Search Standing/Definitions
Reference

Search and AI visibility terms, defined plainly.

Sixteen terms that appear in competitive search analysis, each defined in a sentence or two. No jargon used to explain jargon.

Written for owners and operators rather than for practitioners.

Search Standing  ·  Competitive search and AI visibility analysis  ·  Reviewed annually
The terms[1]

What do these terms actually mean?

Each definition below is written to stand alone. Nothing here requires reading anything else on this page, and nothing is defined using a term that is itself undefined.

Branded traffic
Visits from people who searched the company name. Branded traffic is not customer acquisition: those visitors already knew the business existed. Excluding it from a traffic figure gives the number that actually matters, and for many established companies what remains is close to zero.
Keyword cannibalization
When several pages on one website compete for the same search term. Search engines cannot determine which page answers the query, so they rank all of them weakly rather than one strongly. It is one of the most common and most correctable causes of poor rankings on established sites.
Keyword difficulty
A score, usually 0 to 100, estimating how hard it is to rank on the first page for a term. It is derived largely from the strength of the pages currently ranking. Difficulty below 20 generally means the positions are held by weak or unoptimised pages rather than by strong competitors.
Entity resolution
The process by which a search engine or AI system determines which specific organisation a name refers to, where it operates and what it does. Businesses in cities with shared place names, or with generic company names, frequently fail to resolve cleanly, which prevents local search intent from reaching them.
AI Overview
The answer block Google generates above conventional search results for some queries. It names entities and cites sources. Appearing in it is binary: a company is either in the answer or absent, with no equivalent of position two.
AI visibility
The extent to which AI answer engines mention and cite a business when users ask questions in its category. Measured by mention count, citation count, which specific pages are cited, and the proportion of relevant prompts on which a company appears.
Generative engine optimisation
Abbreviated GEO. The practice of structuring content so it is likely to be quoted in AI-generated answers. In practice it overlaps almost entirely with conventional search optimisation, because AI systems predominantly cite pages that already rank well.
Answer engine optimisation
Abbreviated AEO. Used interchangeably with generative engine optimisation. Both describe writing pages that answer questions directly and are structured so a machine can extract a correct, self-contained passage.
Cited source
A page an AI system draws on when composing an answer. In many industrial and local categories, the most-cited sources for a company are third-party directories and review profiles rather than the company’s own website, which changes what corrective work is required.
Share of voice
The proportion of visibility a company holds across a defined set of search terms, relative to its competitors. It is more useful than raw traffic because it measures position within a market rather than an absolute figure with no context.
Schema markup
Structured data added to a page in a machine-readable format, usually JSON-LD, stating what the page is about. It allows search engines and AI systems to identify an organisation, a service, a location or a set of questions and answers without inferring them from prose.
Local pack
The block of business listings Google displays above organic results for queries with local intent. It is drawn from business profiles rather than websites, which means a company can rank organically and still be absent from it.
Referring domain
A distinct website linking to yours. Referring domain count is a better measure of authority than raw backlink count, because a hundred links from one site carry far less weight than one link each from a hundred sites.
Authority score
A third-party estimate of a domain’s overall strength, generally on a 0 to 100 scale, derived from its link profile and traffic. It is a useful comparator but a weak predictor of performance in low-competition categories, where published page count matters more.
Crawl access
Whether automated systems are permitted to read a site, controlled through the robots.txt file. Blocking AI crawlers such as GPTBot or ClaudeBot prevents a business appearing in AI answers regardless of how well its pages are written.
Commercial intent
A signal that a searcher is looking to buy rather than to learn. Cost per click is the most reliable available proxy: advertisers do not sustain high bids on traffic that does not convert, so a high cost per click on a low-difficulty term indicates value that nobody has claimed organically.
Why the vocabulary matters[2]

Why does the terminology matter to a business owner?

Because the words determine what gets measured, and what gets measured determines what gets fixed. A company told its traffic is growing may not realise the growth is entirely branded, which means search is acquiring nobody.

What a traffic figure hidesA total traffic figure combines branded visits from people who already knew the company with non-branded visits from new buyers. Only the second represents customer acquisition. WHAT A TRAFFIC FIGURE SHOWS WHAT IT HIDES Total monthly visitsWhether visits are brandedGrowth or declineWhether anyone new arrivedRanking keyword countWhich pages competePages receiving trafficWhether AI answers name youAverage positionShare of the available marketTwo companies with identical traffic figures can have entirely different acquisition performance.
The same traffic figure can describe a functioning channel or a directory lookup.

Three of these terms account for most of the confusion we encounter. Branded traffic makes a search channel look functional when it is not. Keyword difficulty is routinely ignored, so companies target terms they cannot win while unclaimed terms sit next to them. And entity resolution is almost never discussed at all, despite being the reason many local businesses fail to appear in their own city.

From our own measurement: in one industrial category analysed in July 2026, the term “pick and pack fulfillment services” carried a keyword difficulty score of 14 and a cost per click of $32.57. The 219-keyword cluster around it averaged a difficulty of 6. Difficulty that low on terms priced that high does not normally coexist: advertisers established the commercial value while the organic positions went unclaimed.

The direction of travel is documented outside our own work. Semrush tracked Google AI Overview appearance rising from roughly 6.5 percent of queries in January 2025 to a peak near 25 percent that July across a ten-million-keyword sample. BrightEdge, measuring nine commercial industries, recorded 48 percent of queries triggering an AI Overview by February 2026. Seer Interactive measured organic click-through on those queries falling from 1.76 percent to 0.61 percent. Prevalence estimates vary by methodology and should be read as a range rather than a figure.

Questions[3]

Questions about the terminology

What is the difference between GEO and AEO?
Generative engine optimisation and answer engine optimisation are used interchangeably. Both describe structuring content so it can be quoted in AI-generated answers. In practice the work overlaps almost entirely with conventional search optimisation, because AI systems predominantly cite pages that already rank well organically.
Is AI visibility the same as SEO?
No, though they are closely related. SEO concerns where a page ranks in a list of results. AI visibility concerns whether a business is named and cited within a generated answer. The distinction matters because there is no position two in an answer: a company is either included or absent.
Why is branded traffic excluded from analysis?
Because it measures recognition rather than acquisition. Someone searching a company by name already knows it exists and would have found it regardless. Excluding branded queries produces the figure that actually indicates whether search is bringing in new buyers, and for many established businesses that figure is close to zero.
What is a good keyword difficulty score to target?
Below 20 for a site with limited domain authority, and below 30 for an established one. Difficulty largely reflects the strength of the pages currently ranking, so a low score usually means the positions are held by weak or unoptimised pages rather than by strong competitors.
Does schema markup improve rankings?
Not directly. Schema markup does not itself raise a page’s position. What it does is allow search engines and AI systems to identify what a page describes without inferring it from prose, which makes a page eligible for richer result formats and easier for an AI system to cite accurately.
Start[5]

Want these applied to your own business?

Send your company and your market. We will run the analysis and send back what these terms actually mean in your case, measured against your real competitors.

No call required and no obligation. Read it, check it yourself, and decide. If the findings are worth discussing after that, we can talk then.

hello@searchstanding.com