the quality problem
Even once calls were being counted, counting them isn't enough. The difference between a call and a call from someone the facility can admit is the whole business, and no platform measures it.
Five steps, starting with measurement.
Everything else was guesswork until the account could tell us what was happening.
Build the tracking before touching the campaigns
Call tracking and form conversion tracking implemented across every entry point, with search term data monitored from the start. You cannot improve what was never measured, and every optimisation decision after this depended on it existing.
For the first time, spend connected to outcome.
Rebuild the account around treatment services
Full restructure with campaigns segmented by service and tightly themed ad groups. Search, call-only, display and remarketing separated so each could be budgeted and judged independently rather than one type quietly absorbing another's spend.
Every campaign type measurable on its own terms.
Split the out-of-state demand by state
Separate call-only campaigns for each state the facility drew patients from, rather than one national campaign. That structure is what later revealed which states were producing calls and which were consuming budget silently.
Per-state performance became visible.
Listen to the calls and classify every one
Each week, every call reviewed and categorised on the report: qualified for treatment, coverage mismatch, referred to another facility, no response, wrong number. Not a sample, all of them.
A cost per qualified call instead of a cost per call.
Cut what the call data condemned
Campaigns that produced clicks but no qualified calls were paused or removed. One state campaign generated nine clicks and zero calls across a reporting week; the recommendation on that week's report read DELETE CAMPAIGN. Another was producing wrong numbers. Both were our own campaigns.
Budget concentrated where qualified calls actually came from.
What the reporting actually looked like.
A representative week, by state. Same ads, same structure, same budget logic, and results that ranged from excellent to zero.
-
FloridaOutpatient PPO enquiry; one not qualified8 clicks · 7 calls 88%
-
New YorkOne referred elsewhere; one callback needed6 clicks · 4 calls 67%
-
New JerseyOne referred to a local provider9 clicks · 5 calls 56%
-
OhioCoverage mismatches12 clicks · 4 calls 33%
-
TexasCoverage mismatch11 clicks · 3 calls 27%
-
PennsylvaniaNo response9 clicks · 2 calls 22%
-
Halfway HouseWrong numbers14 clicks · 2 calls 14%
-
IllinoisDELETE CAMPAIGN9 clicks · 0 calls 0%
-
Week total10 people who needed treatment78 clicks · 27 calls 35%
At roughly $1,611 in weekly spend, that's about $20 a click, $60 a call, and $161 per call from someone the facility could actually help.
Florida converted clicks to calls at 88%. Illinois at zero. Same ads, same structure, and without listening to the calls, both would have looked like reasonable campaigns on a click report.
Why we didn't publish the percentages
We could have led this page with a four-digit percentage increase in conversions. The number exists, and it's technically accurate, because the account had no conversion tracking before we built it.
Improving on something nobody was measuring isn't an achievement worth quoting. Ten qualified calls a week at $161 each, on the same budget that was previously producing almost nothing measurable, is the honest version. It's also the one a treatment centre can actually plan against.
Four things worth taking from this.
Ask what percentage of calls were qualified
If your agency reports call volume, ask how many were people you could actually serve. If they don't know, they aren't listening, and your real cost per lead is several times what they're telling you.
Tracking comes before optimisation
An account without conversion tracking isn't underperforming, it's unmeasured. Any agency proposing campaign work before fixing measurement is planning to optimise toward guesses.
Split geography before you judge it
One national campaign averages a state converting at 88% with one converting at zero. Separate campaigns are what make the difference visible and actionable.
A good agency recommends cuts
If nothing has ever been paused, nothing is being measured properly. An agency paid on spend has no reason to write "delete campaign," which is exactly why it matters when they do.
What went into it.
Six of our services, over two and a half years.
How many of your leads were actually leads?
If your reporting shows calls or form fills without telling you how many were people you could serve, your real cost per lead is higher than you think, often by several times. Give us read-only access and we'll show you the gap. Findings are yours to keep either way.
Read-only access · No obligation · You keep your accounts