Hey there,

Jason Barnard published a piece on Search Engine Land last week that triggered a naming war across the SEO community. His argument: AEO (Answer Engine Optimization) is an incomplete framework. The replacement he's proposing is AAO — Assistive Agent Optimization.

The reasoning cuts to a real shift in how AI systems work. AEO was built for a world where users ask AI questions and get answers. AAO is built for a world where users give AI agents tasks and the agent picks a winner. "Find me a competitive intelligence platform for my 50-person team and book a demo with the best one." The agent doesn't show a list. It acts.

Barnard maps the evolution across four stages:

SEO — be found (in a list of links). AEO — be the answer (to a direct question). AIEO — be the recommendation (when AI summarizes options). AAO — be chosen (when the agent acts without showing alternatives).

Each stage absorbs the one before it. And the numbers back up the urgency. Gartner predicts 40% of enterprise applications will integrate task-specific AI agents by end of 2026, up from under 5% in 2025. In their best-case projection, agentic AI could drive roughly 30% of enterprise application software revenue by 2035 — over $450 billion. Gartner's Anushree Verma described the shift as moving enterprise apps "from tools supporting individual productivity into platforms enabling seamless autonomous collaboration."

That's not a 2030 problem. That's a this-year problem.

But here's my pushback on the framing: for 95% of B2B companies that haven't done basic AEO, telling them it's outdated is premature. You can't be chosen if you're not in the consideration set. The 5-Category AEO Framework — Citation Readiness, Content Structure, Authority Signals, Technical Accessibility, Semantic Clarity — still determines whether agents include you at all.

Do AEO first. Then build the AAO layer on top.

This newsletter walks you through both — the foundational AEO work and the AAO additions that prepare your brand for agent-driven selection.

When we test a client's brand in our multi-LLM analysis workflow, we ask two types of questions.

AEO question: "What are the best competitive intelligence tools for B2B companies?" AAO question: "Choose one competitive intelligence tool for a 100-person SaaS company and explain why."

The same brand that appears in 70% of AEO responses wins only 30% of AAO selections. Recommendation visibility and selection authority are different signals.

Here's why. When an AI engine answers an AEO question, it compiles a list. Multiple brands can appear. The bar is: are you relevant, authoritative, and well-structured enough to be mentioned? When an AI agent answers an AAO question, it picks one. The bar is: does the agent have enough confidence in your brand — pricing, outcomes, fit, differentiation — to act on it without showing alternatives?

Barnard frames this through what he calls the "Algorithmic Trinity" — the three components every AI system uses to make decisions: large language models (for synthesis), knowledge graphs (for entity understanding), and traditional search (for discovery). His argument is that AAO is the only framework that covers all three legs. "Optimizing for one while ignoring the others," Barnard wrote, "is like sitting on a three-legged stool with two legs missing."

The practical implication for B2B content teams: your content needs to work at two levels simultaneously. It needs to be findable and citable (AEO), and it needs to contain the specific decision-making signals that let an agent select you with confidence (AAO).

Here's how to build both layers.

Before optimizing anything, test your current visibility across both AEO and AAO prompts.

Open ChatGPT, Claude, Perplexity, and Gemini. Ask each the same two prompts:

AEO prompt: "What are the best [your category] solutions for [your target market]?"

AAO prompt: "I'm the [target buyer title] at a [company size] [industry] company. I need to [solve the problem your product solves]. Research options and recommend the best approach."

For each platform and each prompt, record:

Run this with 3-5 different task prompts covering your core services.

What you'll likely find: Most B2B companies pass AEO for 1-2 platforms and fail AAO across all four. The gap between "mentioned" and "chosen" is your optimization roadmap.

Here's what a typical audit looks like for a mid-market B2B company:

Platform: ChatGPT — AEO: Mentioned (3rd in list) — AAO: Not selected. Agent chose competitor with clearer pricing and case studies. Platform: Claude — AEO: Not mentioned — AAO: Not mentioned. Content structure too thin for Claude's deeper analysis. Platform: Perplexity — AEO: Mentioned (2nd) — AAO: Selected, but with caveats about limited public information. Platform: Gemini — AEO: Mentioned (4th) — AAO: Not selected. Agent cited competitor's comparison page as primary decision input.

