What AI search optimization is, without the hype
AI search systems answer the question directly and cite a handful of sources. Being one of those sources, or being named when someone asks “which SEO consultant in Kathmandu” or “best trekking company for Everest Base Camp”, is the new visibility problem. The underlying requirements are familiar: pages that answer a specific question clearly, facts that are consistent across the web, structured data, a crawlable site and a brand that other sites mention. What is new is the format of the answer, the way sources are selected and the difficulty of measuring any of it. AI search optimization is the part of my SEO services that applies to this format. It is not a replacement for SEO; the pages that get cited are overwhelmingly the pages that already rank.
Who this is for
- Businesses watching informational traffic fall as AI Overviews answer the question on the results page.
- Brands that want to know whether ChatGPT and Perplexity mention them, and how, when buyers ask for recommendations.
- Companies in Australia, the UK and the USA, where AI Overviews and AI Mode are most widely rolled out, whose category queries now show an AI answer.
- Nepali businesses preparing for the same, as coverage for Nepal and Nepali-language queries expands.
The four services in this sub-silo
What is measurable today, and what is not
| Measurable now | Partly measurable | Not measurable yet |
|---|---|---|
| Referral sessions and conversions from chatgpt.com, perplexity.ai, gemini.google.com and similar sources in GA4 | Whether an AI Overview appears for a tracked query set, and which sources it cites, sampled by tools or by hand from a given location | Impressions and clicks from AI Overviews as a separate line: Search Console includes them in normal totals without a breakdown |
| Rankings, clicks and impressions for the underlying pages, as before | Whether an assistant names your brand for a set of recorded prompts, sampled on a schedule; answers vary by session, location and model version | How many people saw your brand inside an AI answer and did not click |
| Crawl activity from AI crawlers (GPTBot, PerplexityBot, Google-Extended and others) in server logs | Share of voice against competitors within sampled answers, as a trend rather than a precise figure | A reliable “AI ranking” position; the systems do not rank in a stable list |
| Structured data validity and entity consistency across your profiles | Correlation between content changes and later citations, with weeks of lag and noise | Attribution of a sale to an AI answer the buyer read a week earlier |
I report the first column as fact, the second as sampled evidence with the method stated, and I do not report the third at all. Any vendor showing you a precise “AI visibility score” is presenting a sample as a measurement.
How an AI search engagement runs
- Baseline audit. A prompt set built from your keyword map and buyer questions; sampled across AI Overviews and the main assistants; AI referrals and crawler activity pulled from GA4 and logs. Delivered through the AI visibility audit.
- Foundation check. Crawlability for AI crawlers, structured data, entity consistency (name, address, descriptions, founders, products) across site, profiles and directories; fixed with the technical SEO and schema markup services.
- Content work. Answer-first restructuring of the pages that should be cited, new pages for the questions in the prompt set, and original data or explanations that give an assistant a reason to cite you rather than a competitor, planned with content SEO and strategy.
- Off-site presence. Mentions and listings in the sources the systems draw on, through the same outreach as link building.
- Re-sampling and reporting. The prompt set re-run monthly, changes noted with the method, and referrals and rankings reported alongside.
My position on the current claims
Three things are supported by evidence: pages that rank well are far more likely to be cited; clear, direct answers with consistent facts get quoted more often; and off-site mentions influence what assistants say about a brand. Three things are not: that any file or tag guarantees inclusion, that content written “for AI” in a particular style is favoured, or that a stable ranking inside an assistant can be bought or engineered. I will apply the first three and decline to sell the second three. The plain-language explainer for readers who want the background is the AI search optimization guide.
Pricing
AI visibility tracking and answer-format work are included at the Growth and Enterprise tiers on the SEO packages and pricing page; the audit is available standalone.
Next step
Request a free AI visibility check and I will sample ten prompts for your category and report where you appear, where competitors appear and what the underlying pages need. All related services are under SEO services.