Why multi-site attribution is an engineering problem, not an analytics problem
Attribution at a single healthcare location is complicated but manageable. You are connecting a finite set of healthcare marketing activities to a finite patient population. At 50 or 100 locations, the complexity multiplies in ways that cannot be solved by adding more analysts or better spreadsheets. Different markets run different media mixes. Patients interact with advertising in one location’s trade area and book at another. Online-to-offline tracking, connecting a digital ad impression to a completed appointment in a practice management system, requires technology infrastructure that most healthcare software ecosystems are not designed to support natively.
This is a genuine build challenge. The five layers required for full-funnel attribution at multi-site scale are all available as technology. Few platforms have all five integrated consistently across every location because each integration requires ongoing maintenance, and the gaps tend to accumulate quietly until someone tries to run a cross-portfolio attribution report and finds the data is incomplete. Understanding the stack is the prerequisite to building it deliberately rather than discovering the gaps after the fact.
The five-layer attribution stack, and where most platforms are missing layers
Full-funnel healthcare marketing attribution for a multi-location platform requires five layers working together. The first is media tracking: connecting campaign impressions and clicks to identifiable user sessions, which every major ad platform provides but which requires consistent UTM discipline across every campaign and channel. The second is landing page conversion tracking: connecting user sessions to appointment inquiries through form submissions and online booking events. Most platforms have this layer, but inconsistently. Some locations have optimized landing pages, others route traffic to the homepage and lose the conversion signal entirely.
The third layer is call tracking, and it is the one most commonly missing in multi-site healthcare. Dynamic number insertion assigns unique tracking numbers to each campaign and channel, attributing inbound calls to their marketing source at the location level. Without it, the majority of patient inquiries, which in most healthcare specialties still come by phone, are invisible to the attribution model. The fourth layer is CRM or scheduling system integration: connecting tracked inquiries to actual appointments and patient records. The fifth is production data integration: pulling care completion and revenue outcomes from the practice management system and matching them back to the originating marketing activity. This final layer is the most technically demanding and the most financially valuable. It is what makes cost per profitable patient a real metric rather than an estimate.
Milestone one: Connecting media spend to scheduled appointments
The first achievable milestone is media-to-appointment attribution: knowing which channels and campaigns are producing appointment requests, and at what cost. This requires call tracking deployed consistently across all locations and landing page conversion tracking capturing online booking events. When both are in place, the allocation question, which channels are producing appointment inquiries at what cost per inquiry, becomes answerable with data rather than assumption.
Most multi-site platforms can reach this milestone within 60 to 90 days of beginning the build. It is the foundation on which everything else depends, and it is worth reaching before attempting the more complex layers. The most common failure at this stage is inconsistent deployment. Call tracking is active at 80 of 100 locations but not the other 20, which means the 20 without tracking look like low-activity markets when they may simply be unmeasured ones. Consistent deployment across every location, before anything else, is what makes the data trustworthy enough to act on.
Milestone two: Connecting appointments to production revenue
The more valuable and more technically demanding milestone is connecting marketing activity to care plan acceptance and production revenue. This requires integrating appointment data with the practice management system, pulling the production value of completed appointments and matching it back to the campaign that drove the initial inquiry. When this layer is functional, the most financially meaningful CAC metric becomes available: cost per dollar of production by channel, campaign, and location.
The technical challenge here is the gap between the marketing data ecosystem and the clinical data ecosystem. Marketing data lives in ad platforms, call tracking systems, and CRMs. Production data lives in practice management software that was not designed to export to a marketing analytics layer. Bridging that gap typically requires a data integration build, either a direct API connection to the practice management system or a scheduled data export that feeds into a unified reporting environment. This is the integration that most agencies are not equipped to build, and it is the one that most directly determines whether the CMO can report to operating partners in financial terms rather than marketing terms.
The most common failure at this stage is incomplete matching: the integration connects some appointments to their marketing source but not others, because call tracking wasn’t active for those inquiries, or because the patient booked through a channel that isn’t fully integrated. A 70 percent match rate is useful. A 70 percent match rate presented as if it were a complete healthcare marketing attribution produces systematically biased conclusions. Being explicit about match rate, and building toward improving it over time, makes the reporting credible rather than misleading.
What the dashboard shows you when it’s working
With the attribution stack in place, the location-level performance dashboard moves from aspiration to management tool. It shows media spend, appointment inquiries generated, conversion rate, cost per acquired patient, and production revenue attributable to marketing activity, by location and by channel, with trend lines over time and benchmarks across the portfolio.
The first finding when this dashboard goes live is almost always the same: the performance distribution across the portfolio is wider than leadership assumed. The highest-efficiency locations are acquiring profitable patients at two to three times the efficiency of the lowest-efficiency ones, often in markets that look demographically similar on the surface. The second finding is typically a channel that is consuming a meaningful share of the total marketing budget while delivering a disproportionately small share of profitable patient production, invisible in location-level reporting because no single location’s spend on that channel looks unreasonable. Seeing the full picture simultaneously is what makes reallocation decisions defensible. Before the dashboard, those decisions are judgment calls. After it, they are data-driven capital allocation.
“The attribution build is not complete when the technology is connected. It is complete when the data is accurate enough to make a capital allocation decision you would defend in a board meeting.”
What mature attribution looks like at 18 months
The attribution build is not a one-time project. It is a capability that matures over time as match rates improve, integrations stabilize, and the data history becomes long enough to show meaningful trends. At 18 months from the start of a serious build, a well-run attribution infrastructure produces quarterly channel efficiency reports that are reliable enough to drive meaningful budget reallocation, a cost-per-profitable-patient metric that is calculated, not estimated, for at least 80 percent of patient acquisitions, and a location-level performance dashboard that operating partners can read without a marketing translation layer.
That last point is the one that changes the CMO’s position in the organization. When operating partners can see marketing’s financial contribution directly, in the same terms they use to evaluate every other capital deployment in the portfolio, the conversation about marketing’s value stops being a defense and starts being a review. That is the outcome the attribution build is ultimately in service of.
Agency Creative builds the full-funnel attribution infrastructure for multi-site healthcare platforms, including the practice management integrations, call tracking deployment, and reporting layer that most marketing agencies are not equipped to deliver. Let’s walk through what the build looks like for your platform.
Learn how Agency Creative can help boost your brand by calling us at 972.488.1660 or by contacting us online.
See More Work