That pattern — partial AEO visibility, near-zero AAO selection — is the norm, not the exception. The fix isn't one thing. It's a systematic stack of improvements, starting with the AEO foundation and layering AAO signals on top.

Tip: Save your audit results in a spreadsheet and repeat monthly. The prompts stay the same. The tracking becomes your leading indicator of whether content changes are working — faster feedback than waiting for organic traffic data.

For the automated version, start here: revenueexperts.llmauditpro.com scores your content across all five AEO categories in 60 seconds. It won't test AAO selection directly, but it tells you whether the foundation is solid enough for agents to even consider you.

If you failed the AEO prompts — if agents don't mention you at all — the AAO layer is irrelevant. You need the basics in place.

The five categories that determine whether AI engines can find, parse, and cite your content:

Citation Readiness. Does your content contain specific, quotable statements? Agents cite content that makes factual claims with data attached. "We reduced client research time by 90%" is citable. "We help companies work smarter" is not. Go through your top 10 pages and count the number of specific, data-backed claims on each. If a page has fewer than three citable statements, it's invisible to agents.

Content Structure. Are your pages organized with clear headings, semantic HTML, and logical information hierarchy? AI crawlers process structured content far more reliably than walls of text. Every service page needs: H1 (what it is), H2s (how it works, who it's for, what it costs, what results to expect), and H3s for granular detail under each section. FAQ sections with proper FAQPage schema markup are among the highest-impact structural additions you can make — they directly map to the question-answer format AI engines use.

Authority Signals. Can agents verify that you're a credible source? This means author bios with credentials, published case studies with named outcomes, third-party mentions and backlinks from recognized sources, and consistent entity information across your website, LinkedIn, Google Business Profile, and industry directories. Barnard's "entity home" concept is relevant here — your canonical domain needs to be the unmistakable anchor that tells AI systems who you are and what you do.

Technical Accessibility. Can AI crawlers actually read your content? JavaScript-rendered pages, bot-blocking configurations, and slow-loading sites all reduce what agents can access. Check whether your key pages are accessible to GPTBot, ClaudeBot, and CCBot crawlers. If your robots.txt blocks them, you're blocking your own visibility. Our AI SEO Blueprint course covers all 39 technical modules of crawler management, including the specific directives for each AI engine's bot.

Semantic Clarity. Do AI systems understand what entities your content discusses and how they relate? This means consistent terminology (don't call it "competitive intelligence" on one page and "market analysis" on another), clear entity relationships (your company → your services → your results → your team), and schema markup that reinforces those relationships.

If your AI Visibility Score comes back weak on any of these five categories, fix them before moving to AAO. The foundation has to hold weight before you add another floor.

Here's where AEO and AAO diverge. AEO content says "we exist and we're good at this." AAO content says "here's exactly why you should pick us over every alternative, with enough specificity that a machine can make that call."

An agent selecting one brand over another needs:

Explicit pricing or pricing ranges. Agents can't recommend you if they don't know what you cost. "Contact us for pricing" is an AAO failure. "Starting at $2,500/month for mid-market companies" gives the agent a decision input. You don't need to publish your full rate card — ranges, tiers, and starting points work. But something concrete has to be there.

We tested this directly with a client's service pages last month. Their content appeared in ChatGPT and Perplexity answers (AEO pass) but wasn't selected when agents were asked to choose one option (AAO fail). One of three gaps identified: no visible pricing. After adding pricing ranges to two key pages, Perplexity and Claude began recommending them by name within three weeks. The infrastructure for this kind of content optimization is what our AI workflow systems handle at scale.

Direct competitor comparisons. If you don't compare yourself to alternatives, the agent has to figure out positioning on its own — and it often gets it wrong or defaults to the competitor that does provide comparison content. Create comparison pages: "Us vs. Competitor A," "Us vs. Competitor B," and a category comparison showing where you fit. Be honest about tradeoffs. Agents trust content that acknowledges weaknesses more than content that claims to be best at everything.

Specific outcomes per use case. "We improve marketing efficiency" tells an agent nothing. "B2B SaaS companies with 50-200 employees typically see 80-90% reduction in content production time and a 5-10x increase in output within 60 days" gives the agent selection criteria. Map your results to buyer segments. Different company sizes, industries, and use cases should have their own outcome data.

