Your buyers stopped searching. They started asking.

SignlLabs™ is GCW's proprietary AI Signaling practice. It helps brands earn visibility and recommendations across the answer engines people now turn to for answers: ChatGPT, Gemini, Claude, Copilot and Perplexity. It pairs the Maturity Method™ with strategic growth advisory, and the methodology is available for license.

AI Signaling is the practice of engineering how answer engines understand, trust and recommend a brand. The engines do not rank pages. They compose a single answer from the sources they already trust and hand the person that answer directly, often with no link at all. AI Signaling is the discipline of becoming one of those sources. The term, the Maturity Method and the Intent Category Framework were created by SignlLabs.

It starts with a Maturity Method™ audit. We run a defined query set against all five engines, score every observation, and roll the results into a Signaling Maturity Score™ from 0 to 100 with a full breakdown by dimension and by engine.

From there we fix what the brand controls: clear entity definition, consistent messaging, structured data. Then we build what the brand does not yet have, which is third-party validation through earned media, best-of lists, awards and expert coverage. We re-score on a quarterly cadence and adjust the plan against the numbers.

ChatGPT, Gemini, Claude, Copilot and Perplexity. Each engine is scored separately and never blended, because an average hides the engine that stopped recommending you.

Every query gets a 0, 1 or 2 on each engine. Not mentioned is a 0. Mentioned is a 1. Recommended is a 2. We run each query three times and record the median, because answer engines are not deterministic and a single run is not evidence.

A 0 to 100 benchmark of a brand's recommendability, scored across six weighted dimensions and reported per engine. It gives a baseline, a competitive comparison inside the category, and a way to measure progress every quarter. It is GCW's proprietary instrument, not an industry standard, which is the point: it is ours, and we can show you exactly how every number in it was arrived at.

Foundation, 24%, structured identity signals across owned and earned media. Is the brand clearly defined and machine-readable as a distinct entity? Endorsement, 20%, third-party authority citations: reviews, references, credentialed mentions. Do sources the engines trust validate this brand as a leader? Alignment, 16%, consistency between what the brand claims and what outside sources confirm. Presence, 16%, coverage density across the sources AI models retrieve from. Warmth, 12%, sentiment quality; how the brand is characterised when it does come up. Constancy, 12%, recency and freshness of the signals reaching training and retrieval.

The weights are why this is PR-led work: the heaviest lever after Foundation is Endorsement, and Endorsement cannot be built from a brand's own website.

The score maps to five stages, so a number becomes a status a client can act on. 0–20 Unseen. 21–40 Surfacing. 41–60 Standing. 61–80 Trusted. 81–100 The Answer. The ladder is what makes this a maturity model rather than an audit: you get the rung you are on, the rung above it, and a cadence for climbing.

The query set is not random. It is built around ten intent categories spanning awareness, consideration, evaluation, decision and authority, with geographic intent layered on where local presence matters. It is how we make sure we are measuring the questions a real buyer actually asks, not the ones a brand wishes they asked.

Most agencies treat AI visibility as a technical fix: schema markup, metadata, site structure. That is a small piece of the picture.

SignlLabs treats AI recommendation the way GCW has always treated reputation: as something earned rather than built. The heaviest lever in the framework after Foundation is Endorsement, meaning third-party validation from sources the engines already trust. That is PR work, not web development work, and it is why this practice belongs at GCW, backed by a measurement instrument, the Maturity Method™ and its Maturity Ladder™.

No. Traditional SEO ranks pages for a search engine that sends a person to a list of links. Answer engines do not rank pages. They compose a single answer from the sources they already trust and hand the person that answer directly, often with no link at all. The old goal was traffic to a page. The new goal is being one of the sources the engine relies on to build its answer.

AEO focuses on structuring a brand's own content (FAQs, schema, clean headers) so an engine can find and extract it. That is useful, and it addresses only part of the problem, because it stops at the edge of the brand's own website.

An answer engine does not simply answer the question it is asked. It fans a single prompt out into multiple hidden retrieval queries, then composes an answer from the sources it trusts most, which are overwhelmingly earned through third-party endorsement rather than a brand's own site. SignlLabs is built around that reality: we do not only optimise what a brand says about itself, we build the outside validation that determines whether it gets pulled into the composed answer at all.

No. GEO is the academic and industry term for optimising content to be surfaced inside generated answers. In practice it is used as a near-synonym for AEO, with more emphasis on phrasing, citations and statistics inside the content itself. If someone says GEO, they usually mean AEO.

The distinction in one line: AEO and GEO optimise what a brand says about itself. AI Signaling builds what other people say about it, and then proves the movement with a score.

AIO has two meanings and it is worth asking which one: “AI Optimization”, a loose catch-all covering all of the above, and Google's AI Overviews, the generated summary at the top of a results page. LLMO, large language model optimization, is another catch-all usually meaning the same thing as GEO, though some practitioners use it specifically for influencing what a model absorbs during training rather than what it retrieves live.

