Novyra

Engagement Dashboard

Last Updated

—

In Market

The system is getting better at every stage.

Leads are getting cheaper, tracking is getting better, the creative direction is promising, and HubSpot is starting to turn interest into signups. On the closed week Aug 27–Sep 2, Meta delivered 737 leads at $3.25 CPL. C_dark and PreHealth led the creative mix on paid signup efficiency, while Google recorded its first account_created conversion on Sep 2. Every new lead now auto-enrolls in the how-to email sequence, with 441 people in the cohort as of Sep 3. Real app signups reached 75 for the week after removing obvious tests, versus 30 prior. The sequence is working directionally; another week or two will show whether the lift holds. Audience Network stays off, and Reddit waits on bot and datacenter filtering.

Velocity this week: Meta Aug 27–Sep 2 ran $3.25 CPL on 737 leads and $2,397.72, versus $3.47 on 669 leads and $2,318.37 the week before. Blended cost per paid signup is $44.39, better than the $50–70 plan. S2 remains the worst ad at $130.29 per signup, but the gap narrowed versus S1 at $56.51, so re-argue before any kill. Real app signups are 75 after removing obvious tests, versus 30 prior. Google logged its first conversion Sep 2. Reddit is still not live, gated on bot and datacenter filtering.

Phase

2 of 2, In-Market

Engagement Start

May 11, 2026

Live Channels

Meta · Google Search

Email leads, no account

2,863

as of Sep 3 · contact has no signup_date

App signups, all-time

2,189

as of Sep 3 · signup_date

In-market performance

Live numbers from Meta, Google, and the CRM · pulled Sep 3

Meta Cost per Lead

$3.25

vs prior 7d $3.47 (↓ $0.21)

737 leads on $2,397.78 for Aug 27–Sep 2, versus 669 leads on $2,318.37 Aug 20–26. Spend flat, leads up, cost down. Audience Network stayed at zero: Facebook $2,157 / 649 leads, Instagram $230 / 84. Closed week, Pacific account time.

Cost per Paid Signup

$44.39

vs $50–70 plan (better than plan)

Aug 27–Sep 2 ad-level join: $2,397.27 across the four live ads against 54 signups. By ad: C_dark $25.06 at 16.0%, PreHealth $40.14 at 10.0%, S1 $56.51 at 8.3%, S2 $130.29 at 3.1%. 42 contacts in the window had an empty ad stamp, so CPS is slightly overstated.

App Signups (last 7 days)

vs 30 prior 7d (↑ 45)

Aug 27–Sep 2, counting HubSpot signup_date. 91 records in the window; 16 were our own or dev test accounts, so 75 is the real figure. Prior week 43 raw / 30 real. That is a ~320-a-month pace, inside the locked 250–330 target. The 3-use wall still understates app use.

Google Search CPC

$5.86

vs $5.85 prior (flat)

99 clicks / $580.01 Aug 27–Sep 2 vs 128 / $749.14 prior. CTR 9.19% vs 10.15%. First conversion recorded Sep 2 on account_created, after weeks at zero. Budget stays $100/day while volume builds.

How we drive cost per signup

Operating plan · not a forecast

Signups recovered on the closed week: 75 real versus 30 the week before, a ~320-a-month pace that lands inside the locked 250–330 target. The live HubSpot how-to sequence is the most likely driver, and blended cost per paid signup came in at $44.39 against a $50–70 plan. The next two moves are cutting the ad that is not converting and turning Reddit on once its tracking clears. Reddit is not live.

Keep Audience Network off. Optimize on paid signups, not lead CPL.AN share is 0% this week. Facebook carried $2,157.15 / 649 leads. Instagram carried $229.52 / 84. Unknown was $11.06 / 4.

Thursday Sep 3: closed-week cost per signup by ad.S2 remains worst at $130.29, but the gap narrowed sharply. Re-argue before any kill. Videos 3–4 remain unpublished drafts.

Videos 1–2 stay out of new work; the current priority is measuring the live conversion signal and finishing the V3/V4 draft QA.

Google stays $100/day until conversions report. Exception keywords stay paused.

HubSpot how-to sequence is live and auto-enrolling every new lead. 441 in the cohort as of Sep 3.Emails 1–5 are published and sending on schedule. Job of HubSpot is turning existing leads and new Instant Form emails into app accounts at email cost, which is how user count rises without raising media CPS. Real signups went 30 to 75 week over week while this ran.

