Research

Google Ads Strategy

outputs/agent-5-funnel-specs/google-ads-strategy.md

TLDR

  • Google Search traffic is fundamentally different from Facebook: high-intent, full browser, Google SSO available, no ejection needed
  • Quiz funnels should NOT be used for Google Ads — content-rich /lp/_ and /specialties/_ pages already convert at 7.18% (above B2B SaaS benchmark)
  • All 5 variations share: no quiz, SSR landing pages, Google SSO as primary signup, direct "Try for Free" CTA
  • Each variation adds unique landing page enhancements: V2 adds demo preview, V3 customizes H1 per ad group, V4 embeds ROI calculator, V5 adds peer counts
  • Keyword clusters mapped to best variation approach; draft ad copy provided per cluster

Google Ads Funnel Strategy — Cross-Variation Analysis

Why Google Ads Needs Different Entry Points Than Facebook

The Fundamental Channel Divide

Facebook/Instagram and Google Search serve users at fundamentally different points in the buyer journey, which demands fundamentally different funnel architectures:

DimensionFacebook/InstagramGoogle Search
Intent levelLow — interrupted while scrollingHigh — actively searching for a solution
Knowledge stateOften unaware they need AI notesKnows they want an AI scribe, comparing options
Browser contextIn-app browser (no Google SSO, no password manager, unreliable scroll)Full system browser (Google SSO, standard capabilities)
Landing page jobHook attention → build desire → earn commitmentAnswer the query → confirm fit → convert
Quiz roleWarms cold traffic, builds investment, creates sunk costAdds friction for transactional queries, appropriate only for discovery queries
Signup frictionEmail+password only (Google SSO unavailable) → must eject to system browserGoogle SSO available (lowest friction) → no ejection needed
Biggest riskDrop-off at quiz entry (54.52% at Q1)Bounce from poor message match or unnecessary friction

Why This Matters for Twofold

Twofold's quiz funnel at start.trytwofold.com was tried as a Google Ads destination and got a low Quality Score. This is expected — the quiz page has robots: noindex, no <h1>, is pure client-side React, and has no keyword-relevant text content. Google's Quality Score algorithm penalizes pages that don't match the search query.

Meanwhile, Twofold's content-rich /lp/* and /specialties/* pages already perform well for Google Ads: 7.18% conversion rate (above B2B SaaS benchmark of 4.7%), 6.51% CTR, and all ads rated "Excellent" strength.

The strategic conclusion is clear: Google Ads traffic should NOT go to the quiz funnel. It should go to the existing content-rich landing pages, with optimizations informed by each variation's strategy.


Shared vs. Divergent Funnel Components

What All Variations Share for Google Ads

These elements are constant across all 5 variations' Google Ads paths:

  1. No quiz funnel — Google search traffic goes directly to content-rich landing pages, not the quiz. None of the highest-performing competitors use a marketing quiz for Google Search traffic. Quizzes only appear when they serve a functional purpose (Hims: medical qualification, Lemonade: insurance quoting).

  2. Content-rich SSR landing pages — existing /lp/* and /specialties/* templates are the foundation. Server-rendered, semantic HTML, proper H1/H2, structured data (Product schema, 4.9/5 rating), 2,500+ words of keyword-relevant content.

  3. Google SSO as primary signup method — unlike the Facebook flow (where Google SSO is unavailable), Google Ads visitors are in a full system browser with an active Google session. "Continue with Google" should be the first and most prominent signup option. This is a competitive advantage over Freed, which only offers email+password.

  4. Direct "Try for Free" CTA — single, prominent call-to-action. No intermediate steps. Landing page → signup page → app.

  5. No browser ejection needed — Google Ads visitors are already in a full system browser. Recording, microphone permissions, and all app features work natively. This eliminates the biggest friction point of the Facebook flow.

  6. Post-signup experience is shared — once a user signs up (regardless of channel), the onboarding wizard, /new page, and activation flow are the same. The channel-specific differences end at signup.