Most of the market uses these interchangeably. The confusion is worth clearing up on a call, because the acronyms all describe owned-media tactics and none of them describe the earned-authority work that moves the number.

A dashboard tells you the score. We tell you the move, and then we make it. Monitoring is one component of five. The rest of it (earned citation, entity definition, structured answers, wire distribution) is execution no software licence can do for you, and it is where the weighting says the score actually moves.

Technical work like schema markup makes a brand legible to answer engines. It does not make a brand trusted by them. These systems weight third-party validation (the same best-of lists, expert quotes, awards and trade coverage PR has always pursued) far more heavily than anything a brand says about itself. That is why SignlLabs is PR-led at its core, supported by structured data and technical work rather than the other way around.

An engine rarely searches the exact question it was asked. It breaks the prompt into multiple hidden retrieval queries, then composes one answer from what those return. A brand can be invisible to the asked question and still get pulled in through a fan-out query, and the reverse is more common. It is also why a brand cannot keyword its way into an AI answer.

The AI-era version of share of voice: out of the relevant questions in a category, how often is this brand the one recommended, and against whom. That is what the Signaling Maturity Score™ is actually measuring.

Usually entity confusion, which is the single most common reason a brand does not get recommended. An entity is how an engine understands a brand as a distinct, defined thing rather than a string of words. A brand with a clean entity has one name, one description and one set of facts everywhere it appears. When those disagree across sources, the engine cannot resolve who you are, and it names someone it can.

A confidently wrong statement is usually a symptom of weak Alignment: thin or contradictory sources, so the model fills the gap. The fix is authoritative, consistent third-party material, not a correction request. You cannot edit the answer; you can change what the engine finds when it goes looking.

It is the first thing to check. Answer engines use named crawlers, including OAI-SearchBot and ChatGPT-User, PerplexityBot and ClaudeBot. If robots.txt blocks them, none of the rest of the work can reach the engine.

Because zero-click is the default now: the person gets their answer and never visits a website. Fresh, well-sourced third-party coverage moves retrieval faster than a site rebuild does, and the outcome to measure is whether you were named and recommended, not whether someone clicked.

A monthly retainer, run on a repeating quarterly cadence. Weeks 1–2, re-score: run the query set across all five engines and produce the Signaling Maturity Score™ with competitive benchmarking. Week 3, plan: set targets and sequence the work against the weakest weighted dimensions. Weeks 4–10, build: earned media, schema, content, entity and message consistency. Weeks 11–13, report: movement by dimension and by engine, then next quarter's targets.

It is a monthly retainer, scoped to the brand rather than sold as a package, because the work is driven by the score: how much foundation repair the audit turns up, and how much earned placement the weakest weighted dimensions call for. Every engagement starts the same way, with the baseline audit, so the first conversation is about what your number actually is. Ask us and we will scope it against your category.

The first re-score comes at the end of the first quarter, because that is the honest cadence: the engines re-crawl and re-retrieve on their own schedule, and a placement made in week five may not be reflected until they do. Foundation work, meaning entity definition and schema, tends to register fastest. Endorsement and Presence depend on third parties publishing, so they take longer and last longer.

No, and we will not promise a ranking or a specific score by a specific date either. Nobody controls what ChatGPT, Gemini, Claude, Copilot or Perplexity say, and the engines change their retrieval and their models without notice.

What we commit to is the instrument and the work: a measured baseline you can audit, a plan sequenced against your weakest weighted dimensions, the execution to put your brand on the record in the sources those engines retrieve from, and an honest per-engine re-score every quarter showing what moved and what did not. The honest pitch is that we can show you exactly where you stand and exactly what moves it.

Yes, all the way down. Every observation is a 0, 1 or 2 on a named engine, from a query in a published query set, run three times with the median recorded, on a recorded date. Roll those up and you get the dimension breakdown; weight the dimensions and you get the score. There is no step in it you cannot ask us to show you.

Brands in categories where buyers research before they buy and a shortlist forms before anyone visits a website: consumer health and CPG, professional and legal services, energy and industrial, technology, and regulated categories where being described wrongly carries a cost. It works for Fortune 100 companies and early-stage startups alike.

Yes. The practice, including the Maturity Method™, is available for license by public relations firms, advertising agencies and web development companies for use with their own clients. They get a measurable AI Signaling program without building one from scratch: you run it under your own banner, we train the team and stand behind the scoring.

Beyond measurement, SignlLabs provides hands-on counsel on where to invest time and resources, how to compete, and where growth comes next: market sizing, opportunity mapping and a phased roadmap. It works for Fortune 100 companies and early-stage startups alike.

Yes. Geographic intent (“near me”, “in your city”) layers onto any category in the Intent Category Framework™, so a brand that sells in specific markets is scored on how the engines answer there rather than only in the general case.