[]
en

Language

Book a call
Blog
Measurement5 min read · By the New Map team · 21 October 2026

Is GEO worth it for a small B2B company?

Domains cited when engines were asked if GEO is worth it (7 answers)
Domains cited when engines were asked if GEO is worth it (7 answers)Bar chart, Domains cited when engines were asked if GEO is worth it (7 answers): the highest is fullmixmarketing.co.uk at 3; the 8 values range from 2 to 3. Two domains were cited in 3 of 7 answers; six others in 2 of 7.Two domains were cited in 3 of 7 answers;six others in 2 of 7.fullmixmark…3reddit.com3developers.…2epos-market…2forbes.com2hubspot.com2propellex.co2thesalesena…2
Source: New Map measurement, 19 October 2026 · How we measure

GEO is worth it for a small B2B company if it starts with retrieval: Chen et al.1 (preprint, October 2026) linked 32% of within-site citation differences to retrieval rank, against 2.8% for all page features. No data yet says if it pays back; first run the test below on your buyer questions.

What do AI engines say when asked if GEO is worth it?

New Map's measurement2 counted 68 discovery questions; Perplexity gave 68 answers, Claude Code 10 answers, Gemini 68 answers (146 total). Reddit appeared in 20 of 146 answers, YouTube in 16 of 146. That is one week's reading of 68 questions about AI visibility tools on 3 engines (146 answers), not a trend: Perplexity cited a source in only 7 of its 68 answers, and Claude Code's 10 answers are too few to read alone. How New Map measures sets out the method.

On this exact question the sample is smaller. New Map's measurement of the weeks of 12 October 2026 and 19 October 2026 counted 7 answers to "Is GEO worth it for a small B2B company?": 2 on Perplexity, 3 on Gemini and 2 on ChatGPT. The engines named none of the tracked companies, and New Map was named in 0 of the 7 answers.

fullmixmarketing.co.uk and reddit.com were each cited in 3 of the 7 answers. developers.google.com, forbes.com and hubspot.com were cited in 2 of 7 each. Seven answers on 3 engines is too few to rank sources. It shows which pages to study, not a ranking. The chart shows how the cited domains spread across those answers.

Domains cited when engines were asked if GEO is worth it (7 answers)
Domains cited when engines were asked if GEO is worth it (7 answers)Bar chart, Domains cited when engines were asked if GEO is worth it (7 answers): the highest is fullmixmarketing.co.uk at 3; the 8 values range from 2 to 3. Two domains were cited in 3 of 7 answers; six others in 2 of 7.Two domains were cited in 3 of 7 answers;six others in 2 of 7.fullmixmark…3reddit.com3developers.…2epos-market…2forbes.com2hubspot.com2propellex.co2thesalesena…2
Source: New Map measurement, 19 October 2026 · How we measure

Where should a small company put its GEO effort first?

Start with being retrieved. Chen et al. (preprint, October 2026, 158,099 answers) found that ChatGPT had retrieved every page it cited, through its own searches on sub-queries the model rewrote for each answer. Pages were cited more often when their title and paragraphs matched the question, and that was the page trait that predicted citation.

The preprint found no effect inside a site from page speed, summary boxes, structured data or FAQ blocks. AirOps3' playbook (publication date not stated) reports citation rates of 2 to 2.8 times higher for pages with schema, FAQs and clean heading hierarchies than for unstructured pages. Ahrefs4 tested schema directly in May 2026 on 1,885 pages and found no gain against matched controls. The two measured tests point the same way, while the playbook gives no method here, so treat its figure as a hypothesis to test.

Our read, not a measured result: a page whose title and opening paragraph state the buyer's question is a small edit. On 14 October 2026, DEJAN AI5 reported that Gemini 4 Argon returned pages with transparent pricing in 99.7% of tests. That result is specific to that version. The chart sets two figures from the preprint side by side.

What drives GEO results: retrieval rank against page features
What drives GEO results: retrieval rank against page featuresBar chart, What drives GEO results: retrieval rank against page features: the highest is Retrieval rank at 32%; the 2 values range from 2.8% to 32%. Retrieval rank explained 32% of within-site citation variation; all page features 2.8%.Retrieval rank explained 32% ofwithin-site citation variation; all pagefeatures 2.8%.Retrievalrank32%All pagefeatures…2.8%
Source: Chen et al. (arXiv preprint), 10 October 2026

Which sources should a small B2B company's GEO work reach?

Gartner's 2024 survey of 632 B2B buyers found 61% preferred to buy without a sales rep, as Seer Interactive6 reported in October 2026. Separately, the engine studies below look at which sources get cited.

