The Mapping Work Behind a CDP Rollout

A national healthcare network can have hundreds of locations and still struggle to answer basic questions about customer behavior. In this engagement, patient records lived in an electronic health record system, a Snowflake warehouse was being developed, and marketing activity was distributed across advertising, call tracking, and website analytics systems.

When I needed to understand the relationship between calls and patient visits, I requested specific fields and date ranges from the business intelligence team. I then cleaned the extracts and reconciled duplicate records with legitimate repeat visits. That process produced useful analysis, but it required considerable manual work each time.

Why I recommended a CDP

I began discussing a customer data platform with marketing leadership around April or May 2024. My broader interest was making customer information easier to connect, interpret, and use in marketing. Over the following months, I explored approximately eight vendors and attended vendor events to speak with existing customers.

The incoming CTO helped move the initiative forward in summer 2025 and led negotiations with Freshpaint. I took ownership of the implementation around August or September. Healthcare privacy requirements shaped the implementation. The first phase focused on website measurement and Google Analytics, within systems the organization largely controlled. The team wanted to limit disruption heading into its busy fall and winter period.

One clinic could have several identities

The same clinic might appear under its street name in one system, its city in another, and its franchise owner in a third. Acquisitions and ownership changes added historical names. Each location also had analytics destinations to account for. Some had corporate, agency, and owner GA4 properties, alongside overlapping Google Tag Manager containers.

I mapped more than 400 locations by address, aliases, ownership, operator group, agency, and measurement identifiers. The website events were standardized across the network. The difficult work was establishing which location each identifier represented and where its events needed to go.

Validation required the agencies and owners

For six weeks, I met with agency leaders to validate the map, coordinate access to their GA4 properties and GTM containers, and explain the migration. Owners were included in email communications. I also offered office hours for franchisees who wanted to discuss the details or understand the effect on their reporting.

We piloted the implementation on corporate-owned locations. Testing surfaced incorrect location-to-property routing, unexpected event counts, and overlapping tag containers. We corrected routing, updated regular expressions used in the configuration, documented issues, and moved toward a centrally managed GTM deployment.

The franchise rollout then proceeded by agency group. In live walkthroughs, agencies could visit a clinic page while I observed the corresponding event in Freshpaint. We also checked downstream GA4 reporting. Those sessions gave agencies a practical way to understand what they needed to monitor.

Connecting scheduling and check-in data

The implementation also included Solv Health booking and check-in data. My responsibility for that integration was project management: facilitating communication and maintaining momentum between the two vendors. Solv and Freshpaint engineers handled the technical build, including new payloads and pipelines that organized the data for ingestion.

The check-in field was included in the CDP integration. That extended the available information beyond website activity and scheduled appointments to recorded clinic arrival. Check-in and completed treatment remained distinct outcomes.

What the first phase delivered

The first phase established a common website measurement framework, reconciled analytics destinations across the network, and connected Solv booking and check-in information. My responsibilities later shifted toward troubleshooting and further platform development, including a separate consent-management modernization project.

Marketing subsequently pursued Google Ads and Meta activation, while Freshpaint data also fed the broader Snowflake and BI environment. Those later workstreams had their own owners. The contribution I can clearly describe is the implementation foundation: mapping the network, coordinating its stakeholders, validating the migration, and resolving issues before expanding deployment.

For an owner considering similar work, the lesson is practical. Platform selection needs to be followed by a clear inventory, agreed data destinations, stakeholder access, and a rollout that can be tested. In this engagement, those responsibilities occupied a substantial part of the implementation.