Conductor

Search used to end with a click. Somebody typed a question, saw ten links, chose one, and arrived on your site where your analytics recorded them.

Increasingly it ends with an answer. The assistant reads several sources, synthesises a reply, and the person never visits anything — which means the moment your brand was or was not mentioned is invisible to every tool you own.

Conductor's argument is that this layer needs its own measurement, and that measurement is what its platform is now organised around.

Conductor: Visibility In AI Answers

• Tracks your brand across ChatGPT, Gemini, Copilot and Claude as well as traditional search results.
• Four products: Intelligence, Creator, Monitoring and AgentStack, covering measurement through to execution.
• Monitors how AI bots crawl your site, which is a genuinely new category of site health.
• A free AI visibility report as the entry point, which is a sensible way to see the gap before buying.
• Enterprise-scale — a decade of SEO intelligence repositioned around AEO, with comparison pages against the usual suspects.

What Is Answer Engine Optimisation?

Conductor positions itself as the enterprise AEO platform, covering the full lifecycle from AI visibility tracking through content creation to real-time site health.

Answer engine optimisation is the practice of being cited and recommended inside AI-generated answers rather than ranking in a list of links. The distinction sounds academic and is not.

Ranking is a position you can observe. Being included in an answer is probabilistic — ask the same question twice and you may get different sources, because the assistant is generating rather than retrieving a fixed list.

That changes what measurement means. AEO metrics are frequency distributions, not positions: how often you appear across many runs of many phrasings, which is a fundamentally noisier signal than rank tracking ever was.

It also changes what you optimise. Keyword targeting matters less than being the source that states a fact clearly and citably, because the assistant is looking for something to draw on rather than something to link to. Content tooling such as MarketMuse approaches the same shift from the editorial side.

Ranking is a position you can observe. Being included in an answer is probabilistic — ask twice, get different sources.

Why AI Visibility Is Measured Differently

The mechanics of measuring AEO are worth understanding before you evaluate anyone's numbers, including Conductor's.

There is no ranking API for an assistant. Measurement is done by asking questions repeatedly and recording what comes back, which means every vendor's numbers depend on which questions they ask, how many times, in which phrasings and from which regions.

Two platforms measuring the same brand can therefore report materially different visibility, and neither is lying. They asked different things.

This makes methodology the most important thing to interrogate, and the least likely to be published in detail. The questions you actually care about are the ones your buyers ask, and a generic prompt set will not contain them.

The practical guidance is to treat absolute visibility percentages sceptically and trends within one platform seriously. A number that moves consistently on a fixed method tells you something even when the number itself is not comparable to anyone else's.

Four Products And What Each Covers

Conductor splits into Intelligence, Creator, Monitoring and AgentStack, and the split is more meaningful than most enterprise product taxonomies.

Intelligence is measurement — visibility across ChatGPT, Gemini, Copilot, Claude and traditional search, connected to traffic, conversions and revenue.

Creator is generation, producing and optimising content for both search and LLMs. This is the most crowded of the four and the least differentiated, since generating content is something a great many products do.

Monitoring is technical site health with an AI-specific twist: 24/7 tracking of how AI bots crawl your site, with alerts and prioritised fixes. That is genuinely new territory rather than repackaged crawling.

AgentStack is the newest and vaguest — LLM apps, developer tools and turnkey agents for running AEO workflows. Worth understanding concretely before buying, since agent suites are the easiest thing in enterprise software to describe and the hardest to evaluate.

The unifying claim is a single data engine behind all four, which is the standard argument for buying a suite rather than four tools. It is a real advantage when it holds — visibility, content and crawl health genuinely inform each other — and a familiar disappointment when the products turn out to be separately acquired things sharing a login.

The practical way to test it is to ask how a finding in one product produces an action in another. If Monitoring detects that an AI crawler is blocked, does Creator know not to bother optimising those pages? A suite that answers that concretely is a suite; one that answers it in principle is a bundle.

• Interrogate the measurement methodology before trusting any visibility percentage.
• Track trends within one platform rather than comparing across vendors.
• Make sure the prompt set includes the questions your buyers actually ask.
• Take the free visibility report first — it sizes the gap without a commitment.
• Check AI crawler access before optimising content nothing can read.

Monitoring How AI Bots Crawl You

Conductor Monitoring is described as always-on tracking of how AI bots crawl your site, with real-time alerts and prioritised fixes so visibility issues never become revenue problems.

This is the most concrete and least fashionable part of the platform, and probably the most immediately useful.

AI crawlers are not search crawlers. They arrive from different agents, respect robots directives differently, and are frequently blocked by the same infrastructure that protects you from scrapers — often by a security team that made a sensible decision without knowing the consequence.

The result is a specific and common failure: excellent content that assistants cannot read, producing zero visibility for reasons that have nothing to do with the content at all.

Checking crawler access is therefore the first thing to do, before any AEO strategy work. It is a technical question with a yes or no answer, and getting it wrong makes every other investment pointless — the same logic that makes documentation platforms like GitBook worry about machine readability at all.

The Limits Of Optimising For Answers

There are structural limits to AEO that no platform can solve, and knowing them prevents unrealistic expectations of any vendor.

You cannot verify a model's reasoning. Why an assistant cited one source and not another is not observable, and vendors offering explanations are inferring, not reporting.

Model behaviour changes without notice. A model update can reshape which sources are favoured overnight, and unlike a search algorithm update there is no announcement, no documentation and no community consensus about what changed.

Attribution is close to impossible. Someone who read about you in an answer and typed your name into a browser three days later arrives as direct traffic, and no attribution platform — including specialists such as Cometly — can see the assistant that started it.

None of this makes measurement worthless. It makes it directional, which is the same status search rank data had for years before people started treating it as gospel.

There is also a self-defeating dynamic worth naming. If everyone optimises specifically to be cited, the assistants will adjust for exactly that, and the tactics that work now are the ones most likely to stop working — which is the entire history of search optimisation repeating at higher speed.

What survives that cycle is the same thing that survived the last one: being genuinely the best source on a specific question. That is slower than tactics and it is the only durable position, whichever surface is doing the summarising. Content planning tools such as SEO.AI and editorial judgement matter more here than any measurement platform.

Getting Started With AEO Measurement

The mistake with a new measurement discipline is buying the platform before establishing what you would do differently with the answer.

This sequence resolves the cheap technical questions first and reaches procurement only once there is something to act on.

1. Check whether AI crawlers can actually access your site, before anything else.
2. Run the free visibility report to size the gap without committing.
3. Write down the twenty questions your buyers genuinely ask, in their words.
4. Ask those questions manually across two or three assistants and record what appears.
5. Compare that against what a platform reports, and ask how its prompt set was built.
6. Fix the pages where you should obviously be cited and are not, then re-measure.

AEO Words Explained

The vocabulary of answer engine optimisation, without the marketing.

undefined

undefined

undefined

undefined

undefined

undefined