Back to Insights

Call Tracking and the Admissions Handoff Nobody Measures

Call tracking for treatment centers usually stops at the ring. How to measure the admissions handoff, match inquiries to sources, and keep it compliant.

Illustration of a tracked inquiry passing from a website visit through a phone call to an admissions record

You can almost certainly tell me how many people visited your site last month and how many people were admitted. Call tracking for treatment centers exists to account for everything between those two numbers, and in most programs that middle stretch is unmeasured — the admissions line rang some number of times, someone answered most of those calls, and an unknown fraction turned into a bed. When the budget conversation comes around, nobody in the room can say which calls came from which channel, or what happened to the ones that went nowhere.

That blank space is why marketing arguments at treatment programs go in circles. Without it, every channel gets credit for the admissions that happened and blame for the ones that didn’t, in proportion to whoever is arguing loudest. With it, you can see that one channel produces a high volume of short calls that rarely convert while another produces fewer, longer calls that fill beds. Those are different businesses, and you cannot manage them the same way until you can see them separately.

What Call Tracking Has to Answer

Three questions, in this order. Where did this inquiry come from. What happened during the call. Did it become an admission. Most installations answer the first one partially, ignore the second, and leave the third to a spreadsheet somebody maintains by hand until they get busy.

The mechanical part is straightforward: a distinct tracking number for each channel you want to separate — paid search, your Google Business Profile, the website’s organic sessions, print and outdoor, each referral relationship worth naming — all forwarding to the same admissions line so that nothing about the experience of calling you changes. The caller dials a number, your coordinator picks up the same phone, and the platform records which number was dialed.

What makes it worth doing is the second and third questions, and those are not a software purchase. They are an agreement with your admissions team about what gets written down.

Dynamic Number Insertion Needs a Careful Hand

Dynamic number insertion swaps the number displayed on your website depending on how the visitor arrived, which is how you separate organic from paid without publishing two different numbers. It works well inside a browser session. It causes problems the moment someone treats it as a site-wide replacement for your phone number.

Your number does not only live on your website. It is on your Google Business Profile, your state licensure listing, your accreditation profile, insurer and payer directories, referral partners’ pages, discharge paperwork from hospitals that send you people, and whatever print still circulates. Those off-site listings are part of how local search decides you are a real organization at a real address, so the number of record needs to stay consistent across all of them. A few practical rules follow from that:

  • Keep one canonical number everywhere off-site, and let the swapping script touch only what renders on your own pages.
  • Give your Google Business Profile its own treatment rather than a swap. Google’s guidance has historically permitted a tracking number in the primary field when the main number is also listed, but policy language changes, so read the current documentation rather than taking a vendor’s summary of it.
  • Size the number pool for your actual concurrent traffic. If two visitors share a tracking number in the same window, the attribution for both is a guess.
  • Check what the caller sees and hears. Forwarding can mangle caller ID, add a ring delay, or break the path to your after-hours answering service, and all three cost you calls while looking fine on a dashboard.

The install is the easy part. What separates useful call data from decorative call data is whether the source of an inquiry survives the trip from the phone into the record where the outcome is written.

Matching an Inquiry to Its Source After the Phone Goes Down

The tracking platform knows where the call came from. Your CRM or admissions log knows what happened to it. If those two never touch, you have two half-measurements and no answer.

The fix that holds up is writing the source onto the record at the moment the record is created, automatically, rather than asking a coordinator to remember and type it. Admissions staff are mid-conversation with someone who is frightened or exhausted; attribution is not their job and they will not do it reliably, nor should they be judged for that. Where a platform integration can push the source into the inquiry record, use it. Where it can’t, the fallback is a single required field with a short, closed list of options — not a free-text box that fills up with twelve spellings of the same thing.

Then accept that the match will never be perfect, and decide in advance how you’ll handle the honest ambiguities. A family member calls from a number nobody has seen before, on behalf of the person who filled out the form two days ago. An inquiry goes cold for three weeks and comes back through a different channel. Insurance verification stretches the gap between the call and the decision past the window your analytics tool is willing to look at. None of these are reasons to abandon the measurement. They are reasons to report ranges rather than decimal points, and to treat a channel’s pattern over a quarter as the signal rather than any single week.

Recording and Retention Are a Compliance Question

Everything above can be done with metadata: which number was dialed, when, for how long, whether it was answered, and a flag for the outcome. The moment you add recording, transcription, or any analysis of what was said, you are handling some of the most sensitive information your program touches, and you have left marketing’s lane entirely.

That decision belongs to your compliance officer and your own counsel, before anything is switched on. The questions they will want to answer include what the consent requirements are in every state you take calls from, whether your tracking vendor will sign a business associate agreement and what their subprocessors do with the data, how long recordings and transcripts are retained and who can retrieve them, and what leaves your environment when conversion data is passed back to an advertising platform. An agency that enables recording on your account because it makes their reporting look better has made a decision it had no standing to make. Ask what is being captured, where it sits, and for how long — and if the answer comes back vague, that is itself the answer.

The Handoff Numbers Worth Keeping

Once source and outcome are joined, a short list of measures tells you most of what you need. Answer rate and speed to answer, because an inquiry that rings out is a dollar spent on nothing. After-hours volume and what actually happens to those calls. Abandoned calls, counted separately from unanswered ones. The share of inquiries by channel that your own team considers a fit, which is the honest way to tell a good lead source from a loud one. Call-back latency on the ones that leave a message. And finally, inquiry-to-admission rate by source, which is the number that lets you compare channels on the basis of beds rather than volume.

Start with the channels you are actually paying for, since that is where the decisions are, and get the source field onto the record before you add anything else. Once the join holds for a quarter, the rest of the work gets easier: you can run the arithmetic in how to reduce your cost per admission with real inputs, and the fixes in admissions conversion optimization stop being hunches. If the picture that comes back is that the calls simply are not happening, the diagnostic in how to tell why your marketing isn’t working is the better starting point, and it leans heavily on whether your Google Business Profile is doing its share of the ringing.

Most programs discover something uncomfortable in their first clean quarter of call data — a channel they defended for years, or a gap at seven in the evening that nobody had looked at. That discomfort is the point. You cannot fix a handoff you have never watched.