Sooner or later, a client asks, “Can you get us on voice search?”
A few years ago, the process was straightforward. You wrote the way people talk and tried to get the page into the short reply answer box Google shows above the regular results. Voice assistants could read that answer aloud, and voilà, you were in.
But with the rise of AI assistants like ChatGPT and Gemini, it’s not that simple anymore. These tools can pull information from several sources when answering a question. Your client’s site needs to explain what the business does in plain terms, and mentions on other websites can help, too.
An SEO masterclass can teach you how to help clients with both.
How Voice Search Has Changed Since Siri and Alexa
The early voice assistants had short memories. Siri, Alexa, and Google Assistant handled one question at a time. You’d ask, and they’d read back a quick reply. It often came from one website or a business listing. Ask a related question right after, and they’d act as if they’d never heard you.
Today’s assistants keep up. Gemini comes on many Android phones, and Copilot is built into Windows. ChatGPT has a voice mode in its app. You can ask a question and add more detail. The next answer builds on what you already said.
One thing hasn’t changed. The assistant still needs something to read. AI tools can pull from websites and business listings, so the groundwork you’d do for SEO still counts. As you learn AI search optimization, you’ll build on it instead of starting over.
What AI Search Still Has in Common with SEO
Before an AI tool can point people to your client’s website, it has to find the site first. That part works a lot like regular SEO. Google, for example, only links to pages in its AI answers if those pages also appear in its normal search results. If search engines can’t reach a client’s pages, AI tools may miss them as well.
Once an AI tool finds the site, the content needs to be useful. Explain what the business does and answer customers’ questions in plain words. Off your site, keep hours and contact details accurate on Google and other places people check.
Reviews and mentions on other websites can shape what AI tools say about a business. But nobody outside these companies knows how much influence any single review or mention has. If you’re new to how search engines find and read websites, start with an SEO course for beginners. Once you have the basics down, the AI side makes a lot more sense.
Skills Worth Learning for AI Search
AI search asks you to use some familiar SEO skills but in new ways. Most aren’t hard to pick up, but they take practice.
Run real customer questions through ChatGPT, Gemini, Claude, or Perplexity. Check whether your client’s business gets mentioned and whether the details are right. Try the same questions again next month. Compare the answers.
Google Analytics gives you another place to look. It can show visits from sites like chatgpt.com or perplexity.ai, so you can tell clients whether AI search is sending people their way. For a local business, the numbers may start small. That’s fine. You’re building a baseline.
You’ll also hear a lot about structured data. This bit of code labels details like a business’s address and hours so search engines can read them more easily. It’s useful for that, but it isn’t a ticket into AI answers. Google says you don’t need special code to show up there.
What a Monthly AI Check Can Look Like
Let’s say you manage the website for a dental office. Each month, run five questions a new patient might ask through ChatGPT, Claude, and Gemini. One might be, “Is there a family dentist nearby that’s open on Saturdays?” See whether the office comes up. Check what the tools say about it. If an answer links to sources, save those too.
If something’s wrong, track down where it came from. Old hours might come from a business listing or an outdated page on the site. Fix what you can there, then check again next month.
Send the client a short report on what you checked and what you changed. Half a page is plenty. The client can see what you’re doing and what their monthly fee covers.
What You Can Promise Clients and What You Can’t
A client may ask you to get them to the top of ChatGPT. Don’t promise it. You don’t control the answer, and it can change from one search to the next.
To illustrate just how impossible it is to guarantee AI results, data analyst and founder of Spark Toro, Rand Fishkin, shared that you’d have to ask AI the same query an average of 1,500 times before you received the same answer.
Promise what you can control. You can keep the site easy for search engines to reach and the business information accurate. You can also write pages that explain what the company does, then check each month how AI tools describe it.
Set expectations early, so a slow month doesn’t turn into an awkward conversation.
Adding AI Search to What You Offer Clients
Your clients are already asking about AI search. If you build or manage their sites, you’re closer to offering it than you might think.
The opportunity is in the ongoing work. AI answers change. Check how the business shows up each month and report what you find. A one-off project can become a monthly service.
The AI side takes practice. You need to know what’s worth checking and what you can confidently tell a client.
An LLMO masterclass can help you learn the process before you sell the service. Then, when a client asks about AI search, you’ll know what to offer and what to promise.