Every emerging marketing channel goes through three phases: dismissed as a fad, hyped as a panic, then quietly absorbed into the standard playbook. AI search is somewhere between phase one and two right now, depending on which agency's email you read most recently.
This piece tries to do the calibration honestly. We're going to make a case for why AI search matters for some businesses now — but we're also going to tell you which businesses can rationally ignore it for another year.
The case for caring now
Some things that are objectively true in May 2026:
- ChatGPT processes over 2.5 billion prompts daily. A meaningful percentage are commercial — people researching products, services, and providers.
- Google's AI Overviews appear above the traditional results for most informational queries. The clicks that used to flow to top organic results increasingly flow to the cited sources inside the AI block — which are often different sites.
- Perplexity, Claude, Microsoft Copilot, and Gemini are all growing in usage, especially among the demographics most likely to research before buying.
- Most small businesses have done zero AI optimization. Their schema is incomplete, their content isn't structured for citation, and many haven't even allowed AI crawlers in their robots.txt.
When a category is in early adoption, the businesses that move first capture disproportionate brand recall. When AI engines train on the current web, the sites that are well-structured and cited often become the canonical sources — a positioning effect that can persist for years.
The case for being patient
Also true:
- For most small local businesses, traditional Google still drives more pipeline than AI search does. The buyer who needs an emergency plumber tonight is more likely to type "plumber near me" into Google Maps than to consult ChatGPT.
- AI engines are still evolving fast. The exact signals they reward in May 2026 may not be the same signals they reward in May 2027. Over-investing in current tactics could mean optimizing for patterns that fade.
- Measurement is hard. Tracking citation share across multiple engines is non-trivial, and the volume in any single AI engine is rarely sufficient to drive clear monthly decisions for small businesses.
- Many "AI SEO" services in the market are repackaged generic SEO with a new label and a price markup. Buyers can rationally hesitate while the category sorts out.
Who should care now
A few business profiles where AI search is already a real channel and worth investing in:
B2B SaaS and software — buyers research extensively before purchase, often via AI assistants. Being cited inside that research wins deals at the awareness stage.
Professional services with long research cycles — lawyers, accountants, consultants, financial advisors. Clients spend weeks or months evaluating options. AI assistants are common in that evaluation.
Information-heavy businesses — comparison sites, niche media, B2B publications. Your traffic comes from being a trusted information source. AI engines compete for the same role.
Anyone in a "competitive long-tail" niche — where prospects type specific questions and discovery is search-driven. AI search captures these queries efficiently.
Anyone with above-average dependency on inbound traffic — if your business model depends on people finding you organically (vs. paid ads, foot traffic, or word-of-mouth referrals), AI search visibility is becoming non-optional.
Who can wait
Some businesses can rationally focus on other channels for another 6–12 months:
Hyper-local single-location businesses with steady foot traffic and a strong Google Business Profile. Pizza places, hair salons, neighborhood services. Google Maps and review platforms drive most relevant traffic; AI search adoption in these categories is real but slow.
Businesses with extreme word-of-mouth or referral concentration. If 80% of your customers come from referrals or existing customer expansion, paid and AI channels are secondary.
Businesses serving non-internet-native customer bases. Some demographics still research through traditional Google, calls, and in-person. Watch the trend, but you have time.
Businesses in extreme regulation where content distribution is restricted regardless of channel.
The pragmatic middle path
Most small businesses sit somewhere between "must do now" and "can wait." For that majority, the rational approach is:
Spend 1–2 days getting the foundations right. Update robots.txt to allow AI crawlers. Deploy basic schema (Organization, WebSite, FAQPage). Audit your top three pages for question-structure and clarity. None of this takes long, none of it costs much, and all of it benefits traditional SEO too.
Then watch the data. Run an AI visibility check monthly for your top 10 target queries. If you start showing up in citations, double down — the channel is working for you. If you're not showing up after 6 months despite doing the basics, you're either in a category that hasn't matured for AI search yet, or you have a positioning problem the platform can't solve.
The best news in all this: the foundational work for AI search is mostly the same work that already benefits traditional SEO. Even if AI search underperforms your expectations, you haven't wasted anything.
What's the actual downside of waiting?
If you wait 12 more months to start GEO work, here's what you're risking:
- Your category's AI-cited sources become entrenched. AI engines tend to keep citing what they've been citing. New entrants face a brand-recall barrier.
- Competitors who started early have built compound advantages — citation networks, content libraries, schema infrastructure — that take months to catch up to.
- You miss the data you'd have had. Citation share tracking gets more useful over time because you can see what works for you specifically.
None of this is fatal. People who wait can still win. But the entry tax goes up over time.
An honest pitch
We run Reffed. We'd obviously rather you sign up than wait. But the honest version of the pitch isn't "you must do this immediately or die." It's: "The cost of getting the basics right is low, the benefit is real and compounding, and the businesses that wait are paying a slowly rising entry tax."
If that framing resonates and you'd rather not handle the work yourself, run a free audit to see what we'd recommend for your site. If it doesn't, the DIY path is also valid — and the playbook is here.
Find out if AI search matters for your site
Reffed's free audit will tell you specifically — based on your current state and what AI engines see today.