Healthcare SaaSApp Acquisition Signal Architecture

Content was driving acquisition. The app was converting downloaders to subscribers.
Building the analytics infrastructure that traced which channels, content, and creatives drove each outcome, at the subscription level.

Client
Your Doctors Online
Vertical
Healthcare SaaS
Core problem
Web session, app install, and subscription event in three separate measurement systems with no identity connection
Signal environment
Mobile web · iOS/Android app · Subscription billing · Paid media stack · Organic content
The architectural gap
Web session, app install, and subscription event in three separate measurement systems with no identity connection
What changed
Persistent user identity across web and app; BigQuery attribution layer; subscription signal routing to paid media platforms

Your Doctors Online (YDO) is a subscription telehealth platform that connects patients with licensed physicians through a mobile app.

The acquisition model is content-first: YDO produces health information that ranks in organic search, drives mobile web sessions, and converts a portion of visitors to app downloads. From within the app, a portion of users subscribe to a paid plan.

The content attracted the right audience. The app converted a meaningful share of downloaders to subscribers. The organic and paid acquisition channels were generating real volume. Each measurement system in place (web analytics tracking content sessions, Firebase tracking app events) was appropriate for where the business was. The limit the architecture had reached was that three commercial outcomes were happening in three separate measurement environments with no identity thread connecting them: organic web visit, free app download, and paid subscription.

Content and paid media investment could be measured against download volume, but not against subscription revenue.

A user who found YDO through a paid social creative on mobile, downloaded the app, and subscribed three days later was three separate records: a web session, an app install attributed to a media source, and a subscription event with no acquisition context. The Google Analytics 4 (GA4) web session credited the channel for the visit. Firebase attributed the install at the device level. The subscription event in the app carried no reference to the content piece or campaign creative that drove the original web visit.

The result was that content and paid media investment could be measured against download volume, but not against subscription revenue. A creative that drove 10,000 downloads and 50 subscribers was indistinguishable from one that drove 10,000 downloads and 500 subscribers, because the subscription attribution was not in the data. The same problem applied to organic content: an article generating 50,000 sessions had no measurable path to how many of those sessions eventually became paying subscribers.

Paid media bidding optimized toward download volume by default, because subscription revenue was the outcome that mattered commercially but the signal that was not reaching the platforms.

Building the analytics infrastructure

Persistent user identity across web and appA custom user identifier set at the web session and carried through the app's install attribution layer, so the app install and all subsequent in-app events, including the subscription start, can be joined back to the originating web session and acquisition context.
BigQuery as the attribution warehouseGA4 web session data, Firebase app events, and subscription billing events joined in BigQuery by user identifier. The warehouse layer that holds the complete acquisition-to-subscription view for every user, traceable to channel, campaign, content piece, and creative.
Subscription vs download attribution modelsSeparate conversion event architecture and attribution logic for free downloads and subscription starts. The two outcomes are attributed independently, because the channels and content formats that drove each often differed significantly. Reporting that shows both together and separately.
Content-to-subscription attributionOrganic web sessions tagged with content-level parameters that persist through the app install attribution chain. Each article's contribution to subscription revenue measurable directly, with editorial investment evaluated against subscriber yield rather than session volume.
Paid media subscription signal routingSubscription conversion events with revenue value forwarded to Meta and Google via their respective conversion APIs at the app level. Bidding algorithms given the subscription signal rather than the download proxy, so media optimization targets what the business actually needs to measure.
Analytics infrastructure and analysis engineThe BigQuery models and reporting layer that give the growth and editorial teams access to the attribution data at the channel, content, and creative level, maintained operationally rather than produced as a one-time export.

Paid media optimization shifted from download-volume targets to subscription-revenue targets.

"They always help us focus and point us towards that North Star. Our teams have been aligned from the start, and they work to keep us on track."
— Raihan Masroor, Co-Founder, Your Doctors Online

Media allocation changed at the channel and creative level based on what the attribution model showed.

The creative-level view of subscription vs download attribution revealed consistent differences in format performance: content formats designed around symptom and condition queries drove higher subscription yield than content designed for high-volume health information searches, even when download volume was similar. Creative investment shifted accordingly.

SEO and content investment became measurable against subscription revenue. The editorial team could see, for the first time, which topics and content formats were producing subscribers rather than just traffic.

The business could answer the question the growth strategy depended on: which channel, which content, and which creative drives subscription revenue, as a distinct question from which drives downloads. The two signals, once separated, produced materially different allocation decisions.

If paid media is optimizing toward downloads when subscriptions are what fund the business, the attribution layer is where the conversation starts.

The Assessment maps the signal gap between the outcome you're measuring and the outcome that matters.

Start here

The Measurement Architecture Assessment — a 30-minute call, a fixed-fee diagnostic, a clear picture of where your signal layer breaks and what to do about it.