Free search before signup stays open while we measure HubSpot.Locked Aug 26 with Gerard. Forced signup comes after we see what the reach-out does to signup rate versus user rate, in a few weeks to a month. No session-replay tooling.

Dev: signup tracking (gclid on the homepage, not /login) and the results-reset fix, so Search and Meta can optimize on real signups.

Sep 1 $20K media only against this plan. Reddit is not live.Campaign shell is built and paused. Gated on the FiveRivers bot and datacenter filter, which still has no ship date, plus server-side events on the stored click ID and a joint dry run. Spec is $100/day, US, Feed + Conversation, third-party off. Reddit is the next signup lever now that the HubSpot sequence is running.

Engagement Roadmap

Foundation → In-Market → Scale

May–Jun

Foundation & strategy

Jul

Meta launch

Aug 18

Google Search serving

Aug 20

Placement cleanup + email program

Now

Attribution build + Reddit relaunch

Sep 1

$20K media budget

In market this week

Cleaner traffic, signup rate up

Aug 20–26

Audience Network stayed at zeroFacebook $1,823.94 / 537 leads. Instagram $226.26 / 64. Versus $525 AN spend and 144 AN leads the week before.

Lead-to-signup rate rose even as CPL rose5.5% this week (33 / 604) versus 4.7% prior (38 / 802) versus 2.9% Aug 18–24 (20 / 689). CPL $3.41 versus $2.91.

How-to sequence live and auto-enrolling, 441 in cohortEmails 1–5 published and sending. Enrollment is automatic on every new lead, no manual batches. First 150 went Aug 24 from hello@ at a 3.11% bounce on A1.

Google Search held at $100/day, first conversion inAug 27–Sep 2: 99 clicks, $580.01 spend, $5.86 CPC, 9.19% CTR, and 1 conversion on Sep 2 after weeks at zero. Exception keywords stay paused.

Last closed CPS remains $69 (Aug 11–17)This week’s paid-attributed signup count is missing, so no new CPS. Re-read Aug 27.

Coming up for Novyra

Close CPS, hold the next send

Aug 26 – Sep 5

Thursday Aug 27: 7-day cost per signup by adVideos 3–4 stay off until Tom says publish after that read. Videos 1–2 stay out until the leftover TZ badge fill is done.

How-to sequence running on autopilot441 enrolled as of Sep 3, joining automatically as leads arrive. Emails 1–5 published. First 150 went Aug 24 from hello@ at a 3.11% bounce on A1.

Free search before signup stays openLocked Aug 26. Measure HubSpot reach-out vs signup rate for a few weeks to a month, then consider forcing signup. No session-replay tooling.

Signups recovered: 75 real last 7 days versus 30 priorA ~320-a-month pace, inside the locked 250–330 target. The live HubSpot sequence is the most likely cause. Reddit is the next lever and is still not live, gated on the bot filter, server events, and a joint dry run. Spec $100/day, US, Feed + Conversation, third-party off.

Google stays $100/day until conversions reportException keywords stay paused. No raise until the signup event proves what those clicks are worth.

Dev: gclid on the homepage, not /login, plus the results-reset fixSo Search and Meta can optimize on real signups. Reddit stays gated on this build.

September 1: $20K media only against this planAugust held at $16,350. Reddit still gated on the tracking build.

GA4 vs CRM reconciliationApp-side analytics and CRM tell different activation stories. One trusted number before the Oct–Nov investor window.

Working personas

Oncology-first focus locked May 19, 2026

Patient · primary

Cancer patient

U.S.-based, 50+, in treatment or recently diagnosed

Overwhelmed by ClinicalTrials.gov, looking for someone to filter the noise. Reachable through disease-specific Facebook groups, oncology Reddit communities, advocacy sites, and Google search. Aha moment: matched to a real trial, fast, in plain language. Novyra gives patients agency.

Caregiver · primary

Caregiver

Often adult child of older patient, sometimes spouse

Frequently the actual searcher, especially for patients 60+. Doing the research the patient can't, translating it back. Reaches Novyra through advocacy orgs first, then Facebook groups, then late-night Google searches. Aha moment: a concrete path forward to take to the patient. "Here are real trials your mom may qualify for."

