Short answer: a WordStream by LocaliQ survey of over 300 US small businesses found that SMBs now name social media (64%) as a main source of website traffic more often than organic search (52%), that 40% report at least some traffic loss from Google algorithm changes and the shift to AI search, and that 48% say they are extremely or very familiar with GEO. The GEO work they actually do is ordinary page work: clear headings (35%), readability (26%) and technical fixes (24%). This article summarises that third-party survey on SEO, GEO and traffic. It is not InsightWonder research.
The full report is on WordStream's site, and LocaliQ publishes a ten-point summary of the same survey. We checked every figure below against those two pages. Where the report makes a claim it attributes to another study, we either traced it to the original or left it out.
What the SMB survey measured, and what it cannot tell you
WordStream surveyed over 300 small businesses in 24 sub-industries across the US and deliberately included businesses that have no website. It split them into three size groups: solopreneurs, businesses with two to 10 employees, and businesses with 11 to 100 employees. 22% of respondents are in home services, so the sample leans towards local, service-led businesses rather than pure online retail.
Keep three limits in mind before reading any percentage:
- It is self-reported. "We lost traffic" and "we track AI referrals" are what owners said, not what their analytics showed.
- It is American. Which social platforms, directories and marketplaces matter will differ in other markets.
- The size groups do not see equally well. A business with someone watching analytics will notice a traffic drop that a sole trader without goal tracking will miss. The report raises this itself.
A note on this revision: we rebuilt the page to make clear from the first paragraph that it summarises someone else's survey. We re-checked every figure at its source, replaced a statistic the report attributes to another company's study with an Ahrefs figure we could verify directly, and dropped a description of the publisher we could not confirm. One figure is stated two different ways within the report itself (how many businesses without a site plan to launch one); we use the version both of the publisher's pages agree on.
Where SMB traffic comes from, and who is losing it
Asked to tick every main source of website traffic, respondents named social media, organic search and referrals most often. It was not only the larger firms: 57% of solopreneurs and 62% of businesses with two to 10 employees named social. AI was chosen least often, although 18% of businesses, and 11% of one-person shops, count AI sources among their top traffic drivers.
This does not mean search stopped mattering. 72% of respondents still describe their SEO as very or somewhat effective, rising to 89% among businesses with 11 to 100 employees. What the answers show is that discovery is spread across more places. A customer may see a post, ask an assistant for a recommendation and then look the business up before getting in touch, and each of those steps needs something accurate to land on.
Traffic loss is real, and larger firms report more of it. 40% of businesses reported at least some traffic loss attributed to Google algorithm changes and the shift to AI-powered search, and 20% said their traffic was down considerably. Among businesses with 11 to 100 employees the figure was 46%.
The report offers two explanations for the size gap and presents both as theories. Larger businesses track their traffic more closely, so they see declines that smaller ones miss. And smaller businesses are more likely to be local, and the report says queries that include a location name are less likely to trigger an AI Overview. For a sense of what an AI Overview does to clicks, Ahrefs re-ran its study on December 2025 data and found that the presence of an AI Overview correlated with a 58% lower average clickthrough rate for the top-ranking page (Ahrefs). That is a correlation across a large keyword set, not a measurement of any particular small business.
SMB familiarity with GEO splits along company size
48% of respondents say they are extremely or very familiar with generative engine optimization, and only 5% are not familiar with it at all. The average hides a wide gap by size, and monitoring follows familiarity almost exactly:
| Group | Extremely or very familiar with GEO | Monitor AI referrals and mentions |
|---|---|---|
| All respondents | 48% | Half |
| 11 to 100 employees | 65% | 70% |
| Two to 10 employees | 39% | 41% |
| Solopreneurs | 34% | 32% |
39% of solopreneurs said they were not too familiar or not familiar at all with GEO. Put plainly, the smallest businesses are the least likely to know whether assistants mention them. The same split shows up in how SEO is run: half of SMBs manage it in-house, 57% of solopreneurs do, and 19% of one-person brands use an agency against 53% of the largest businesses. The businesses with the least outside help are also the ones least likely to be watching AI answers.
The report argues that GEO can level the field because many AI citations come from pages outside Google's top ten, and it cites another company's study for that. The comparable figure we could verify at its source is Ahrefs' analysis of 863K keyword results pages: 37.9% of URLs cited in AI Overviews also appeared within the first 10 result blocks, 31.2% came from positions 11 to 100 and 31.0% from beyond the top 100 (Ahrefs). That covers AI Overviews only, and it does not mean a weak page gets cited. It does mean that ranking position and being cited are no longer the same contest.