What Diverges Per Variation

ComponentFacebook FlowGoogle Flow
Entry pointQuiz at start.trytwofold.comContent page at trytwofold.com/lp/_ or /specialties/_
Pre-signup engagementQuiz questions → loading → results → signup formLanding page content → "Try for Free" CTA → signup page
Signup methodsEmail + password only (in-app browser)Google SSO (primary) + email + password
Browser ejectionRequired — auto-eject after signup to system browserNot needed — already in system browser
Signup framing"Save your workspace" (quiz completion continuation)"Start Free" or "Try for Free" (standard CTA)
Personalization before signupDeep — quiz answers personalize results and setupLight — specialty from page context, note format from onboarding
Variation-specific elements on landing pageN/A (quiz IS the variation)See variation-specific sections below

Variation-Specific Google Landing Page Enhancements

Each variation can contribute unique elements to the Google Ads landing pages:

VariationEnhancement for Google Landing Pages
V1: Speed Start v2None — standard content pages work well. Focus on CTA optimization and Google SSO prominence.
V2: Instant ValueAdd interactive note preview element to the landing page. User selects specialty → sees sample note generation. "Try before you sign up" on the landing page itself.
V3: Pain PathCustomize H1/hero messaging per ad group to match the pain cluster. Treatment Plans ads → "AI Treatment Plan Generator" hero. Therapy Dictation ads → "Dictate Notes in 30 Seconds" hero.
V4: ROI CalculatorEmbed a simplified ROI calculator on the landing page: "How much time do you spend on notes? [slider] → You could save X hours/week." CTA: "Start saving — Try Free."
V5: Peer ProofAdd specialty-specific peer data to landing pages: "4,200+ therapists use Twofold" (on therapy pages), "Trusted by 2,100+ psychiatrists" (on psychiatry pages). Real-time or near-real-time user counts.

Keyword Cluster → Variation Mapping

Different keyword intents demand different landing page approaches. Here's which variation's Google Ads approach works best per keyword cluster:

Cluster 1: Specialty-Specific Keywords

Keywords: "AI therapy notes", "psychiatry AI scribe", "psychology documentation", "social work notes AI", "counseling notes generator", "physical therapy AI scribe"

Best approach: V3 (Pain Path) + V5 (Peer Proof)

  • Use /specialties/* template pages with specialty-specific testimonials, FAQs, and feature descriptions
  • H1 must match the specialty keyword exactly (already done: "Best AI Medical Scribe for Psychiatry")
  • Add specialty-specific peer counts (V5 enhancement): "Trusted by 4,200+ therapists"
  • These are medium-intent keywords — user is researching solutions for their specialty

Current performance: Psychiatry (8.64% CVR, best in portfolio), Psychology (10.00%), Mental Health (6.56%)

Draft ad copy:

  • Headline 1: "AI [Specialty] Notes — 30 Seconds"
  • Headline 2: "Built for [Specialty] by Clinicians"
  • Headline 3: "HIPAA Compliant · From $19/mo"
  • Description: "AI scribe built specifically for [specialty]. Generate [format] notes from sessions in 30 seconds. Trusted by 20,000+ clinicians. HIPAA compliant. Try free for 7 days."

Cluster 2: Product Feature Keywords

Keywords: "AI progress note generator", "SOAP note generator", "therapy dictation software", "treatment plan generator", "AI medical scribe", "therapy charting software"

Best approach: V4 (ROI Calculator) + V2 (Instant Value)

  • Use /lp/* template pages with feature-specific content
  • H1 must name the specific feature ("AI SOAP Note Generator")
  • Add ROI calculator widget (V4 enhancement): "How much time do you spend on notes?"
  • Or add note preview (V2 enhancement): "See a sample SOAP note"
  • These are high-intent transactional keywords — user knows what feature they want

Current performance: Note Types (11.22% CVR, best cost efficiency at $198.70/conv), Progress Notes (2.62%), Therapy Dictation (7.83%), Therapy Charting (5.00%, best cost at $47.91/conv)

Draft ad copy:

  • Headline 1: "AI [Feature] — Try Free"
  • Headline 2: "[Feature] in 30 Seconds"
  • Headline 3: "20,000+ Clinicians Trust Us"
  • Description: "Generate [feature] from session recordings or dictation in 30 seconds. 50+ note formats. HIPAA compliant. From $19/mo. No credit card needed."

