What does AI say about your business?
Right now, AI systems are answering questions about your business, and agents are trying to use your website. The audit shows you exactly what they see, say, and can do. Evidence, score, roadmap.
$2,500
DELIVERABLEScore, evidence & roadmap
RESPONSEWithin 1 business day
DELIVERYUp to 1 week
Eight dimensions,
evidence-based.
Every point on the scorecard maps to a test, an observed behavior, or an implementation artifact. No AI-readiness theater.
- 20 high-intent prompts, run across 8 AI platforms and search surfaces
- API-based testing with a neutral system prompt: no memory or personalization bias
- Google AI Overviews & Microsoft Copilot captured live, logged out, screenshots kept
- Factual accuracy: services, pricing, policies, proof
- Browser-agent task testing: research, quote, book, retrieve
- Crawlability, structured data & source-of-truth review
- Benchmarked against 2 competitors: yours, or the AI-optimized ones we find
- Your Agent Readiness Score, with a prioritized roadmap
0/100
- AI Discoverability12/15
- Machine Understanding13/20
- Accuracy & Consistency9/15
- Structured Data7/10
- Agent Accessibility8/15
- Trust & Authentication3/5
- Transaction Capability6/10
- Agent Actions10/10
Every surface a buyer asks.
The same 20 prompts run on every platform below, the same day. Six by API with a neutral system prompt; two search surfaces with no public API, captured live from a clean, logged-out session. Screenshots kept for all of it.
- APIChatGPTOpenAI, web search on
- APIClaudeAnthropic, web search on
- APIPerplexitySonar, search-native
- APIGeminiGoogle, grounding on
- APIGrokxAI, live search on
- APIMeta AILlama models
- LIVEAI OverviewsGoogle AI Overviews & AI Mode, captured live
- LIVECopilotMicrosoft Copilot & Bing, captured live
Voice assistants. Siri, Alexa, and Google Assistant now hand their answers to the same models above (Gemini, ChatGPT, and vendor models), so what they say tracks what those platforms say. We note them in the report; we don't run them as separate suites.
Scored against six levels.
The score plots your business on our six-level maturity model, from Discoverable to Actionable. See the full framework.
- LV.1DiscoverableCan agents find you?
- LV.2UnderstandableCan they interpret your offer?
- LV.3TrustedCan they verify you?
- LV.4AccessibleStable interfaces?
- LV.5TransactableCan agents pay?
- LV.6ActionableReal workflows, safely?
Request your audit.
A few qualifying questions to help us better understand your digital footprint. Submitting the form is no obligation: we'll confirm fit and scope with you before anything is started or billed. We respond within 1 business day; the full audit takes up to 1 week.
Questions from the field.
What exactly does the Agent Readiness Audit include?
We run 20 high-intent buying and comparison prompts across eight AI platforms and search surfaces (ChatGPT, Claude, Perplexity, Google Gemini, Grok, and Meta AI by API; Google AI Overviews and Microsoft Copilot captured live), test factual accuracy about your services, pricing, and policies, benchmark you against 2 competitors, run browser-agent tasks against your site (research, quote, book, retrieve), audit crawlability, structured data, and source-of-truth consistency, and deliver your Agent Readiness Score with screenshots, findings, and a prioritized roadmap.
Which AI platforms do you test?
Six by API: ChatGPT (OpenAI), Claude (Anthropic), Perplexity, Google Gemini, Grok (xAI), and Meta AI (Llama). Two search surfaces that have no public API, Google AI Overviews and Microsoft Copilot (Bing), are captured live from a clean, logged-out session. Voice assistants such as Siri and Alexa now hand their answers to these same models (Gemini, ChatGPT, and vendor models), so what they say tracks what the platforms above say; we note that in the report rather than running them as separate suites.
How do you keep the AI test results unbiased?
Every prompt runs through each platform's API with a neutral system prompt, not through a logged-in chat interface. That means no chat memory, no personalization, and no prior-conversation bias skewing what the models say about you. For the two surfaces without a public API (Google AI Overviews and Microsoft Copilot), we capture the live results from a clean, logged-out session and keep the screenshots. Either way, the results reflect what a fresh prospect's AI assistant would actually report.
Which competitors do you test against?
Two. You're welcome to suggest them and we'll factor that in, but if our research finds competitors that are better AI-optimized than the ones you named, we'll use those instead. The benchmark only means something if it's measured against whoever is actually winning the AI answers in your market.
How much does the audit cost?
The audit is a fixed fee of $2,500. No hourly surprises, no scope creep. One price, one complete deliverable.
How long does the audit take?
We respond to every request within 1 business day to confirm scope and kickoff. From there, the audit takes up to 1 week to complete, depending on the competitors and workflows you want tested.
Do I need to give you access to my website or analytics?
No access is required for the core audit; we test what AI systems and agents can see and do from the outside, exactly as they would. Optional read-only access to Google Search Console or analytics makes the roadmap sharper.
What happens after the audit?
The audit stands alone: you can hand the roadmap to any team. Most clients ask us to execute the priority fixes (website modernization, AI visibility optimization, or agent interfaces), and audit findings scope that work precisely.