Optimizing for visibility in generative AI platforms seems easy enough. Enter a sitemap URL into Bing Webmaster Tools and Search Console, and, voilà, Microsoft Copilot, Google Gemini, and even ChatGPT access and reference the content.
But Clint Butler reminds us it’s not that simple. He says the problem is knowing which platform to optimize, as the tactics differ. And rarely do those platforms mention the source, much less link to it. Google, incredibly, uses one source to power AI Overviews but links to others in the citations.
Clint is a decorated, 20-year military veteran who now runs Digtalteer, a prominent search-engine-optimization agency. I spoke with him last month on the state of AI search, ecommerce optimization tactics, schema markup, and more.
Our entire audio dialog is embedded below. The transcript is edited for clarity and length.
Eric Schwartzman: What’s working now for search engine optimization?
Clint Butler: The elephant in the room is artificial intelligence, large language models, and the types of content you can make with them — text, video, audio — and how business owners can leverage that within their marketing, including search.
You and I are old school. We’ve written content by hand. Now we can put our knowledge into the models and get a nice content base. It doesn’t take us as long to push out articles because we can edit, review, and publish them more efficiently.
Schwartzman: Is it more about using AI to generate content, or should we optimize our existing content to appear in LLMs such as ChatGPT, Claude, or Gemini?
Butler: The problem with that is knowing which model to optimize. If you’re going for the Microsoft version, admission is simple. Add your stuff to Bing Webmaster Tools, and your site is in Copilot.
As far as I can tell, Google uses its top-ranking organic pages to generate AI Overviews. But it then links to other pages for citations!
It’s worth asking whether achieving a top organic ranking is worth it. Google should fix that citation practice.
Schwartzman: As for LLMs, we’ve got OpenAI, Gemini from Google, and Llama from Facebook. Presumably content would enter all of the training sets if it hit LinkedIn, Bing Webmaster Tools, and the web for Google, as well as Facebook and X.
Butler: That’s another misnomer. Let’s say you have an ecommerce site selling scented candles. You publish a lot of content about scented candles.
Just because the AI bots crawl you doesn’t mean you will be in the data sets or properly cited. Gemini is the most prominent AI to get in front of people because it powers AI Overviews.
And, again, just because Google uses your content doesn’t mean you get a link. The rules vary depending on the LLM. Bing and Copilot are a bit better at citing the source. You would more likely get a click to your scented candles out of Copilot than Google’s Gemini.
Schwartzman: Are you saying SEO becomes less important? Should merchants look at other channels for traffic?
Butler: The SEO priority for ecommerce merchants is Google Merchant Center. That’s what Google uses to populate many search snippets and product carousels where AI Overviews typically don’t appear.
Ecommerce merchants trying to generate traffic and awareness through informational marketing may have a problem, although the latest data that I’ve seen from AccuRanker and Sistrix suggests it’s not insurmountable if you’re in the top three listings. The click-through rate for those placements is only down from 26% to 24% with AI Overviews.
So long as you’re in the top three, you’re okay.
Schwartzman: You work with ecommerce clients. What are their common SEO mistakes?
Butler: It depends on the platform, but product names and category optimization are the two big ones. Say you have a scented Halloween candle and call it “Eric’s Freaky Friday Halloween Candle” versus simply “Jasmine Candle.”
People search for jasmine-scented candles but not for your fancy name. Use that fancy name in your product description but not the name.
Schwartzman: Is there a balance? Say I’m searching “jasmine candle,” and the SERP choices are “Jasmine Candle” and “Jasmine Candle: Eric’s Freaky Friday Jasmine Candle.” I’m likely going with the one that isn’t so vanilla.
Butler: That’s true. But the problem is that Google will likely truncate after “Jasmine Candle.” Searchers won’t see the “Freaky Friday” part. Experiment instead with inserting sizes or maybe even colors.
Schwartzman: What is a healthy click-through rate for a query such as “scented candles”?
Butler: Search Console suggests 2% to 5% is good, but keep in mind where that data comes from. It’s not 100% accurate. In my experience, take what you see in Search Console with a grain of salt. If you want 100% accurate data, run a Google Ads campaign.
Search Console is useful for basic decisions, however.
Schwartzman: You’re an expert on Schema.org structured data. You offer a course on how to write schema. You’ve even ranked blank pages just with schema. But it’s not practical for ecommerce merchants with large catalogs to write advanced schema for each item. What should they do?
Butler: Much of the advanced product variant data on Google search results come from Merchant Feed, not from schema inserted by the seller. So the first step for merchants is setting up Google Merchant Feed.
Beyond that, merchants can use schema selectively for rich snippets. Say a seller has 100 products, and 10 generate most of the revenue. Implement a nice product schema on just those 10.
There are helpful tools, too. Shopify users can leverage its Google & YouTube tool. Fill in the fields — pricing, shipping, categories, imagery — and the tool will populate Google Merchant Feed, which, again, drives SERP carousels.
Schwartzman: How can readers get in touch?
Butler: My agency is Digitalteer.com. I’m on LinkedIn and X.