Structured "who this is for" and "who this isn't for" content. Agents making selection decisions need fit signals. A page that says "Best for mid-market B2B companies with marketing teams of 5-50 people" helps an agent match your offering to the user's stated situation. A page that says "We serve businesses of all sizes" gives the agent no differentiation signal.

Transaction or engagement pathways. Barnard's point about "actionability" is worth paying attention to here. An agent that selects your brand may also need to take the next step — booking a demo, starting a trial, requesting a proposal. If that pathway is buried behind three clicks and a contact form with 12 fields, the agent can't execute. Clear, accessible CTAs with structured data markup (schema for offers, services, contact points) make your brand not just selectable but executable.

Common mistakes that kill AAO readiness:

Most companies make the same three errors when they start optimizing for agent selection. Spotting them early saves months of wasted effort.

Mistake 1: Hiding pricing behind "Contact Sales." The instinct to gate pricing is understandable — you want a conversation, not a price comparison. But agents can't start a conversation. They can only work with what's publicly available. Every competitor that publishes pricing ranges gives agents a decision input you're withholding. You don't need exact numbers. "Implementation starts at $25K for mid-market companies, with monthly retainers from $2,500" is enough.

Mistake 2: Writing for humans only. Your homepage might read beautifully to a CMO, but if it uses metaphors instead of facts, an agent has nothing to work with. "We turn marketing chaos into revenue growth" doesn't tell an agent anything. "AI operating systems for B2B companies with 50-500 employees, specializing in content automation, competitive intelligence, and AEO optimization across ChatGPT, Claude, Perplexity, and Gemini" tells an agent exactly what you do, who you serve, and what platforms you work with. Both sentences can coexist on the same page — the human-friendly version up top, the agent-readable version in structured data, meta descriptions, and the body text.

Mistake 3: Optimizing for one AI platform. ChatGPT has the largest consumer user base, so companies often test and optimize only for ChatGPT. But Claude, Perplexity, and Gemini each process information differently and weight different signals. A brand that wins on ChatGPT might not even appear on Claude. The multi-LLM analysis methodology we use across all client deployments exists because single-platform optimization creates blind spots. Always test across all four.

Barnard repeatedly emphasizes the "entity home" — the single canonical page or domain that unmistakably defines your brand to AI systems. If the agent can't anchor your brand to a clear, authoritative entity, it will choose a competitor it understands better.

For most B2B companies, your entity home is your homepage + your primary service pages. Here's what makes an entity home strong:

Consistent naming. Your brand name, tagline, and core description should be identical across your website, LinkedIn company page, Google Business Profile, Crunchbase, industry directories, and any other platform where your brand appears. Inconsistency creates entity confusion — the agent isn't sure whether "Revenue Experts AI" and "Revenue Experts" and "RevX" are the same company or three different ones.

Organization schema on every page. Implement Organization schema markup on your homepage with your legal name, URL, logo, founding date, address, social profiles, and a description. This is the machine-readable version of "who we are." Our Context Engineering Masterclass covers how to build the information architecture that both AI agents and traditional crawlers use to understand entity relationships — 12 modules, 60+ exercises, 50+ reusable templates.

Author entities for your content. Every article, guide, and case study should have a named author with a bio page that includes credentials, expertise areas, and links to external profiles. Google's E-E-A-T framework and AI agent trust signals both rely on verifiable author entities. If your blog posts say "by Admin" or have no author attribution, you're weakening your entity signals at the content level.

Corroboration across platforms. The more places AI systems can verify your brand information — and the more consistent that information is — the higher confidence the agent has in selecting you. Wikipedia, industry directories, media mentions, podcast appearances, conference speaking pages, and social profiles all contribute to entity corroboration. The AI Visibility Score includes entity consistency analysis as part of its Semantic Clarity category.

This is the content type that separates AEO-ready brands from AAO-ready brands. Most B2B sites have "what we do" content. Almost none have "why pick us over the alternatives" content structured for machine consumption.

What to create:

Category definition pages. "What is [your category]?" pages that define the space, explain what buyers should look for, list evaluation criteria, and position your approach. When an agent researches a category, these pages become the framework it uses to evaluate options — including yours.

