Count Qualified Calls, Not Calls

Working with a residential substance use disorder treatment provider, I heard the CEO sum up a frustrating week in two sentences:

“We got several calls, but of them two were viable… The revenue doesn’t justify the ad strategy.”

The problem was real. The cause wasn’t what it looked like. “Several calls” sounds like activity. Two viable calls is the actual result. And for a while, the ad platform couldn’t tell the difference.

Why is “calls” the wrong number?

In addiction treatment, a raw call count includes:

• family members researching options for someone who isn’t ready yet
• people who need a level of care you don’t provide
• people whose insurance you don’t accept, and who can’t self-pay
• callers from a state you can’t admit from
• alumni and current patients’ families calling about something else
• vendors, lead brokers, wrong numbers and hang-ups

Every one of those shows up as a “conversion” if you let it. Your campaigns then do exactly what they’re told: find more people who make phone calls. That’s not the same as finding more admissions.

Step one: what counts as a qualified call?

Before you can count qualified calls, marketing and admissions have to agree on what one is. The best definition I’ve seen came straight from a good admissions team. It covered four things:

• Can they pay? Is the insurance accepted, or is there a realistic ability to self-pay?
• Are they a clinical fit for this level of care?
• Can we admit them? Are they within the geography you serve?
• When is a lead closed? In this case, six attempts over ten days before a lead is marked closed.

Write it down, and make sure the people answering the phone apply it the same way every time.

Step two: is Google actually hearing it?

This is where it quietly goes wrong.

Most call-tracking platforms only send calls marked as “converted” back to Google Ads. If nobody is marking them, or the setting that passes them along is switched off, Google’s bidding has no conversion data at all. It spends your budget with no feedback on which searches and ads produce admissions.

The fix is unglamorous: a named owner for tagging calls every day, a working sync to Google, and switching the campaign’s goal from any call to qualified calls. Once the account has the right signal, it has something real to learn from.

Step three: can the definition be too strict?

A common shortcut is to use call length as a stand-in for quality. The provider had been counting calls over one minute. To improve lead quality, they raised the bar to three minutes.

It sounded sensible. It backfired. The CEO:

“Our admissions took a huge hit in June because we tried this strategy, did not work out.”

A stricter definition means fewer conversions for the algorithm to learn from. Starve it of data and it stops finding the people you want. They went back to the old threshold.

The better answer came from looking at real calls instead of guessing. An analysis of actual call outcomes found that a cutoff around four minutes kept roughly three-quarters of genuinely viable calls while filtering out about three-quarters of the non-viable ones. That is the kind of evidence you want before you change what the algorithm is optimizing for. And even then, change it gradually and watch admissions, not just leads.

Watch for the same mistake in your reports

The same problem hides in reporting. If your “cost per lead” or “cost per appointment” counts callbacks, reschedules and family follow-ups as new inquiries, the number will look better than it is and week-to-week comparisons stop meaning anything. Keep new inquiries separate from everything else, and report cost per admission alongside cost per lead.

A checklist you can use this week

1. Write down your definition of a qualified call, covering ability to pay, clinical fit, geography and when a lead is closed. Get admissions and marketing to sign off on it together.
2. Name the person who tags calls, and spot-check the tagging daily for the first month.
3. Confirm the conversion signal is reaching your ad platform. Don’t assume. Look for conversions in the account itself.
4. Optimize to qualified calls, not all calls, and not an arbitrary call length.
5. Change thresholds with evidence. Pull a sample of real calls before you tighten anything, then change it in steps.
6. Report cost per admission, not just cost per lead.

Your ad platform is only as smart as the signal you give it. Give it the right one.

Not sure your campaigns are learning from the right signal? Book a free 30-minute call.