Cluster 3: Compliance-Driven Keywords

Keywords: "HIPAA compliant AI notes", "HIPAA AI scribe", "HIPAA compliant therapy software"

Best approach: V1 (Speed Start v2) — standard content pages with heavy compliance messaging

  • Use /lp/hipaa-compliant-ai-note-taker page
  • Lead with compliance: BAA included, end-to-end encryption, recordings never stored
  • These are trust-seeking keywords — user's primary concern is security, not features

Current performance: HIPAA (4.02% CVR, below portfolio average)

Draft ad copy:

  • Headline 1: "HIPAA Compliant AI Notes"
  • Headline 2: "BAA Included · No Extra Fee"
  • Headline 3: "End-to-End Encryption"
  • Description: "HIPAA-compliant AI note taker with signed BAA included. Recordings are never stored. End-to-end encryption. Trusted by 20,000+ clinicians. Try free."

Cluster 4: Competitor Targeting Keywords

Keywords: "[Competitor] alternative", "best AI scribe for therapists", "Freed vs Twofold"

Best approach: V5 (Peer Proof) + dedicated comparison pages

  • Create comparison landing pages: /compare/freed, /compare/deepscribe, etc.
  • Lead with Twofold's advantages: price ($49/mo vs $90+/mo), Google SSO, template customization
  • Add switching testimonials: "I switched from [Competitor] and saved $500/year"
  • These are late-stage high-intent keywords — user is deciding between options

Draft ad copy:

  • Headline 1: "Looking for a [Competitor] Alternative?"
  • Headline 2: "Half the Price, Same Quality"
  • Headline 3: "20,000+ Clinicians Chose Twofold"
  • Description: "Switch from [Competitor] to Twofold. Same AI scribe quality at $49/mo (vs $90+). HIPAA compliant. 50+ note formats. Google SSO. Try free — no credit card."

Cluster 5: Pain-Driven Keywords

Keywords: "reduce clinical documentation time", "therapy notes taking too long", "clinical documentation burnout"

Best approach: V3 (Pain Path)

  • Route to pain-matched landing pages
  • H1 echoes the pain: "Stop Spending Hours on Therapy Notes"
  • Immediate empathy + solution in above-the-fold
  • These are problem-aware but solution-unaware — need education + proof

Draft ad copy:

  • Headline 1: "Finish Notes Before Your Next Session"
  • Headline 2: "AI Writes Your Notes in 30 Seconds"
  • Headline 3: "HIPAA Compliant · From $19/mo"
  • Description: "Stop charting at 9 PM. Twofold's AI listens to your sessions and generates notes in 30 seconds. Trusted by 20,000+ clinicians. Try free."

Overall Recommendation: Google Ads Funnel Architecture

Google Search → Keyword-matched /lp/* or /specialties/* landing page
                    ↓
            "Try for Free" CTA (single, prominent)
                    ↓
            Signup page (Google SSO first, then email+password)
                    ↓
            Onboarding wizard (3 steps: specialty, practice size, note preferences)
                    ↓
            /new page with guided activation (same as Facebook post-signup)

Why No Quiz for Google Traffic

  1. Research consensus: 40+ sources confirm direct CTA outperforms quiz for transactional search queries. Quiz completion for paid search is 35-50% (vs 55-70% organic).
  2. Competitor validation: Freed (direct competitor, most sophisticated Google strategy) uses zero quiz — only keyword-specific content pages with "Try for Free" CTA.
  3. Quality Score: Quiz pages fail Google's Quality Score requirements (no keyword-relevant content, client-side only, noindex).
  4. Intent mismatch: A clinician who searched "SOAP note generator" already knows what they want. A quiz asking "What's your role?" is friction, not value.
  5. Twofold's own data: Current /lp/* pages already achieve 7.18% CVR without any quiz — above the B2B SaaS benchmark of 4.7%.