Versus pages. One page per major competitor: "[Your brand] vs. [Competitor]." Structure each with: use case comparison, feature comparison, pricing comparison (ranges are fine), ideal customer profile for each, and an honest assessment of where each option is stronger. These pages perform well in both traditional search and AI agent queries because they directly match the comparison structure agents use internally.

Use-case-specific landing pages. Instead of one generic services page, create pages for each buyer segment: "[Your solution] for SaaS companies," "[Your solution] for professional services firms," "[Your solution] for e-commerce." Each page should include segment-specific outcomes, relevant case studies, and pricing context for that segment. When an agent is asked to find a solution for a specific type of company, these pages give it a direct match.

Decision guides. "How to choose a [your category] solution: the buyer's checklist." Position your evaluation criteria as the industry standard. Include your brand as one option among several (agents trust content that doesn't only promote itself). Structure with comparison tables, scoring rubrics, and clear recommendations for different scenarios.

This content strategy serves both channels. For AEO, it creates citation-ready, structured information that agents reference when answering questions. For AAO, it provides the specific decision inputs — pricing, comparisons, fit criteria — that let agents select with confidence.

Tip: Structure your versus pages with this template. Every comparison page should follow the same format — it makes them easy to produce and it gives AI agents a consistent parsing structure:

Opening paragraph: What both solutions do (category context). Section 1: Who each solution is best for (company size, industry, use case). Section 2: Feature comparison (table format — agents parse tables extremely well). Section 3: Pricing comparison (ranges are fine, but be specific about what each tier includes). Section 4: Customer outcomes (metrics from each — cite specific numbers where available). Section 5: Your honest recommendation (which situations favor you, which favor the competitor). Closing: CTA with a clear next step (demo, audit, trial).

Produce one of these per major competitor. Five versus pages covering your top five competitors gives agents a complete decision framework — with your brand as the author setting the evaluation criteria.

One more content type worth building: the "three research modes" content.

Barnard identified three ways users and AI systems discover brands: explicit research (direct queries — "best CI tools"), implicit research (background queries agents run without the user seeing them), and ambient research (proactive AI recommendations without any query at all — like Google Discover or AI-generated suggestions in Chrome).

Each mode needs different content. Explicit queries need category pages and comparison content. Implicit queries need deep service pages with structured data and clear entity signals — the kind of pages agents pull from when building context behind the scenes. Ambient recommendations need consistent publishing velocity, topical authority signals, and high E-E-A-T scores.

If you're only optimizing for explicit research (the queries you can see), you're missing two-thirds of the discovery surface. Our AI Revenue Intelligence Systems address all three modes by automating the content, monitoring, and competitive tracking across each.

One of the problems with optimizing for AI agents is that each platform processes information differently. ChatGPT might recommend you while Claude ignores you. Perplexity might cite your comparison page while Gemini cites a competitor's.

This is why we use multi-LLM cross-validation across all our client work — running the same prompts through ChatGPT, Claude, Gemini, and Perplexity, comparing results, and optimizing for the gaps.

How to set up your own testing cycle:

Monthly AEO test. Run 5 category-level questions ("What are the best X tools?") through all four platforms. Track: mentioned (yes/no), position in the list (1st, 2nd, 3rd+), and attributes cited. Log results in a spreadsheet. Over 3 months, you'll see whether your AEO visibility is improving, stable, or declining.

Monthly AAO test. Run 5 task-oriented prompts ("Choose one X for a [specific buyer profile]") through all four platforms. Track: selected as primary (yes/no), selected as one of multiple (yes/no), not mentioned (yes/no), and reasons cited for selection or rejection. This is the harder test. Most brands fail it initially.

Competitor tracking. Run the same prompts with competitor-specific variations. "Compare [your brand] and [competitor] for [use case]." Track how agents describe the comparison. What advantages do they assign to each? What language do they use? This reveals exactly what content you need to create or strengthen.

Content response cycle. When you find a prompt where a competitor wins and you don't, analyze what the agent cited. Visit the competitor's page. Identify the specific content that gave the agent confidence to select them. Then create content that matches or exceeds those decision signals. Test again 3-4 weeks later. AI engines reprocess content faster than traditional search — you'll often see changes within a few weeks of publishing improved content.

The Context Prompting Engineering course covers how to design these testing frameworks systematically — including prompt templates for competitive analysis, scoring rubrics for visibility tracking, and the multi-agent architectures that automate the testing cycle.

