Brands are worried about how they show up in LLMs as calls for AIO strategies reach a fever pitch.
There’s urgency to this topic: an estimated 30% of shoppers are using AI for product research, according to a recent Brainlabs report. A year ago, that figure was just 12%. With any new popular methodology comes a slew of misconceptions.
Digiday caught up with four AI visibility agency experts to debunk some of them:
**Myth: There is a proven AI visibility playbook. **
The old SEO playbook — keyword research, backlink building and content volume — no longer guarantee a top spot in search results. But agency execs say an AI visibility playbook is an evolution of SEO. Many of its foundational principles still apply, like quality backlinks, high-quality content and site health, says Nate King, a manager of product strategy at RPA.
“It’s 80 to 90% of the same [SEO] tactics and strategies is what you’re doing for AI search optimization,” King said, “but that 10 to 20% is actually super critical and important.”
Agencies are testing a few things, like structured data markup, earned media and citations. For example, the team at Go Fish Digital is currently exploring the through line between paid social and AI visibility, David Dweck, president at Go Fish Digital, told Digiday.
Myth: AI visibility is about optimizing your website.
LLM crawlers pull from everywhere across the web, whether it be Reddit, YouTube, Wikipedia or a brand’s own site. Solely focusing on building out FAQ pages and over optimizing product feeds is a red herring approach. The bigger fishes are earned media and citations across social media, in news headlines and even traditional search engines, per the execs.
“We’re seeing owned, we’re seeing earned, we’re seeing social, we’re seeing user generated, we’re seeing commerce. It’s all influencing these answers,” said Heather Physioc, chief discoverability officer at VML.
Meaning, a cross-channel marketing strategy is en vogue when it comes to AI visibility.
Myth: There are silver bullets.
If you ask Physioc, “anyone promising anything in this world is a liar,” and that’s especially true when it comes to promises around AI visibility. That hasn’t stopped the flood of generative engine optimization (GEO) startups or AI visibility hacks. Agency execs say any AI visibility hacks aren’t a silver bullet and may only work briefly.
Keyword stuffing, backlink games or similar tricks won’t work in the AI visibility-era, where LLMs function more like humans. Even things like llm.txt file, a standardized text file placed in a website to help AI crawlers understand site structure, has shown little impact, execs said.
Adding more schema or semantic markup isn’t a winning strategy either, according to Ryne Landers, expert lead of SEM/SEO at Markacy.
“My main hypothesis is you’ve got to build relevant content and not manipulate. They’re too smart,” said Dweck. He added, “Brands really need to focus on providing value to the marketplace, and the LLMs will surface that.”
Platforms themselves have taken notice of some marketers’ so-called hacks. For example, Reddit has already started cracking down on so-called AI marketing slop.
Myth: AI visibility should be measured like SEO.
In what’s being called a zero-click environment, marketers are scrambling to find new ways to determine the downstream impact of AI visibility. Here, click-thru-rate doesn’t apply as it did with traditional SEO. Instead, marketers are looking at things like share-of-voice, tracking things like how often a brand appears across the various LLMs (and within what context).
“The way a lot of marketers evaluate media spend or ad spend is sentiment, share of voice, and awareness, or recall or top of mindedness, mindshare,” said Markacy’s Landers, “You have to treat the LLMs the same way.”
Myth: Brands should write for AI instead of humans.
While more shoppers are turning to LLMs to start their product hunt, there’s still a human sitting at the other end of the computer and brands should keep that in mind, per the execs. Write content for human beings rather than AI bot crawlers. “Neglecting the human at the end of the process or the human who is asking the question is a recipe for disaster,” said VML’s Physioc.
More in Marketing
Michaels claims Google-powered AI assistant doubles conversion rate of traditional search
Michaels shared early results of its new Google Gemini-powered AI assistant.
[
How AI costs are quietly reshaping principal media deals ](https://digiday.com/marketing/how-ai-costs-are-quietly-reshaping-principal-media-deals/)
AI doesn’t have a price. Principal media is absorbing it anyway.
Future of Marketing Briefing: The case for and against the death of the big brand ad
Marketing has a new favorite argument: is it still smarter to talk to a million people at once, or to find the thousand people they already trust?