The Front Lines
SubscribeContent Studio
Colin Daw.
VP of Growth · Virta Health
Growth-driven leader with breadth of experience across high-impact operating roles (growth/commercial/ops/GM). Combining a mechanical engineering foundation, a Stanford MBA, and a penchant for solving cross-functional problems, I contribute most strongly as an intellectually curious, empathetic leader who elevates and integrates the skills and perspectives on my team. Outside of work, I enjoy spending time with my family and in the outdoors, and I've been a nationally-ranked cyclist with national and regional titles in both Road and Mountain Bike.
Guest
Colin Daw
VP of Growth
Company:
Virta Health
Location:
Aptos, California, United States
Loading episode...
Listen onSpotify

When the easy button isn't easy: How Virta Health rebuilt its member acquisition engine

Colin Daw is a mechanical engineer who never planned to run a marketing function.

That background turns out to matter. When he inherited a plateaued growth operation at Virta Health in mid-2023, he didn't approach it as a marketer. He approached it as an engineer staring at a machine with untested components and no baseline data.

"Marketing is a machine, it's a system, a complex one," Colin said in a recent episode of Unicorn Marketers. "We basically hadn't come anywhere near our ceiling for direct email access. A/B testing was a rarity, let alone a weekly occurrence. And we had barely scratched the surface on direct mail."

That diagnosis — methodical, systems-level — shaped everything that followed.

The constraint nobody had fully solved

Virta Health uses individualized, AI-powered nutrition therapy to reverse metabolic disease, starting with type 2 diabetes. The company launched with a five-year prospective clinical trial and has spent a decade building the clinical proof to match its ambition.

But the business model creates a structural constraint that upstream marketing can't solve.

Virta operates B2B2C. Employers and health plans contract with Virta and sponsor access to the program for their members — meaning eligible members pay nothing. That arrangement gives Virta something most consumer health companies don't have: direct contact information for a highly targeted eligible population. It also creates a gating layer. Before a single message goes out, a client has to grant what Colin's team calls "access" — the data and permissions to reach members directly.

Between a third and half of Colin's 20-person growth team exists specifically to acquire that access, maintaining client relationships and making the case for direct outreach to their populations.

"When we align with a client that it's a shared goal to enroll more folks, it tends to be reasonably easy to get that client to agree to email, mail, SMS — all our best practices," Colin said.

The alignment is genuine — Virta's program saves payers money in health outcomes every month. But the access still has to be sold, client by client, before any channel can perform.

The channel that looked broken before it worked

With access unlocked, Colin started testing everything. The biggest surprise was direct mail.

Early format testing produced a result that cut against instinct. Color brochures — the intuitive choice — hit the junk pile. What actually converted in healthcare was a plain legal-size windowed envelope framed around "important benefits information." No glossy design. No obvious branding. It looked like something from a benefits administrator.

"A lot of people think like a color brochure is going to do great," Colin said. "I think it hits the junk mail pile more often than not."

The measurement problem compounded the challenge. Unlike email, direct mail produces no open rates, no click data, no real-time signal. Recipients don't reliably scan QR codes or enter vanity URLs. "Mail only shows up if people take action and apply," Colin said, "and half the time they don't even scan your QR code or enter your vanity URL that can be tracked." That forced Virta to build time-based attribution models — measuring application spikes in the window following a send — before they could confidently read whether the channel worked at all.

Most teams declare a channel dead before solving the measurement model. Virta kept building.

Over two years, direct mail scaled 50-100x. By 2025 it accounted for 40-50% of all member enrollments and contributed to more than a doubling of Virta's overall enrollment rate. It is now the largest single line item in Colin's budget. Email accounts for roughly the other half of volume. Together the two channels drive approximately 80% of Virta's application pipeline.

Cutting what the business model made structurally unwinnable

While mail was scaling, paid Meta was getting cut.

The problem wasn't creative quality or targeting sophistication. It was HIPAA. Audience list uploads were off the table. Conversion signal pass-back that could identify protected health information was prohibited. Hashed lookalike audiences didn't generate performance. Vendors promising compliant conversion data pass-back and employer-level targeting — identifying who was actually covered for Virta — were tested and fell short.

The root issue Colin identified: there is no Meta algorithm that knows which employer or health plan has sponsored Virta access for its members. Until Virta's coverage extends across a large enough share of the American adult population for broad untargeted reach to work by volume, the channel is structurally mismatched to the business model.

"That's been an assumption we've had to challenge and test and really muddle through for two years," Colin said. "And I could confidently come into 26 with a very low paid budget and double down on other channels."

Two years of testing to confirm a structural constraint is an expensive lesson — and a precise one.

Reframing the competitive threat

When GLP-1 weight loss drugs went mainstream, the initial read inside Virta was close to alarm. A program built on sustained behavior change suddenly had to compete against an injectable producing dramatic results with minimal apparent effort.

"You'd think we're screwed, right? You really would," Colin said.

The picture looked different from the B2B side. Self-insured employers, ASOs, health plans, and government payers including Medicare were watching GLP-1 drug costs climb with visible concern. "They're all scared of the cost of these drugs," Colin said. "So that's been a tailwind in our B2B."

On the member side, Virta's own survey data showed a majority of American adults don't want to be on a weight loss drug for life. And for those already on GLP-1s, Virta's clinical results offered a specific, defensible counter-narrative: close to 80% first-year retention in the clinical trial, 60-70% retention for the broader behavior change program, and 13% average one-year weight loss without drugs.