Barnard's most forward-looking argument is about actionability. Right now, most AI agents recommend. Soon, many will execute — booking demos, starting trials, placing orders, scheduling consultations. Google's MCP (Model Context Protocol) rollout is building the infrastructure for exactly this: agents that can interact with services programmatically.

Barnard wrote on his own site that "the Zero-Sum Moment isn't approaching. It's here." When an agent books a hotel for you, there's one choice. No list. No comparison. The brand that the agent trusts is the brand that gets the transaction.

For B2B companies, full transactional agent integration is still 12-18 months away for most use cases. But the preparation starts now:

Simplify your conversion pathways. If booking a demo requires filling out a 12-field form, an agent can't complete it on behalf of a user. One-click booking links, Calendly-style scheduling, and simple trial signups with minimal friction are agent-friendly.

Implement structured data for your services. Service schema, Offer schema, and ContactPoint schema tell agents what you offer, what it costs, and how to engage. This is the machine-readable version of your sales page.

Create API-accessible entry points. This is further out for most companies, but worth noting: brands that expose booking, scheduling, or quoting functionality through APIs will be the first ones agents can transact with directly.

Monitor MCP and agent protocol developments. Google's MCP, Anthropic's MCP, and OpenAI's function calling are all creating the standards for agent-to-service interaction. Following these developments now — even if you're not implementing them yet — keeps you ahead of the curve.

Here's the priority sequence. Don't skip ahead — each phase builds on the one before it.

Days 1-30: AEO foundation.

Run the 4-platform audit from Step 1. Record your baseline across all AEO and AAO prompts. Run the AI Visibility Score audit on your homepage and top 5 service pages. Fix technical accessibility issues first — robots.txt blocking, JavaScript rendering problems, missing meta descriptions. These are binary: either crawlers can read your page or they can't. Add FAQPage schema markup to your top 3 service pages. Go through each page and add at least 3 citable, data-backed statements. Remove or rewrite any "Contact us for details" language around pricing — add ranges or starting points.

Days 31-60: AAO signal layer.

Create your first 3 versus pages (top 3 competitors). Add "who this is for" and "who this isn't for" sections to every service page. Implement Organization schema on your homepage. Build or update author bio pages for everyone publishing content. Create one category definition page ("What is [your category]?"). Run the 4-platform audit again. Compare to Day 1 baseline. You should see AEO improvements; AAO improvements take longer.

Days 61-90: Decision-support content and testing cycle.

Create 2-3 use-case-specific landing pages for your top buyer segments. Build one decision guide ("How to choose a [your category] solution"). Set up your monthly multi-LLM testing cycle as a recurring calendar event. Implement Service schema and Offer schema on your service pages. Run your third 4-platform audit. By now, AEO visibility should be measurably stronger across all platforms. AAO selection should be appearing on at least 1-2 platforms for your strongest use cases.

Beyond 90 days: Continue the monthly testing cycle. Expand versus pages to cover all significant competitors. Start monitoring MCP and agent protocol developments for the transaction layer. Build comparison tables into existing content where agents need structured data to make selections.

The naming debate will continue. Barnard counted at least six competing acronyms for the same discipline — AI SEO, GEO, AEO, AIEO, LLM optimization, entity SEO — and argued AAO should replace them all. Whether the industry settles on AAO or something else, the underlying shift is real: AI systems are moving from showing options to making choices.

For B2B companies, the playbook is:

Start with AEO. Build AAO on top. Test relentlessly.

The brands that AI trusts in 2026 will be the brands agents select in 2027 and beyond. That trust is built now, one content decision at a time.

Sources referenced: Jason Barnard, Search Engine Land (Feb 25, 2026); Gartner press release (Aug 26, 2025) on enterprise AI agent adoption; Kalicube Algorithmic Trinity framework.

Until next week,

Elizabeta Kuzevska Co-Founder, Revenue Experts AI revenueexperts.ai | onlinemarketingacademy.ai Connect on X: @ekuzevska · Connect on LinkedIn

P.S. Last week's prediction came true — Google's February Discover update finished rolling out after 22 days, and the data from NewzDash shows real consolidation: 172 unique domains in the US Top 1,000 dropped to 158. We broke down the full B2B playbook in this week's Medium article. If you missed it, check the link in bio