Primary user
Founders & marketing leads
Easier Validation / Phase 0
See where Meta clicks stall before the submit.
Easier Validation audits the real mobile click path behind Meta ads, records the funnel screen by screen, and layers on a fast simulated audience panel so founders and marketing leads can see what turns curiosity into drop-off.
Primary promise
Mobile funnel leak diagnosis
MVP boundary
Pre-submit only
Single Ad Audit
This is the first vertical slice of the real product: one direct ad snapshot in, one screenshot-backed teardown out. Use a landing-page override when you want a guaranteed click-path run while we harden automatic destination detection.
Intake Workbench
Start with a precise source. Easier Validation will normalize the input, tell you whether it is ready to run, and show how much manual resolution is still needed before the audit begins.
Resolution preview
See whether the intake is audit-ready, needs an advertiser match, or still needs a shortlist decision.
Choose an input path, paste your values, and preview how the audit will resolve them.
Inputs
The intake is intentionally precise. We are optimizing for reliable advertiser resolution and reproducible audits, not loose screenshot guessing.
Ad Library Advertiser URL
Default path. Best for low-friction intake and strongest automatic advertiser resolution.
Direct Ad Snapshot URLs
Most precise path when the user already knows which ads matter.
Library IDs
Useful for agencies or operators copying exact references from research workflows.
Facebook Page URL
Resolver path when the advertiser page is known but the Ad Library URL is not.
Instagram Handle
Resolver path for brands that are easier to identify via Instagram than Facebook.
Phase 0 Run Flow
Phase 0 is intentionally narrow: one direct ad snapshot, one mobile walkthrough, one screenshot-backed teardown. The engine uses deterministic browser automation for navigation and screenshots, then AI to interpret what happened and package it into a premium report.
1
Resolve advertiser
Normalize the input, fetch the best advertiser candidate, and surface a human confirmation step only when confidence is weak.
2
Select ads
Auto-pick 3-5 ads when reliable reach data exists, otherwise let the user confirm the shortlist in one step.
3
Walk mobile funnel
Open each ad path with Playwright mobile emulation, capture viewport and full-page screenshots, and stop before submission.
4
Score friction
Use vision + structured extraction to log message match, trust, clarity, offer continuity, and obvious leak points.
5
Run cohort panel
Generate a best-guess ICP, let the user tweak it, then simulate 8-12 audience members plus interpretation variants.
6
Return report
Bundle screenshots, decision logs, leak priorities, and audience reactions into a high-clarity report that tees up a call.
Decision Log
We do not expose raw chain-of-thought. Instead, each key screen gets a structured decision log that feels like a live teardown without becoming messy or opaque.
Ad promise
Observed: Creative promises a savings angle and a quick route to the offer.
Reaction: Cold traffic understands the headline, but the CTA feels broad rather than specific.
Fix: Mirror the exact benefit and action in the first landing-page screen.
Landing hero
Observed: Hero image is premium, but the offer language competes with brand storytelling.
Reaction: A high-intent click still has to work out what to do next.
Fix: Collapse to one primary promise, one proof point, and one CTA.
Form threshold
Observed: The form asks for extra fields before value is obvious.
Reaction: Lower-trust visitors hesitate and start looking for escape hatches.
Fix: Move optional questions later and restate what happens after submit.
Audience Panel
The MVP uses an internal simulated panel because it is quicker, cheaper, and easier to tune. The user gets a best-guess audience profile first, then can tweak the most important psychographic and demographic controls before the panel runs.
Editable controls
RoleCompany sizeSectorGrowth pressureBuying sophisticationBudget sensitivityPrimary painPrimary objectionRisk toleranceDecision dynamics
Why internal first
- It adds another dependency before we know the best persona controls and report format.
- The internal panel is enough for a fast, hands-off first release.
- A provider abstraction keeps the door open for Ask Rally later without blocking this launch.
Panel composition
- 8-12 audience members for balanced reactions
- Multiple interpretation variants around the same ICP
- Consensus themes plus disagreement signals
Demand-gen lead at a 40-person SaaS company
Sees the flow through a pipeline-efficiency lens and cares about proof over polish.
Founder-led ecommerce operator
Feels performance pressure personally and is highly sensitive to wasted clicks.
Marketing director in a service business
Balances brand, lead quality, and team adoption more than pure CPL.
Recommended Stack
Mixing languages across a broader suite is fine. For this MVP, keeping the product surface in React and TypeScript gives you the cleanest path to reuse and maintain.
| Layer | Choice | Why |
|---|---|---|
| Product app | Next.js App Router + React + TypeScript | Keeps the suite coherent and makes reporting, auth, and shared UI easier to reuse. |
| Automation | Playwright worker | Deterministic for navigation and full-page screenshots, with AI used for interpretation rather than blind clicking. |
| Model orchestration | OpenAI Responses API | Good fit for image understanding, structured outputs, and selective tool-driven reasoning. |
| Queueing | Redis + BullMQ | Simple way to run multi-ad audits asynchronously without tying long jobs to request time. |
| Persistence | Postgres + S3-compatible storage | Structured run state in Postgres, binary artifacts in object storage. |