While Google still dominates the search industry, AI platforms like ChatGPT, Perplexity, Copilot, and Gemini are taking slices of the pie. And those slices could eventually have a big impact on traditional search traffic.
You might have even started to see some website visits trickle through from AI in your analytics. As more and more businesses see a decline in organic traffic, looking for new sources–like AI referral traffic–will become more important. Want to get started? Follow these tips to help drive referral traffic from AI search and measure results.
Contents
What is AI referral traffic (and why it matters)
AI referral traffic is any traffic driven by an AI tool or platform. This includes ChatGPT, Perplexity, or any other platforms we mentioned earlier.
As more people use AI search engines to find information, and as more AI search engines pop up, you have the opportunity to drive traffic from those searches.
In the example above, we searched for how to fix a sink leak on ChatGPT, and it provided this list of sources. From here, we could click off to one of the source links to get more information about our inquiry.
Are you using AI the right way? Free download >> Emergency Guide to AI in Marketing
6 ways to get more referral traffic from AI search
Try these tips to see more referral traffic from AI search to your website.
1. Make it easy for AI crawlers to find content on your site
AI crawlers (or spiders) sift through an endless supply of pages on the web, trying to match users with the most relevant content. If your site is slow or hard to navigate, you’re basically presenting crawlers (and users, for that matter) with a mess.
Here’s how to make it easy:
Improve page speed and mobile friendliness
Page speed is a Google ranking factor, so your site needs to load fast for Google. But it should also be fast for AI. Crawlers are built to scan and summarize content quickly, and there’s a chance they might time out and not fully crawl everything if your site is slow.
Google’s Page Speed Insights tool makes it easy to see how fast your pages load and what elements can be addressed. Simply plug in the URL of the page you want speed insights for and get results in just seconds.
Some quick fixes:
- Compress images with tools like TinyPNG or Squoosh
- Try using lazy loading
- Use a content delivery network (CDN) to speed up load times globally
- Eliminate unnecessary plugins and scripts. If your site is running five different tracking codes, it might be time to clean things up.
Your site also needs to be mobile-friendly, and the most common way to achieve this is through responsive web design. A responsive site automatically adjusts your page sizes, layout, button types, and more to work better on mobile devices.
Plus, AI search is built to answer questions on the go, whether someone’s using voice search in a car or scrolling through results on their phone. If and when you show up in AI search results, you don’t want visitors to arrive at a clunky design.
Use internal links for optimal crawlability and discoverability
Internal links connect different pages of your website, making it easier for AI crawlers to navigate between them. These links also help the crawlers find all the content on your site. If a page exists without any links pointing to it (i.e., it’s an orphan page), a crawler might not see it.
You can add internal links in your navigation, footers, within each piece of content, and in other places.
As the crawlers move through your site, they collect more information about your content that influences whether or not you’ll be mentioned and listed in AI results. Internal links have great benefits for SEO performance.
A well-linked site helps AI understand the relationships between your content, making it easier for the platforms to cite your pages when users search topics related to your site.
Tip: Make sure you can reach any page of your site in three clicks or fewer from any other page. One way to do this is through topic clusters.
2. Target conversational queries
91% of all search queries are considered long tail, or conversational. Sure, people still search broad terms like “garden hoses.” But with AI, people talk to it like they would a friend, asking questions like:
- What’s the best garden hose that won’t kink?
- Which hose is best for a big backyard?
Tools like ChatGPT, Perplexity, and Copilot can generate ideas and variations of how users ask about a topic. Even though these don’t have actual search data (like volume or intent), the ideas can still be worth considering since they come from AI itself–the very place you want to get traffic from.
For example, instead of targeting just “garden hoses” in your content, ask an AI tool to provide questions and conversational phrases about the topic. You might get responses like:
- What’s the best garden hose for water pressure?
- Which garden hose lasts the longest?
- What size hose do I need for my yard?
AI search engines will return clear, direct answers in response to conversational queries. So, your page content should do the same.
Also consider the “People also ask” (PAA) section in Google results for more ideas. Google practically hands you a roadmap to conversational search. The PAA section lists related questions you can use and answer in your content.
As an example, when I search “garden hoses,” this is what the PAA section shows. And when I click on a question, even more questions will appear at the bottom of the list.
Need help finding the right keywords? Try our Free Keyword Tool!
3. Improve the E-E-A-T signals of your content
A Seer Interactive study about what drives brand visibility in AI answers showed that brands ranking on the first page of Google have a strong correlation.
And when they filtered the data to focus on solution-oriented websites, there was an even stronger correlation, proving that quality content is important.
What does that mean for you?
To rank high in search engines, and thus appear in AI search, your content needs to be the most reliable, well-supported option for the topic. And this is where E-E-A-T comes in: Experience, Expertise, Authoritativeness, and Trustworthiness. These are the guidelines Google uses to evaluate content quality.