Physician · primary

Treating oncologist

Community or academic oncologist, trial coordinator

Resource-strapped, burned by digital tools that add work. Will not adopt a tool that adds steps. Reachable via professional journals, CME, conferences, and group sponsorships. Not via LinkedIn for work. Aha moment: 10-20 second workflow replaces ClinicalTrials.gov. Novyra gives doctors simplicity.

What we found in Novyra's data

May 22 pre-launch audit · historical baseline behind the current setup

Pixel measurement broken

$14.52

reported CAC is not real. No event_id, dual Pixel + CAPI without dedup, pixel firing on localhost + unrelated domains.

Resolved by design: CAC now measured in the CRM (ad-level join to real signups), not from the pixel. Clean pixel/CAPI rebuild ships with the attribution work.

Video format crushes static

20×

$3.91 CAC on top video vs $97 on worst static. Video CTR 5-6% vs static 0.18%.

Mike's production is 80%+ video for cold prospecting. Statics reserved for retargeting only.

/beta is the right landing page

$3.91

CAC for video ads landing on /beta. Static ads to / hit $55-97 CAC.

Earlier "redirect /beta to /" recommendation reverses. Short engagement = signups completing, not bouncing.

April 2026 performance cliff

3×

CAC tripled across Europe, Asia, Australia in April with frequency hitting 5-7.

Audience fatigue + SAC tightening. Confirms 2025 numbers don't port to 2026 plan.

Channel concentration

63%

of sessions from Facebook paid

Multi-channel mix (Meta + Google + endemic) diversifies dependency.

Form-start rate

0.45%

32 form_starts from 7,043 sessions, 90d

Funnel leak at top. Signup flow rebuild scheduled Week 3.

Where Novyra needs to act

Items waiting on Novyra

Reply owner for medical replies

The how-to sequence is live and auto-enrolling, 441 people in the cohort as of Sep 3, so replies are arriving continuously now rather than in one batch. Some will be medical questions that need a named human at Novyra to see them. One name from Gerard settles the reply owner.

Action needed

Dev tracking build: the three items gating Reddit

Click-ID capture at email submit, two server-side conversion events, and bot filtering. Spec delivered Aug 21 with the Google snippet included. gclid must land on the homepage, not /login, plus the results-reset fix. Reddit spend starts only after these land and a joint cross-device test passes. Timeline confirmation requested from FiveRivers.

In flight

HubSpot billing onto a Novyra payment method

Marketing Hub Pro ($1,000/mo + $250 contacts tier + $60 seats, next charge Sep 20) went on Tom’s card to unblock the purchase. The SOW contracts software in Novyra’s name — either move the card or paper the pass-through.

Action needed

Google Ads billing to the Novyra card

Google Search is spending ~$100/day. Billing should sit on Novyra’s Mercury card; Tom enters it on a call, five minutes.

Action needed

Risks Novyra is managing through us

Items we’re watching so they don’t become surprises

Monthly pace vs locked target

~320

Aug 27–Sep 2 ran 75 real signups, a ~320-a-month pace. Locked target is 250–330, so the gap is closed on this one week. One week is not a trend; the read to watch is whether it holds through Sep 9.

Last closed CPS is $69 (Aug 11–17, inside the $50–70 plan). This week’s paid-attributed count is not closed. Re-read Aug 27. Signup rate this week is 5.5%, up from 4.7% prior.

Reddit could launch blind

Mitigated

Reddit optimizes on a server event we send. The magic-link signup often completes on a different device than the ad click, which silently breaks attribution if unhandled.

Spend is gated on the dev build plus a joint cross-device dry run. No event, no budget.

Healthcare ad policy, both platforms

Active

Google accepted only 2 of 20 keyword themes on upload (cancer-specific terms rejected); Meta’s health restrictions keep tightening. Copy runs “may qualify” language everywhere.

Certification posture is clean and monitored weekly. Growth on Google comes from bid and budget on the accepted terms, not from keyword expansion.

Two activation numbers exist

Open

App-side analytics claim a much higher lead→signup rate than the CRM join shows. The Google tag gap this week proved app-side numbers need verification.

Reconciliation owed before the Oct–Nov investor conversations, so the deck number survives diligence.

Decisions already made

For your awareness, no action needed

Free search before signup stays open

Locked Aug 26. Keep the current free uses so we can measure what HubSpot reach-out does to signup rate versus user rate. Forced signup is later, in a few weeks to a month, not this week. No session-replay or click-video tooling.

