The Method

The infrastructure agents use to find, evaluate, and choose your business.

Three scores. One question. When an AI agent evaluates your business with no human in the loop, what does it find and what can it actually do with you? Most businesses have never thought to ask.

Overview

How the three scores build on each other.

Agents only take action on businesses they can first find, understand, and verify. ARS gets you in the room. AAS decides whether you transact. AVS is the single number that expresses both.

The baseline: The commercial internet averages ARS 54. Most businesses are parseable. Almost none are optimized for agent evaluation. Zero companies in our scored dataset have reached Agent-Native.

Step 1 · Discovery

ARS

Agent Readability Score · 0–100

"Do agents know you exist and do they trust what they find?"

Step 2 · Action

AAS

Agent Action Score · 0–100

"Once agents find you, can they actually do anything with you?"

Composite · Verdict

AVS

Agent Viability Score · weighted

"Taking it all together, are you winning or losing in the agent economy?"

The key distinction: The site that wins with humans is not always the site that wins with agents. A beautifully designed homepage with compelling copy can score lower than a sparse, unglamorous page that is properly structured for machine reading. Agents do not read your paragraphs. They look for signals, and most sites are not sending the right ones.

Five bands · one scale

Range	Band	What it means
0 – 19	Agent Invisible	No meaningful presence in agent decision-making. Losing addressable pipeline invisibly.
20 – 39	Agent Detected	Agents find you but cannot verify or transact. Commercially inert in autonomous evaluations.
40 – 59	Agent Considered	Agents include you in research but route to readier competitors at the decision layer.
60 – 79	Agent Ready	Agents reliably recommend and transact with you. Certification threshold. Advantage compounds.
80 – 100	Agent Native	Default agent recommendation. Citation equity compounds into retrieval patterns for years.

Explore the five scoring bands

Every score resolves to one of five bands. Drag the marker to see what agents can and cannot do with a business at each level.

INTERACTIVE · DRAG THE MARKER OR TAP A BANDSALESFORCE.COM SITS AT AVS 24

salesforce.com

CERTIFICATION 60+

020406080100

Agent Detected

20 – 39

Agents find you but cannot verify or transact. Commercially inert in autonomous evaluations.

ARS

0 – 100

Step 1 · Discovery & comprehension

Agent Readability Score: can AI agents find and understand you?

"Can an autonomous AI agent find you, understand you, and trust what it finds, with no human involved?"

Why ARS comes first: Think of it as the entrance requirement. An agent that cannot confidently identify who you are and what you do will not spend more time on you. It moves to the next result. No ARS, no evaluation. A high AAS does not matter if the agent never got that far.

ARS measures how readable your business is to machines. Not how compelling your content is, how structured it is. The homepage your designer is proudest of may score lower than a plainer page built for automated evaluation. Agents do not read. They scan for signals.

A low ARS means agents encounter you but cannot classify what you are or confirm you are legitimate. Present in the crawl, absent in the evaluation. Get ARS right and you are in the room. What happens next is AAS.

ARS dimensions: what agents need to find you

Entity and identity

Can agents definitively establish who you are and what category you operate in?

Content legibility

Can agents extract structured, verifiable information from your digital presence?

Access and permissions

Have you explicitly declared your policy on agent access and interaction?

Discovery configuration

Is your infrastructure configured to be reliably found by automated crawlers?

Full dimension breakdown and prioritized remediation roadmap are delivered with every Comprehensive Report. This summary tells you what matters, not how to fake it.

AAS

0 – 100

Step 2 · Evaluation & action

Agent Action Score: can AI agents actually use your product?

"Once an AI agent understands your business, can it take meaningful commercial action without any human intervention?"

The insight most businesses miss: If completing a commercial action with your business requires a human at any point, a sales call, a demo request, a form that triggers an email, an agent cannot transact with you. It will not wait. It moves to a competitor who has built for autonomous interaction. Most businesses have never tested whether an agent can actually use their product.

AAS measures whether an autonomous agent can actually do something with your business. Not whether you have a sales team, whether an agent could skip the sales team entirely, evaluate your offering, and initiate a transaction without any human involved.

Low AAS means you show up in the research and disappear at the decision point. Agents route to competitors they can act on. High AAS means you are one of the few businesses an agent can take from evaluation to transaction without involving a single human.

AAS dimensions: what agents need to act

Pricing accessibility

Can agents determine your pricing without human involvement?

Action availability

Can agents initiate a commercial action with your business autonomously?

Credibility signals

Can agents verify your legitimacy through structured, machine-readable proof?

Protocol readiness

Is your infrastructure configured for direct, autonomous agent interaction?

The average AAS across our scored dataset is 18. Most businesses can be found. Almost none can be used. That gap is where your revenue is silently going.

AVS

Composite

Composite · the verdict

Agent Viability Score: are you ready for the agent economy?

"Taking both discoverability and actionability together, how commercially prepared is this organization for autonomous AI agent procurement?"

AVS is the composite verdict, a single number predicting how your business performs when agents are doing the buying. The weighting is proprietary and deliberate. It favors actionability over discoverability, because an agent that can find you but cannot transact with you is worth exactly nothing to your revenue. The formula is not published here. It is delivered with every paid assessment.

Agent Readiness Report · salesforce.com · April 2026

ARS

41 / 100

Agent Considered

AAS

11 / 100

Agent Invisible

AVS

24 / 100

Agent Invisible

Diagnosis: This business is broadly readable. Agents can identify its category and core offering. However, the action layer is largely absent. The business appears in consideration sets but is systematically excluded at the decision layer in favor of competitors with higher AAS scores.

The Comprehensive Report includes manual review, full-site crawl, live agent simulation across four buying scenarios, and a competitive context probe where relevant. It also includes a complimentary re-scan after your fixes ship to verify the new score.

See where you stand

See where you stand.

Your ARS score across all ten dimensions, benchmarked against your industry, with your top three priority fixes. Free, same day.

Get your free ARS score Or email [email protected]

Free preview No commitment Delivered same day