The strategic response was a deliberate positioning shift. Virta stopped trying to compete with GLP-1s and started positioning as the off-ramp from them. The framing is precise and intentional: Virta doesn't tell clients to block drug access. It offers members a clinically proven path to either enhance the outcomes of a GLP-1 they're already on or eventually discontinue it.

"We have yet to be positioned by our clients as the blocker," Colin said. "We're positioned as that off ramp."

Stacking compounding gains

By 2026, the growth playbook had evolved again. The era of exponential single-channel bets gave way to a portfolio of six or seven simultaneous experiments — each designed to add incremental enrollment volume the team would then compound.

The experiments are concrete: lifecycle-triggered mail sent only to members who have engaged with email but not yet applied; AI-powered outbound calling trained on Virta's data to serve as a consultative voice during the consideration journey; member-facing webinars reintroduced after a two-year pause. The first webinar test, run scrappily in March 2026, drew over 3,000 RSVPs from eligibles across more than 300 clients.

"Different seasons mean different expectations," Colin said. "Now we're looking at: how do we eke out 5% better in this channel? Or how do we find a channel that's going to add 3% to next year's enrollments out of nothing this year? You have to keep stacking those compounding gains."

The machine that was barely running in 2023 now has a methodology. The next phase is building the experiments that compound on top of it.

Six takeaways from this conversation.

Actionable for Marketing Unicorns marketers

  1. Define "access" as a resource your team actively manages and sells.
    In Virta's B2B2C model, "access" is the internal term for the data and permissions needed to reach end users — member email addresses, mailing addresses, and approval to contact them. Roughly half of Colin's 20-person growth team is client-facing, and their primary job is acquiring that access from employer and health plan clients. For B2B2C founders, the growth ceiling isn't always channel performance — it's often how much of your eligible audience you're actually permitted to reach. Building a dedicated function around unlocking access is what makes every downstream channel work.
  2. Test mail format economics before drawing conclusions on the channel.
    When Colin took over growth in mid-2023, direct mail had barely been touched. Early testing revealed the obvious formats — color brochures — performed worst, hitting the junk pile. What worked in healthcare: a legal-size windowed envelope framed around "important benefits information," with the actual content iterating from there. The critical operational wrinkle: unlike email, mail attribution is opaque. Recipients don't scan QR codes or enter vanity URLs at anywhere near full rates, so Virta uses time-based attribution to measure performance. Founders scaling a new channel should build the measurement model before declaring the channel dead — Virta nearly missed a channel that became 40-50% of their enrollment volume.
  3. Match talent specificity to channel maturity. Colin ran mail himself in 2023.
    In 2024 he hired a high-caliber generalist — someone who had founded a company and led healthcare initiatives at the White House — who scaled the channel from a few hundred thousand pieces a quarter to several million. At that point, with mail becoming a multi-million dollar budget line and the largest paid channel, they brought in a dedicated mail channel manager: a specialist in mail economics, vendor logistics, automation, and CAC management. The sequencing matters — a specialist hired too early has nothing to optimize; a generalist left too long caps the channel's ceiling.
  4. When a channel structurally can't work, stop testing it.
    Virta ran paid Meta for two years before cutting it to near zero. The constraint wasn't creative or targeting skill — it was HIPAA. They couldn't upload audience lists, pass back PHI-adjacent conversion signals, or use hashed lookalike audiences effectively. They tried vendors promising compliant conversion data pass-back and employer-level targeting to identify who was covered for Virta. None of it converted. The underlying problem: there's no Meta algorithm that knows who an employer or health plan has sponsored, and until Virta covers a large enough share of the American adult population for broad targeting to work, it never will. Two years and a structural constraint diagnosis later, Colin entered 2026 with confidence in a near-zero paid social budget. The lesson: know the difference between a channel that needs more testing and one your business model makes structurally unwinnable.
  5. Reframe a market threat as a positioning wedge.
    When GLP-1s went mainstream, the initial read was existential — everyone wants the easy button, and Virta requires behavior change. What Colin observed instead: payers across the board (self-insured employers, ASOs, health plans, Medicare) are scared of GLP-1 drug costs, creating a genuine B2B tailwind. On the member side, Virta's own survey data shows a majority of American adults don't want to be on a weight loss drug for life. Virta's response was to position as the off-ramp — not the blocker. The deliberate framing: Virta works alongside GLP-1s to enhance outcomes and provides the clinically proven path for members who want to eventually discontinue. Founders facing a well-funded market entrant should ask whether that entrant is also creating demand or awareness they can capture, and whether their positioning can complement rather than oppose.
  6. Run a portfolio of small bets once your primary channel matures.
    In 2023-2024, Virta was chasing exponential moves — mail 50-100x, enrollment rates doubling. By 2026, the frame shifted to a portfolio of six or seven simultaneous experiments, each targeting incremental gains. Concrete examples from the transcript: lifecycle mail triggered by specific funnel behaviors (engaged with emails, not yet applied), AI-powered outbound calling trained on Virta's data, and member-facing webinars. The first webinar test — run scrappily — drew 3,000+ RSVPs from eligibles across 300+ clients and produced strong attendance and follow-up rates. The expectation calibration is explicit: mature channels get optimized for 5% gains; new experiments target adding 3% to annual enrollments from a standing start. Founders scaling through multiple growth phases should expect the bet size and growth rate to evolve — and build an experiment management process that can hold six or seven parallel tests without losing rigor.

More episodes

Christina Bottis
Chief Global Marketing Officer · Mural
Matt Hooper
Chief Marketing Officer · Quantexa
Kerry Haring
Head of Data Partnerships · Handshake
Stephanie Lottridge
Head of Content & Comms, Newspack · Automattic