A complete AI visibility report, before you buy one

When an assistant is asked what customers think of you, it does not read minds — it synthesises whatever public evidence it can find: your own site copy, your reviews, directory listings, professional bodies, and the way other people write about you. In your case that evidence is consistent but thin. The picture that comes back is of a small, experienced consultancy whose clients value patience, plain English and the fact that certification actually gets achieved on the first attempt. Words that recur across answers are 'thorough', 'straightforward', 'good value' and 'responsive' — reassuring words, but generic ones. They are the vocabulary of a competent supplier, not of a specialist people seek out by name.

Sentiment is clearly positive: nothing in the observed answers is critical of your work, and no assistant raised a concern about quality, reliability or conduct. That matters, because negative sentiment is far harder to correct than absent sentiment. What is missing is depth. Assistants can tell a prospective client that you are well regarded; they cannot tell them what it is like to work with you, how long an engagement typically takes, what happens when an audit finds a non-conformity, or which kinds of organisation get the most from you. Where a buyer asks a comparison question, that missing texture is the reason another firm's answer reads richer than yours.

The strongest customer signal in your favour is longevity — repeated references to a long-established firm with accredited consultants. The weakest is specificity. There are no named client outcomes attached to you in the public record, no sector stories, and no quantified results. Reviews exist but are few and cluster in one place, so an assistant treats them as suggestive rather than conclusive. In practice this makes your reputation legible but not persuasive: it supports you once you are already on a shortlist, and does very little to get you onto one.

There is also a quiet mismatch between what your clients appear to value and what your public material emphasises. Client-side language centres on the experience of being guided through a stressful process; your own material centres on standards, scope and process. An assistant answering 'who is good to work with?' has to reach for the first, and finds little of it in your own words — so it borrows the vocabulary of whoever did write it down.

AI assistants recommend the businesses they can describe with confidence. Every suggestion below comes from that single mechanism: where an answer had to be hedged or omitted, there was nothing authoritative to draw on. The first and most repeated suggestion is service-level depth. You are strongly associated with ISO 9001 and reasonably with ISO 14001, but information security and AI governance barely register — despite being work you do. A single sentence on a combined services page is not enough; assistants want a dedicated page per standard, written in the language a buyer would use, with scope, process, timescales and worked examples.

The second theme is proof. Assistants repeatedly note the absence of named outcomes. Case studies do disproportionate work here because they contain the three things a recommendation needs at once: a sector, a problem and a result. Three case studies covering different sectors would let an assistant answer 'who has done this for a company like mine?' — currently one of the questions where you are least likely to be named.

Third is geography. Your coverage is described in regional terms, which is safe but weak. Buyers ask in town names. Where a competitor has a page for a specific town and you do not, the competitor is named and you are not — even when you serve that town and they barely do. Location pages only work when they carry genuinely local substance, so pair each one with local client evidence rather than a template.

Fourth is transparency about cost and engagement. Almost no answer could indicate what working with you involves or what it costs, so assistants steer cautious buyers towards firms that publish indicative figures or a clear engagement model. You do not need a price list; a typical range, a description of what shapes the cost, and a plain outline of the stages of an engagement would remove the ambiguity.

Fifth, and easiest to overlook: machine-readable basics. Consistent business details across your site and the major directories, structured data describing your services and areas served, and a clear about page that states plainly what you do and for whom. These are unglamorous, cheap and fast, and they raise the confidence with which every other answer is given.

Differentiation in AI answers is not decided by who is best. It is decided by who is easiest to characterise. When an assistant compares you to the firms it also knows, it reaches for the sharpest distinguishing feature it can support with evidence. For you that feature is dependable, hands-on regional delivery for organisations achieving certification for the first time. That is a real and commercially valuable position — but it is also a narrow one, and it quietly excludes you from every question about scale, complexity or specialism.

The competitor most often named alongside you occupies the opposite space. It is described in national terms, is connected to a wider spread of standards, and publishes sector-specific material that assistants can cite. In head-to-head comparison answers it therefore reads as the more capable option, even where your delivery quality is equal or better. The gap here is documentary, not operational: it has written down the things you have done but not published.

A second competitor differentiates on sector rather than geography. By concentrating its published material on a small number of industries, it is named almost automatically for questions carrying that context — a much cheaper form of dominance than trying to be visible everywhere, and one that is available to you in the sectors where your experience is deepest.

Your genuine advantages — direct access to senior consultants, a first-time pass record, continuity of the same consultant across an engagement, and a regional presence that means someone can actually attend site — are only weakly represented in AI answers, because they are stated as claims rather than demonstrated with evidence. Assistants discount unsupported claims heavily; every competitor says it is experienced and responsive. What they do not all have is a named client saying it, a date, and a number.

The strategic conclusion is narrow and encouraging: you are not losing comparison answers on capability. You are losing them on the availability of evidence. Turning three or four of your strongest engagements into published, specific, sector-labelled stories would change how you are characterised in exactly the answers where higher-value buyers are making shortlists.