Thin, basic content doesn’t cut it. AI search prioritizes content that fully answers someone’s query.
Let’s say you sell project management software, and you’re creating a comparison piece titled “Top 10 Project Management Tools.” Instead of just listing the tools with a description and price for each, you should aim to stand out by putting together helpful, in-depth information that’s backed by data and reputable sources.
For the article, you could improve its E-E-A-T signals by:
- Including side-by-side feature comparisons (e.g., how the tools stack up for workflow automation).
- Sharing common pitfalls to avoid, like failing to onboard your team properly.
- Tailoring recommendations for freelancers, startups, and enterprises.
- Using stats with cited sources.
- Quoting industry experts.
- Referencing case studies to back up claims.
- Using images and videos to support concepts.
4. Earn more brand mentions on the web
Stephen Wolfram wrote that ChatGPT “looks for things that in a certain sense match in meaning.” It’s trained based on what it has learned from the web.
And Rand Fishkin of SparkToro created a video about large language models (LLMs) using mentions as their currency.
So, to get links and traffic from LLMs, you need to get your brand mentioned on sites that cover topics related to your business. You want to signal to AI that you’re a notable brand in your space.
If you sell project management software, Google some of your target keywords and see what ranks. Reach out to those sites to:
- Engage with industry blogs and media. Get featured in “Top 10” roundups and sponsored features.
- Collaborate with well-known project management experts on webinars, podcasts, or guest blog posts.
- Cross-promote with complementary brands. Team up with a time-tracking app or collaboration tool.
5. Optimize for featured snippets and AI Overviews
If your content is a wall of text, it’s like handing over a puzzle with missing pieces. But if you break things up with clear headings, short paragraphs, and easy-to-read lists, you’re basically spoon-feeding the AI exactly what it wants to show users.
Look at the AI Overview for the query “What size hose do I need for my yard?” It’s all bite-sized information with headings and lists. You can’t argue that it’s not organized. Or not straight to the point.
Here’s how to optimize for those kinds of results:
- Answer the intent of the query without any fluff. AI is going to concisely give users the information they want, so you should too. Get to the point right away.
- Use H1s, H2s, and H3s properly. Your main heading (H1) should set the stage, while subheadings (H2s and H3s) break content into logical sections. You can even go H4 and beyond if the content is deep enough to warrant that.
- Put sets of bite-sized information in bullet points, like this very set you’re reading now.
- Put sequences and steps into numbered lists.
By structuring your content with AI crawlers in mind, you’re increasing the odds of getting featured in AI-driven results and attracting human visitors to your site.
Want to learn how AI can supercharge your ads strategy? Free guide >> 9 Powerful Ways to Use AI in Google & Facebook Ads
6. Adapt your content strategy for specific AI platforms
Pay attention to how the various AI platforms respond and what they respond with. Tailor your content to those characteristics.
For example, when I start searching on Perplexity, it provides auto-suggestions. And when I search “project management for,” this is what comes up:
If you’re targeting certain types of businesses, this can be a goldmine of ideas to create content around. Experiment with your target keywords to see what comes up.
Perplexity also produces related questions at the very end of its responses. You could create content that answers these exact questions. These are the questions that appeared at the end of “project management for engineers:”
As another example, take ChatGPT.
Chris Long of Go Fish Digital shared a post on LinkedIn, suggesting you search your target keywords in ChatGPT (one at a time) to see which ones trigger the search function.
Once you learn which ones trigger a search response (as opposed to chat), look at the type of results. Are they listicles? Are they comparison pieces?
Take inspiration from what’s there and build a strategy to optimize for those on your site.
How to track site traffic from AI search
Naturally, you should monitor the traffic you get from AI to see:
- Which platforms bring you the most traffic.
- The types of content that are most effective for getting AI traffic.
- The topics generate the best quality AI traffic–look at conversions, time on site, etc.
If you use GA4, here’s how to identify traffic from a few of the platforms:
I recommend creating a custom channel group, which will allow you to easily compare data against other channels like organic search, direct, etc.
Skip down to steps 3 and 4 in Nico Brooks’ guide, Tracking AI Traffic in GA4, to see how you can do this. It’s easier than you think!
And then you can break things down from there, by source (AI platform), by landing page, etc.
Start driving more AI referral traffic
This will likely include a bit of trial and error. Figure out what’s most effective for your site in bringing in AI traffic, and then do more of that.
While organic and paid traffic from search engines like Google aren’t going anywhere, it doesn’t hurt to incorporate some of these tactics into your strategy to help you capture traffic from these often-used AI tools.
Want to learn how we can help? Reach out for a demo today.