The One Exception: Commercial Investigation Queries

For discovery-phase keywords ("best AI scribe for therapists", "clinical documentation software comparison"), a lightweight interactive element (not a full quiz) could improve conversion:

  • V2's note preview: "Select your specialty → See a sample note"
  • V4's ROI calculator: "How many sessions/week? → See your savings"

These are not quizzes — they're value-delivery tools embedded in the landing page. They serve the searcher's intent (evaluating options) while collecting data for personalization.


Technical Requirements Checklist

All Google Ads landing pages must meet these requirements:

Already Met (Current /lp/_ and /specialties/_ pages)

  • Server-side rendered (SSR) — content visible to Googlebot
  • Semantic HTML with proper H1/H2 hierarchy
  • Structured data (Product schema, 4.9/5 rating, Organization, WebSite, WebPage)
  • Meta descriptions present
  • Mobile-responsive layout
  • Multiple CTA placements (5-8 "Try for Free" buttons per page)
  • HIPAA/compliance badges visible
  • Testimonials with credentials
  • FAQ section
  • No robots: noindex

Improvements Needed

  • Fix mobile signup page responsivenessapp.trytwofold.com/get-started renders desktop layout on mobile (HIGHEST PRIORITY — affects 82.9% of traffic)
  • Promote Google SSO on landing pages — add "Sign up in seconds with Google" messaging near CTAs (not just on signup page)
  • Strip navigation for Google Ads traffic — use UTM parameters to serve minimal nav (logo + "Try for Free" only) to prevent exit paths
  • Fix user count inconsistency — "18,000+ therapists" on some pages vs "20,000+ providers" on others; pick one and standardize
  • Add product screenshot/demo above the fold — current pages push product visuals below fold
  • Create comparison pages for competitor targeting keywords: /compare/freed, /compare/deepscribe
  • Expand keyword-specific page coverage — create dedicated pages for Note Types (11.22% CVR), Progress Notes, and Therapy Charting ad groups
  • Migrate high-performing ad groups to /specialties/* template — it has richer structured data, more FAQs, specialty-specific content. Psychiatry (8.64% CVR) already uses it.
  • Add page speed optimization — each 1-second delay = 7% conversion drop. Target < 2.5s load time.
  • Consider /lp/* page variants per variation enhancement — V2 note preview, V4 ROI calculator, V5 peer counts

For Variation-Specific Google Enhancements

  • V2: Interactive note preview component (client-side, hydrated after SSR page load)
  • V4: ROI calculator widget (client-side, embedded in /lp/* pages)
  • V5: Dynamic peer count API (returns specialty-specific user counts for landing page display)

Performance Monitoring

Key Metrics to Track (Per Ad Group)

MetricCurrent BenchmarkTarget
CTR6.51%7-8% (after message match optimization)
CVR7.18%9-10% (after signup page fix + Google SSO promotion)
CPA$303.92$200-250 (through CVR improvement)
Google SSO usage rateUnknown (not tracked per channel)Track and target 50%+ for Google Ads traffic
Activation rate (Google)UnknownTrack separately; expect 55-65% (no in-app browser issues)
  1. Test 1: Current signup page vs mobile-responsive signup page (highest impact fix)
  2. Test 2: Full navigation vs stripped navigation for Google Ads traffic
  3. Test 3: Current landing pages vs pages with Google SSO messaging ("Sign up in seconds with Google")
  4. Test 4: /lp/_ template vs /specialties/_ template for same keyword clusters
  5. Test 5: Standard page vs page with interactive element (V2 note preview or V4 ROI calculator)