How to Show Up in AI Search: ChatGPT, Perplexity, and Google's AI

Type a question into ChatGPT, Perplexity, or Google today and you often won't get a page of ten blue links. You get one written answer, with a few sources named under it. If your business is one of those named sources, you get found. If it isn't, you're invisible, even when you're the better business. So the practical question is simple: when someone asks an AI about a business like yours, can it read you clearly enough to cite you?

What is AI search, and how is it different from ten blue links?

Old search handed you a list and let you choose. You typed a query, Google returned links, and you clicked one. AI search does the choosing for you. Tools like ChatGPT search, Perplexity, Google's AI Overviews, and Gemini read across many pages, write one answer, and cite a handful of sources they leaned on. The reader often never scrolls a list. They read the answer and maybe click one of the two or three sources named.

This is already how a lot of people search. Google's AI Overviews, the AI summary at the top of many results, now reach 2 billion monthly users and appear in more than 200 countries and territories (TechCrunch, Google). In the US, a majority of adults (60%) say they now read those AI summaries, and about 42% use chatbots to search for information (Pew Research Center). More than 800 million people use ChatGPT every week (TechCrunch). Answer engines like Perplexity work the same way: they synthesize a reply and cite their sources inline.

The shift matters because the math changed. Ten links meant ten chances to be seen. One synthesized answer that names three sources means three slots. Getting into that short list is the whole game, and it's a different job from ranking a page on Google.

How does AI search actually decide which sources to cite?

Not by picking the best business. By picking the one it can read clearly and verify in more than one place. An AI answer is built from patterns across many sources. When several independent pages describe you the same way, same name, same service, same city, the model gets confident enough to use you. When information about you is thin, contradictory, or locked inside an image or a PDF it can't read, it stays vague and cites someone else. In practice, a few signals decide whether you make the short list:

• Legibility — can a machine read, in plain text, what you do, who it's for, and where? Facts buried in a logo, a hero image, or a scanned menu don't count.
• Corroboration — do independent sources (directories, reviews, press, other sites) describe you consistently? One mention is a claim; several matching ones read as a fact the model will repeat.
• Relevance to the exact question — does a page of yours already answer the specific thing the person asked, in the words they used?
• Freshness — many of these tools pull live web results, so recently updated, accurate pages get picked over stale ones.

There's a reason clear beats clever here. The signals that make a person trust a source, plain writing, real specifics, consistent facts, are the same ones that make a model quote it. If an assistant keeps skipping you, the cause is usually mechanical, not mysterious. It's worth understanding why AI can't find your business before you spend anything trying to fix it.

How do I show up in AI search? The concrete moves

There's no button and no ad slot. Showing up in AI search is a handful of habits that together make you easy to read and easy to verify. Do them roughly in this order.

Publish clear content that answers real questions

This is the foundation. Write pages that answer the exact questions your customers ask an assistant: whether you're open Sundays, what a service costs near them, whether you handle same-day work, how you're different from the obvious alternative. Use plain sentences and put the answer near the top. When your page already answers the question a person typed, you're the obvious thing to quote. A steady habit of publishing those answers is one reason blogging still helps your AI visibility.

Get genuinely referenced across the web

AI trusts what it sees corroborated. So the goal is honest presence, not link tricks: a listing in the directories your industry actually uses, a mention in local press or a community roundup, a profile on the review site your customers check, a guest post or interview on a relevant blog. Each real reference is a small vote that you exist and do what you say. You can't fake this at scale, and models are getting better at spotting people who try. This is also how you get recommended by ChatGPT instead of skipped.

Keep your site clean and structured for machines

Read your homepage as a stranger, or as a robot. Can it tell, in plain text, what you sell, who it's for, and the area you serve? Put those facts in words, not only in graphics. Add an about page with real detail. Where it fits, add structured data (schema) so your name, address, hours, and services are labeled in a format machines parse without guessing. Clean, fast, well-organized pages are easier to read, and easier to read means easier to cite.

Collect real reviews and keep your business details identical everywhere

Reviews are corroboration at scale: many independent people describing what you do and how well. Ask every happy customer and send a direct link. Then make your name, address, and phone identical across every listing, character for character. If you're "Sunrise Bakery" on your site, "Sunrise Bakehouse" on Facebook, and a third phone number on an old directory, a model can't tell whether that's one business or three, so it hedges and leaves you out.

Let AI crawlers actually reach your pages

None of this helps if the tools can't read you. Check that your robots file isn't accidentally blocking AI crawlers from the pages you want cited, and that your key information isn't trapped behind a login or a script that never renders as text. Some sites also add an llms.txt, a plain-text map that points assistants to their cleanest, most important pages. Support is still uneven and it won't force a citation, but it's low effort. Here's what llms.txt is and how to make one.

Is answer engine optimization just SEO with a new name?

Mostly it's the same work aimed at a new prize. Answer engine optimization (also called AI search optimization, or GEO, generative engine optimization) is the practice of making your business easy for AI to find, understand, and cite. Most of what helps you rank on Google, clear pages, real reviews, consistent listings, structured data, also helps an AI understand you. The overlap is large enough that you shouldn't treat AI search as a separate project bolted on the side.

But the win condition is different, and that's worth taking seriously. With classic SEO the reward is a click from a results page. With answer engine optimization the reward is being the source the assistant summarizes and names, sometimes with no click at all. So you write for the answer, not just the ranking: put the plain answer where a machine can lift it, and make sure your name travels with it.

How do I appear in Perplexity and other answer engines specifically?

The same fundamentals carry across every tool, so don't build a separate plan for each. To appear in Perplexity, which cites its sources inline on almost every answer, the levers are the ones above: pages that answer the question directly, genuine references from sites it crawls, and clean structure it can read. ChatGPT leans on its training plus live web results; Google's AI Overviews and Gemini lean heavily on Google's understanding of you, so a complete Google Business Profile does a lot of the work there, especially for a small business. Different front doors, same house: be clear, be corroborated, be readable.

How long does this take, and can anyone guarantee it?

No one can guarantee it, and be wary of anyone who promises a spot in an AI answer. Nobody controls what a model says, and there's no paid placement inside an AI Overview or a ChatGPT reply today. Site and listing fixes can help within weeks, as pages get re-crawled and tools pull fresh results. The deeper signal, being genuinely referenced and reviewed across the web, builds over months, the way real reputation always has. It's slow on purpose, and that's the point: a competitor can't copy it overnight, and anyone selling a shortcut is selling link spam models increasingly ignore or fake mentions that get you dropped.

What should I actually do this week?

Pick three things and finish them. Make your business name, address, and phone identical everywhere. Complete your Google Business Profile with real photos and a few honest reviews. Rewrite your homepage and one FAQ page in plain sentences that answer what customers actually ask an assistant. That trio covers the three things every engine checks before it names you: consistency, presence, and legibility.

This is the gap Laspi's WordPress plugin is built to close: it writes SEO- and GEO-shaped articles for your site, grounded in your real business and structured the way AI reads, so the answers to your customers' questions actually exist on pages a model can cite. You still decide what's true about your business. It just makes sure that truth is written down clearly, and often enough to be found.

None of this is a trick, and that's the good news. You don't have to out-clever an algorithm. You have to be a real business that's easy to describe, easy to verify, and easy to read, in the words your customers use and in the places machines look. Do the boring things well, and getting cited in an AI answer stops being luck and starts being the natural result.