Open your Google Ads account, your Facebook Ads Manager, and your analytics tool on the same day and add up the conversions each one claims. The total will almost certainly be higher than the number of new patients you actually saw. Each platform takes credit for what it touched, and nobody checks the sum against your schedule.
This is the attribution problem. It is not a flaw in your marketing. It is a measurement gap, and closing it is worth the effort because it tells you where the next dollar should go.
Why the platforms disagree
Each tool measures from its own vantage point. Google Ads counts a conversion when someone who clicked an ad later takes a tracked action. Facebook may count people who merely saw an ad and later booked. Analytics tools use their own rules for splitting credit between channels. A patient might see a Facebook ad, search your name on Google, click a map result, and call. Four systems can each claim part of that story.
On top of that, phone calls and walk-ins often happen off the web entirely. If your front desk does not record where people came from, the best-performing channel may be invisible.
Decide what you are counting
Before choosing tools, define the unit. For a practice, the useful unit is a booked new patient first visit, and ideally one that showed up. Leads, clicks, and form fills are steps along the way, not outcomes. Keep a clear definition:
- Lead: a call, form, chat, or text from a new person.
- Booked: a first visit scheduled.
- Showed: the person attended.
- Retained: the person continued past a defined point, such as the second visit.
Report each stage, so you can spot where a channel leaks. A channel that generates many leads and few bookings has a conversion problem, not necessarily a traffic problem.
Build a three-layer tracking system
Layer 1: Automated tracking
- Call tracking numbers for each major channel, such as one on the website, one on Google Ads, and one on the Google Business Profile. Many call tracking tools record the call and tag its source. Keep notice and consent requirements in mind when recording.
- Form tracking that captures the source, medium, and campaign with each submission.
- Consistent campaign tags on ad links so every click can be traced to a campaign and keyword group.
- Conversion events set up in your ad platforms and analytics, aligned to the same definitions.
Layer 2: Human-reported source
Automation misses word of mouth, direct visits, and referrals. At intake, ask every new patient a simple question: "How did you hear about us?" Give a short list plus an "other" box. Put it in a required field in your system so it cannot be skipped. Then add a second prompt: "Is there anything you searched for or saw that made you call?" The answers are often illuminating.
Layer 3: A single source of truth
All of this data needs one home. A CRM that connects leads to appointments is ideal. Without it, assign someone to log each new patient's source in a spreadsheet at the time of booking. A weekly twenty-minute reconciliation is better than a perfect system nobody uses.
Reading the data honestly
Expect some disagreement between layers. A patient may say "my friend recommended you" even though a call-tracking record shows an ad click. Both can be true, because people are often influenced by several touches. Record both a first-touch and a last-touch source when you can, then look for patterns rather than assigning exact credit.
Be careful about concluding too much from small numbers. If a channel produced three patients in a month, one extra booking changes the story by a third. Look at quarter-over-quarter trends and avoid declaring a winner on a few data points.
Questions a good report answers
Build a simple monthly table with one row per channel and columns for spend, leads, booked, showed, and the cost per patient booked. Then answer:
- Which channels produced the most booked first visits?
- Which produced the best-quality patients, such as those who stayed in care?
- Where do leads drop between inquiry and booking?
- Which channels have no tracking yet and need it?
- What will we change next month because of this?
Cost per patient booked is more reliable than cost per lead, because it ties spending to a result your schedule can verify. Where you cannot calculate it yet, say so in the report rather than guessing.
Pitfalls to avoid
- Counting leads that are not new patients, such as existing patients, vendors, or wrong numbers
- Mixing time periods, such as comparing this month's spend to leads that booked from last month's work
- Giving credit to all channels equally when your data does not support it
- Ignoring organic and referral channels because they have no ad dashboard
- Changing tracking mid-month without noting it
Share the report with your team in a short monthly meeting. When the front desk sees which channels produce patients, they understand why logging the source matters, and the data gets better. Celebrate improvements in tracking quality as well as in results.
Start small
You do not need a data team. Pick one change this month: set up call tracking, add the intake question, or fix the form source field. Then review one report together at the end of the month. Our PatientSurge® system builds attribution reporting into the practice's CRM so the numbers are ready each month. If you would like help untangling what your current numbers mean, book a strategy call and we will go through your channels with you.