Locked Aug 26

Audience Network excluded account-wide

Aug 20, on placement-level evidence: converting ads ran 8% of spend there, zero-signup ads ran 62–82%, and it was pricier per lead even for the converters. Applies to every campaign; the “allow limited spending” loophole is off. Standing rule going forward, for Novyra and every account we run.

Done Aug 20

Lifecycle email lives in HubSpot; Marketing Hub Pro purchased

Gerard approved the upgrade; purchased Aug 20 ($1,000/mo Pro + $250 contacts + $60 seats, monthly, 7,000-contact tier). Welcome email moved from the app into HubSpot and is live on every signup; one system now owns every patient-facing email.

Purchased

Patient emails are anonymous by design

No first names anywhere in the lifecycle emails — “Hi there” for everyone. Right posture for a cancer audience, and batched sending (150 → 350 → 500 → rest) protects the new sending domain’s reputation.

Locked Aug 17

Reddit restart spec agreed with Reddit’s team

$100/day for a 10-day burst, US, Feed + Conversation placements only, third-party network off, one server-validated qualified-session event per session as the optimization signal, bot filtering on our side. Click IDs only in every payload — no personal data to any ad platform.

Agreed Aug 20

August budget held at $16,350; $20K starts Sep 1

The SOW’s $20K media budget begins September 1; August stays at the agreed $16,350 with the ~$6K/month of planned placement-test spend cancelled as unnecessary — the exclusion achieved the same thing for $0.

Confirmed

PPC strategy + $10K/mo budget approved

Gerard signed off May 29. Google + Meta funnel approach, $10K/month paid media, 250-330 signups/month at $35-50 CAC once optimized. "Let's press go on all this."

Approved

Signup conversion definition locked

Gerard confirmed May 30. A conversion = user signs up + runs a search + clicks on a trial result. Stricter than email capture, looser than enrollment. Anchors all CAC measurement, pixel events, and Target CPA bidding.

Confirmed

IRB approval not required

Confirmed June 1. Novyra is a matching tool, not a research-conducting entity, and its privacy policy states no medical data is collected. The IRB requirement in Google's clinical trial recruitment policy applies to advertisers sponsoring or running trials, which Novyra isn't. Removes IRB from the launch critical path. Google Ads advertiser certification remains the only compliance gate.

Confirmed

Creative split: patient-heavy

8 launch videos with 6 patient + 2 caregiver (refined from 4/4 on the May 29 call). Monthly refresh of 4 total. Video-first. FDA-compliant copy reviewed before every batch.

Locked

Therapeutic focus: oncology first

Locked May 19. Future expansion candidates: brain health / Alzheimer's, pediatric rare disease.

Confirmed

B2C direction, three primary personas

Investor-confirmed May 18. Three personas: cancer patient (50+), caregiver, treating oncologist. Physicians reached through professional channels (Doximity, Sermo, UpToDate), not Meta.

Confirmed

Reddit added as a Phase 1 launch channel

Reddit moved from Phase 2 deferred to Phase 1 launch channel per Gerard's June 8 direction ("as soon as we have the opportunity to do some Reddit stuff, let's do it") and confirmed June 11. Reddit launches in parallel with Google Search and Meta, not after. Working numbers from David Collins (Reddit agency client partner): $2/click planning estimate, projected $30-60 CAC per qualified signup at a $1-2K monthly test budget within the $10K/month total envelope. Creative direction is organic-looking Reddit-native posts (static images and very short videos). No special healthcare licensing required for matching-service model. Approval timeline 24-48 hours. David handles compliance review. Reddit Pixel installation timed to kickoff. Patient-advocate influencer track remains a Phase 2 candidate.

Phase 1

Filmed advisor interviews

Greenlit. Subjects from the advisory team selected based on availability and on-camera fit. Feeds Meta video creative for cold prospecting.

Greenlit

Working group

Team

Gerard Heath

Founder & CEO

John King

Operations

Mike

Strategy

Thomas Zandstra

Marketing Lead, Acculead

Advisory team in interview rotation
Savi Glowe (clinical strategy) · Nick (data & tech) · Chris Mason, PhD (genomics, Weill Cornell) · Dr. Munira Moosajee (medical oncology)