Ahrefs found in October 2026 that Wikipedia made up 71.9% of external citations in branded AI Overviews. Only one competitor-owned domain appeared across the 57 source lists it checked.

Muck Rack's May 2026 study of more than 25M cited links found earned sources made up about 84% of citations on unbranded questions. Muck Rack notes that "earned" there includes Wikipedia and social sources, so it does not mean PR drives 84%.

A citation is also not a recommendation. Ahrefs' July 2026 study of 9,886 answers found 43% of answers citing a brand's own event page never named the brand. A small company should aim for mentions it can authentically earn, in sources the engines already cite.

How stable are GEO results, and does that change what it is worth?

In consecutive runs of an identical prompt, 20% of brands stayed visible; visibility held at 30% from one run to the next, according to AirOps' playbook (publication date not stated). Its method is not given, so read it as a direction.

Profound's September 2026 study of 1.19M page-engine lifecycles found the median page loses half its citation share within 11 days of its peak. Pages cited by five or more engines held their citations 5.5x as often. Peec AI's September 2026 study of 120M chats found weekly moves under about 5 points can be chance.

For a small company, judge GEO on repeated readings over weeks, not on one check. The chart shows the AirOps figures.

How can a small B2B company test whether GEO is worth it?

GEO is worth it for a small team when the test is cheap and the readings repeat. How a brand's visibility in AI answers is measured explains the counts used below.

  1. Write the questions your buyers would ask an engine before they know you, and the ones they ask when they check you by name.
  2. Ask each question on several engines, more than once, and record the brands named and the domains cited.
  3. Mark the cited domains where your company could authentically earn a mention.
  4. Rewrite the pages that should answer each question so the title and first paragraph match it.
  5. Read the same questions again after several weeks, and treat small weekly moves with caution.

The diagram shows the loop.

The test for whether GEO is worth it for a small company
The test for whether GEO is worth it for a small companyDiagram, The test for whether GEO is worth it for a small company: Pick buyer questions, then Ask engines repeatedly, then Read brands and sources, then Fix pages, earn mentions, then Read again after weeks. The test is a loop: ask, read, fix, then read again.The test is a loop: ask, read, fix, thenread again.repeatPick buyer questionsAsk engines repeatedlyRead brands and sourcesFix pages, earn mentionsRead again after weeks
Source: New Map, 20 October 2026

What to do

  1. Write ten buyer questions, including this one, and ask each on at least three engines more than once. Log the brands named and the domains cited for every answer.
  2. Rewrite the title and opening paragraph of the page that should answer each question, so they match the question in the buyer's words.
  3. If you want the readings and fixes handled, New Map measures what engines answer about your brand and drafts the fixes; see plans and pricing.

Frequently asked

Will AI engines name my company when asked about GEO?

New Map's measurement counted 7 answers to this question across 3 engines. The engines named none of the tracked companies, and New Map was named in 0 of the 7.

Do FAQ blocks and schema help a small company get cited?

A Chen et al. preprint (October 2026) found FAQ blocks and structured data had no within-site effect; Ahrefs (May 2026) found 1,885 pages adding schema gained nothing against matched controls. AirOps reports 2 to 2.8 times more citations, undated. Test on your pages.

Can I judge GEO from a single check of one engine?

No. According to AirOps' playbook (undated), 20% of results remained visible across five consecutive runs of the same prompt and 30% stayed visible from one run to the next. Repeat the same questions over several weeks before deciding.

Sources

  1. Chen et al. (arXiv), What Drives AI Citations? Content, Retrieval and Model Knowledge in Generative Engine Optimization (GEO), arxiv.org, 10 October 2026
  2. New Map measurement, How we measure, 19 October 2026
  3. AirOps, The GEO Strategy Playbook for Marketers and Organic Growth Teams, airops.com
  4. Ahrefs, AI Overviews Currently Appear for 83% of Branded Searches, ahrefs.com, 2 October 2026
  5. DEJAN AI (Dan Petrovic), Gemini 4 Argon's personality: no hype, hard facts, fewer tables, dejan.ai, 14 October 2026
  6. Seer Interactive, Competitive Analysis for Paid Media: How To Find the Broken Stage, seerinteractive.com, 7 October 2026
  7. Zhang et al. (arXiv), VoteGEO-Bench: Measuring the Impact of Generative Engine Optimization on Electoral Visibility, arxiv.org, 15 October 2026

See where you stand

Book a call and find out how visible your brand is across the AI engines your customers already ask.

Book a call

For agencies: See the agency package