INTRODUCING
From answer to action.
Carry conversational state from AI answers to your site, render a matching experience, and prove it with metrics like Promise-Fulfillment Rate.
Teams use 2nd Half to keep their AI ads honest—and to lift conversion by aligning the landing experience to the promise made in the assistant.
AI Ads: 2025 Reality Check
Where the action is—and how 2nd Half completes the journey
AI answers are now an ad surface. Google's AI Overviews, Microsoft Copilot, Amazon's Rufus, and Perplexity all place sponsored units inside the conversation—not just beside it. The catch? Clicks still travel with classic IDs (gclid/msclkid/UTMs). There's no standard way to carry the conversation's state across the click, and cross-surface measurement is thin. That's the gap 2nd Half fills.
Inventory (what's live)
Ads appear inside AI answers across Google, Copilot, Rufus, and Perplexity—more like guided tiles and conversational units than blue links.
Plumbing (how clicks flow)
Outbound links still use traditional click tracking. There's no native "promise payload" handoff, so state from the conversation is lost on arrival.
Measurement (what's missing)
Platforms tout engagement, but answer → follow-ups → click → action isn't measured end-to-end. Last-click can't see whether your page kept the promise.
Supply dynamics (who has what)
Google/Microsoft control reach; Amazon owns shopping intent; Perplexity is the independent testbed; OpenAI is staffing up for on-platform ads.
Who to watch
Microsoft Advertising
Copilot Showroom formats built for chat.Amazon (Rufus)
Sponsored units inside shopping conversations.Perplexity
Sponsored Questions: clean, assistant-native inventory.OpenAI
Hiring for ad infra; expect on-platform placements.No standard state handoff. No trustworthy measurement.
That's why 2nd Half exists.
Relay
Carry the assistant's structured promise (entities, constraints, claims) across the click via a signed payload or edge header—alongside gclid/msclkid/UTMs.
Score
Measure the journey honestly with new KPIs:
PFR (Promise-Fulfillment Rate) — did the page match the promise?
FIS (First-Intent Success) — did users complete the promised action without detours?
CAT (Conversation-to-Action Time) — how fast from answer to action?
Integrity — were claims rendered from a registered truth source?
How 2nd Half plugs into each surface
Google AI Overviews → Append Relay payload to your deep links; Score computes PFR on first paint and binds to gclid for modeled conversions.
Microsoft Copilot (Showroom) → Capture selected filters/options as a Promise Object; render a matching variant; feed FIS back to Microsoft Ads.
Amazon Rufus → For sponsored units that exit to your site, Relay preserves shopping constraints (price/size/brand); Score verifies Integrity against your catalog.
Perplexity → Ideal for pilots: run Sponsored Questions; measure PFR vs. CPA; publish your uplift.
OpenAI (anticipated) → Be first to wire a Promise Object when inventory opens; 2nd Half is already spec'd to carry and verify it.
Proof & outcomes
Higher PFR correlates with lower CPA and faster decisions.
Delta Banners disclose nearest-compliant adjustments (e.g., 90 → 120 days rollout) and preserve trust.
Promise Integrity Reports identify the top broken promises to fix—by campaign, query, or assistant thread.
Carry state from answer to action in under an hour.
See how promise-keeping predicts conversion better than CTR.
Product Pillars
2nd Half Relay
Carry assistant/ad state across the click with a signed payload or edge header. Works alongside gclid/msclkid/UTMs.
2nd Half Score
Measure the journey the way it happens now: assistant → click → render → action. Track PFR (Promise-Fulfillment Rate), FIS (First-Intent Success), CAT (Conversation-to-Action Time), and Claim Integrity.
2nd Half Compose
Optionally render server-side variants that match the promise—grounded by your specs, eligibility, and CPQ. If there's no exact match, show the nearest compliant option with a transparent delta.
Last-click can't see the promise.
Assistants set specific expectations—price bands, SKUs, coverage, timelines. If the landing page doesn't match, users bounce and Legal worries. 2nd Half captures the promise, verifies what you rendered, and shows you what to fix.
PFR
does your page keep the promise?
FIS
did users complete the intended action without detours?
CAT
how fast from answer to action?
Integrity
were all claims sourced?
How it works
Promise In
Ingest a structured promise from the assistant/ad (or reconstruct from UTMs/gclid).
Relay
Pass state via signed query param or edge header; bind to a Conversation ID (CID).
Render & Score
Render your page; compute PFR/FIS/CAT and verify claims against your truth sources.
Close the Loop
Upload modeled conversions (with PFR) back to ad platforms and your warehouse.
Metrics
PFR
Promise-Fulfillment Rate (0–1)
Semantic + token match between the promise and your hero/CTA/modules.
FIS
First-Intent Success
Completed the promised action without navigation detours.
CAT
Conversation-to-Action Time
Time from final assistant answer to action.
Integrity
Claim Integrity
% of rendered claims backed by a registered source.
ICP Use Cases
Cloud & AI Infra
Keep region/SKU/price/SLA promises and accelerate scoped opportunities.
Cybersecurity
Map frameworks and coverage to a reference architecture; prove it on the page.
Telecom/FinServ/Insurance
Eligibility-matched offers with audit-ready claims.
🔒 Cybersecurity
From Framework Ask to Scoped Opportunity
Secure the handoff from assistant to action.
When prospects ask assistants for SOC 2 + HIPAA, 5k endpoints, ZTNA in 90 days, 2nd Half carries that promise to your site, renders a matching reference architecture, and proves alignment with Promise-Fulfillment Rate (PFR).
Typical assistant promises we capture
Frameworks & regs: SOC 2, ISO 27001, HIPAA, PCI-DSS, NIST 800-53
Coverage: EDR/XDR + MDR, M365/Google Workspace, SIEM/SOAR, ZTNA/SASE
Scope & scale: endpoint counts, regions, clouds, MFA/SSO, DLP
Constraints: rollout windows, SLA/SLO, budget bands, tooling preferences
What 2nd Half does on your landing
Relays the promise (frameworks, coverage, scale, timeline) via a signed payload or reconstruction from UTMs/click IDs.
Renders to the promise with a modular page: Reference architecture (e.g., EDR→XDR→SIEM→SOAR with ZTNA edge), 90-day rollout plan (waves, change control, success criteria), Evidence: audited frameworks supported, SLAs, case studies, Clear CTAs: Request scoped assessment or Book readiness review
Verifies claims against your truth sources (what you actually support). No off-registry statements.
Scores the visit with security-specific KPIs.
Security-grade metrics we track
PFR
alignment between promised frameworks/coverage and your hero/CTA/modules (0–1)
FIS
visitor requests the scoped assessment without detours
CAT
seconds from final assistant reply to action
Claim Integrity
% of rendered statements backed by your registry
Control Coverage Match
% of promised controls mapped to your reference design
Framework Alignment Score
evidence that SOC 2/ISO/PCI requirements are addressed on-page
Outcomes teams see
Fewer broken-promise bounces; higher qualified assessments/POCs.
Faster sales engineering (handoff includes structured scope).
Lower legal/compliance risk (claims are signed + auditable).
30-day pilot for security programs
- 1
Map promises: choose 3 intents (e.g., SOC 2 + HIPAA MDR, ZTNA rollout, Managed SIEM).
- 2
Wire Relay & Score: capture promise → render matching modules → emit PFR/FIS/CAT.
- 3
Truth registry: allowed frameworks, stacks (EDR/XDR/SIEM/SOAR vendors), SLA language.
- 4
A/B against your current pages; deliver a Promise Integrity Report with lift in qualified assessments.
See how often your pages keep the promise—and what to fix first.