Which AI surfaces does AgentLens measure?

ChatGPT, Claude, Gemini and Grok through their vendor APIs, and Google AI Mode read as a live search surface. Perplexity is not measured because no real adapter exists for it.

Does AgentLens change my website?

No. It generates deployable artifacts such as JSON-LD, llms.txt, robots.txt and meta tags. You review them and your own pipeline deploys them.

How is the readiness score calculated?

Four pillars at 25% each: Schema Coverage, Content Structure, Agentic Brand Familiarness (Share of Model) and Shopping Assistant. Each is scored 0–100 with a 95% confidence interval and its sample size.

Will I know the cost before a run?

Yes. Every run starts with a spend-free estimate that shows pinned models, expected calls and a hard cap at 115% of the estimate. Nothing runs until you confirm.

Is AgentLens generally available?

It is in beta, verified in a shared staging environment. Beta pilots run there; no production environment is claimed.

Trust

The boundary is the product.

AgentLens turns evidence-backed diagnoses into reviewable artifacts, then measures what changed. It never silently modifies a customer website or activates a commerce integration.

What AgentLens does not do

Stated plainly, because a platform that tells you what it cannot do is a platform you can check.

No silent changes to your site

AgentLens never deploys to your CMS, storefront, robots.txt, DNS or merchant account. It generates; you apply.

No guaranteed rankings or sales lift

Scores and projections are decision support. Model behaviour changes; we show confidence intervals and significance, not promises.

No transactions

No checkout, no payment, no autonomous purchasing on your behalf.

No surfaces we cannot measure

If there is no real adapter for an AI surface, it is not offered or advertised. Perplexity is not on the list today for exactly that reason.

Controls

Your keys, your data boundary, your export formats.

Bring your own keys: OpenAI, Anthropic, Gemini and xAI keys are saved encrypted at rest and can be tested, rotated and deleted. Owners and admins only.
Roles and quotas: Signup with email verification, role-based access, domain ownership, usage quotas and a durable audit history per tenant.
Tenant data boundary: FastAPI, Next.js and PostgreSQL on Google Cloud Run; tenant-scoped data, roles and quotas; 205 public API operations under a locked OpenAPI contract.
Immutable snapshots: Report snapshots keep their metric definitions. Metric versions are pinned so dashboards never silently change meaning.
Exports that travel: Executive diagnosis, buyer narrative, treatment pipeline and peer benchmark as PDF and PPTX; GEO snapshots add XLSX and JSON.
Analysis and measurement stay apart: Drafting and classification may run on the platform model plane; measurement probes always run on the real vendor model.

Where the platform stands today

Beta. Verified in staging. Honest about the gates still open.

Beta

verified in staging

no production environment is claimed

4,080

automated test cases

3,064 backend · 834 frontend · 182 E2E

measured AI surfaces

ChatGPT · Claude · Gemini · Grok · Google AI Mode

163

brands in the benchmark catalog

99 fashion · 47 banks · fintech, tech, lodging

205

public API operations

under a locked OpenAPI contract

Implemented

Audit, Fix Engine, Content Optimizer, commerce protocols, feed ingestion, SaaS tenancy, BYOK, Growth Cockpit, saved analyses, automations, mention quality, spend estimate and confirm.

Staging-verified

Services, migrations and the account catalog re-verified 2026-07-30; every deployed image carries its Git commit. Data plane and GCP preflight (13 of 13) last re-run 2026-07-20.

Pilot-gated

GEO Intelligence (benchmark runs, automations) and LLM Discovery are enabled in staging but gated for customer beta: cost-capped banking proof, report reconciliation, reliability and two-week operating gates are open.

No production environment is claimed. Beta pilots run on the staging platform. Code inventory as of 2026-09-10; last incremental staging verification 2026-07-30.

Questions

Asked before every pilot.

Which AI surfaces does AgentLens measure?: ChatGPT, Claude, Gemini and Grok through their vendor APIs, and Google AI Mode read as a live search surface. Perplexity is not measured because no real adapter exists for it.
Does AgentLens change my website?: No. It generates deployable artifacts such as JSON-LD, llms.txt, robots.txt and meta tags. You review them and your own pipeline deploys them.
How is the readiness score calculated?: Four pillars at 25% each: Schema Coverage, Content Structure, Agentic Brand Familiarness (Share of Model) and Shopping Assistant. Each is scored 0–100 with a 95% confidence interval and its sample size.
Will I know the cost before a run?: Yes. Every run starts with a spend-free estimate that shows pinned models, expected calls and a hard cap at 115% of the estimate. Nothing runs until you confirm.
Is AgentLens generally available?: It is in beta, verified in a shared staging environment. Beta pilots run there; no production environment is claimed.

Check it yourself

Request a report, read the evidence, decide.

We reply within five business days.

Request your report