What SMBs do for GEO, and where AI referrals land
Asked what they changed to improve visibility in AI search, respondents described ordinary page work. Structuring pages with clear, descriptive headings led at 35%. Optimizing readability, with short paragraphs and lists, followed at 26%, and technical SEO such as site speed and mobile compatibility came third at 24%. The largest businesses ranked encouraging external brand mentions (33%) and structured data (30%) as their second and third priorities.
The report notes that just about every tactic on the list is also SEO best practice. Our reading of the size difference is that newer teams start by fixing their own pages, while more experienced teams also work on what other sites say about them. Both kinds of work are covered in our guides to writing passages an engine can lift and third-party corroboration.
Businesses that monitor AI referrals were also asked which pages those visits land on. The homepage came first, followed by product or service pages and contact pages, while FAQ pages and news or press pages were the least likely. 45% of businesses with two to 10 employees also named their About page, against 33% of sole proprietors and 22% of the largest businesses.
That is a useful corrective to the idea that GEO means publishing a pile of new FAQ content. The pages assistants send people to are mostly pages a business already has. If the homepage does not say plainly what the business does, for whom and where, that is the first fix.
SMB SEO problems are concentrated; GEO problems are not
| Rank | Biggest SEO challenge | Biggest GEO challenge |
|---|---|---|
| 1 | Generating traffic (42%) | Competitors' content appears in AI summaries instead (24%) |
| 2 | Creating high-quality, SEO-optimized content (32%) | Tracking and attributing AI referral traffic (22%) |
| 3 | Keeping up with algorithm updates (28%) | Declining organic traffic (22%, tied) |
| 4 | Tracking ROI and attributing SEO value (20%) | Spread across the rest |
The report points out that no GEO challenge received more than 30% of the vote and the top four sat within 10 percentage points, while in SEO the top challenge received nearly 20% more votes than the fourth. A flat list like that usually means people do not yet agree on what the hard part is. The largest businesses put traffic and attribution difficulties first (28%) and competition second (27%). Attribution sitting near the top is the tell: it is hard to prioritise a channel you cannot see.
If you start measuring, keep your brand name out of the questions. Ask an assistant whether a named business is any good and it will name it. In one of our own runs, 13 of 15 generated questions contained the brand name and the score read 13% when the honest value was zero. Building a prompt set explains how to write questions that avoid this.
What the SMB survey means for your next steps
If you have no website
Roughly a third of respondents have no website, including over half of sole proprietors and 25% of businesses with multiple employees. Among them, 43% never plan to launch one, rising to 56% of sole proprietorships. Their reasons are practical: 50% rely on word of mouth or local referrals, 35% use social media, marketplaces and directory listings instead, 19% are unsure of the payback, 17% are focusing on other marketing, and 14% and 7% cite time and money.
AI is not changing many minds. 52% say the growth of AI will have no impact on the decision, 14% say it makes them less likely to build a site and 29% say more likely. Businesses that do have a site value it: 94% call it at least somewhat important to their goals, 70% sell directly from it, 69% treat it as a major source of leads, and 61% say website traffic has become more important over the past couple of years.
For a business without a site, the profiles it does control, such as its Google Business Profile, directory listings and social pages, are what an assistant can find, so keeping those accurate and consistent matters more, not less. Our guide to local GEO covers that situation.
If you have one: five steps, in order
- Rewrite the homepage first. It is the most-named AI landing page in the survey. Say what you do, for whom and where in the opening text, in plain words.
- Give each service or product line its own page. Product and service pages were the second most-named landing pages. One combined services page cannot answer several different questions well.
- Fix headings before anything clever. It is the most common GEO action in the survey, it costs time rather than money, and it helps ordinary search as well.
- Work on what others say about you. The largest businesses put external brand mentions second. Check that directories, local press and supplier pages state the same facts your own site does.
- Measure without your brand name in the question. Use a fixed set of buyer questions, ask several assistants, and treat a change as real only when it repeats in the next run.
The order follows the survey's own evidence rather than a theory: start with the pages visitors already land on, then fix structure, then work outside your site, and only then judge whether anything moved.
Sources
- WordStream by LocaliQ, The Big SMB Website Trends Report, last updated 2026-03-24.
- LocaliQ, 2026 SEO insights, last updated 2026-08-17.
- Ahrefs, AI Overviews and clicks, updated study.
- Ahrefs, AI Overview citations and the top 10 results.