Measuring What Matters in the AI Search Era

Search is shifting, traffic patterns are changing

In our recent post, Is AI Changing How People Find Content?, we discussed how online search metrics are shifting as AI assistants grow in popularity. The biggest takeaway is that a drop in search engine traffic to your website is to be expected, but visitors coming to your site from AI assistants are much more engaged than traditional visitors from search engines.

For websites whose primary goal is to inform or provide information to searchers (e.g., extension resources, research hosting, or educational content), this traffic shift hits differently than it does for lead-generation sites (e.g., colleges, departments, program enrollment). AI assistants summarize your content and give searchers the answer they need without ever sending them to your website. This is a fundamental change in how your content reaches people.

It doesn't mean your content strategy is failing! It means the metrics you use to evaluate website engagement need to evolve.

The mindset shift: from volume metrics to quality metrics

For a long time, pageviews were the standard measure of content success. More visitors generally meant your content was reaching the right audience. That made sense when discovery happened primarily through search engines.

As AI assistants begin to absorb the search activity that used to send visitors to your website, pageviews alone are no longer a reliable indicator of whether your content is doing its job. For informational websites especially, the more meaningful question is no longer how many people visited? It's did the people who visited actually engage with what they found?

The visitor journey has changed

The traditional visitor journey assumed that discovery and evaluation happened on your website. A visitor would find you through a search engine, scan your page to decide if it was worth their time, and then either engage or leave. Your goals were built to measure that process (pageviews, bounce rates, and session duration).

The old journey

1. Search - visitor enters keywords into a search engine
2. Scan - visitor reviews a page of results and selects a link
3. Land - visitor arrives on your website
4. Evaluate - visitor decides whether the content is relevant
5. Engage or bounce - visitor reads and explores further or leaves
6. Return - if the experience was valuable, the visitor comes back

The new journey

1. Prompt - visitor asks an AI assistant a question in natural, conversational language
2. Synthesize - the AI summarizes from multiple sources, including potentially your content, and generates an answer
3. Cite - the AI may or may not surface a link to your website as a source
4. Land informed - if the visitor does click through, they arrive already knowing the core answer, with a specific follow-up need
5. Engage - the visitor is looking to verify, expand on, or act on what the AI already told them
6. Return - if your content satisfies, the visitor is more likely to come back directly, establishing you as a reliable source and bypassing search entirely

How the journey change impacts website traffic

The most significant change is that the discovery stage now happens entirely without you. Your content may be informing answers that thousands of people receive without ever generating a single pageview on your site.

This creates two distinct audiences you need to think about separately:

• The AI-mediated audience - people who receive your information through an AI assistant and never visit your site. For extension and research websites, this audience may actually represent your mission being fulfilled, even if we can't measure it with your analytics.
• The actual visitor - someone who clicks through from an AI citation, types your URL directly, or finds you through a link. This person is further along in their thinking than a traditional search visitor. They know roughly what they're looking for and have a reason to be on your specific site. This is why we consider them more engaged and willing to take action on your website.

Where you should focus your efforts to maximize return

In the old journey, the critical moment was the first few seconds after landing. Could you hold their attention long enough to prove your content was worth reading?

In the new journey, they already believe you're a credible source. They think you might have something valuable. The question is whether your content is set up to keep them, encourage exploration, and build enough trust to bring them back.

For informational websites, that means:

• Depth over breadth - a thorough article on a specific topic is more valuable than a surface-level overview
• Clear paths forward - when a visitor finishes a page, is it obvious where to go next?
• Highlight authority - make it clear what makes your website a trusted source. Use author credentials, citations, publication dates, and institutional affiliation. These things matter more when a visitor is coming to verify or go deeper.

Where your analytics fit in

The new journey means your analytics will only ever show you part of the picture. Accept that limitation and build your measurement strategy around what you can see:

• Who is arriving, and from where?
• What do they do once they get there?
• Do they come back?

Those three questions tracked consistently over time will tell you more about your content's real impact than any pageview total ever did.

Key Performance Indicators (KPIs) to reconsider

• Pageviews - A pageview happens when a visitor views a page on your website. Just because your content has fewer overall pageviews doesn't necessarily mean it's not important or resonating with searchers. AI assistants may be using your information as a source, and visitors are getting the answers they are looking for. That's AI doing its job.
• Bounce rate - A bounce is counted when only one page is viewed on your website. Bounce rates need more context to be useful. Traditionally, a high bounce rate indicated that visitors didn't find what they were expecting and left your site immediately. Now we know that visitors coming to your website from AI assistants have much more context than traditional visitors. They may be getting the answer they are looking for straight away and moving on.
• Session duration - Session duration is the time spent on your website. It isn't the most useful metric because an informed visitor may need less time to get value from your content. Like bounce rate, it needs channel context to be meaningful. A short session from an AI referral visitor may still represent a successful visit.

KPIs to prioritize now

• Acquisition channel type - people come to your website from search engines, AI assistants, other websites, directly, social media, etc. In the visitor journey, the stage before a person lands on your website and triggers a pageview is an impression. Where are visitors first seeing you? Learn more about the new AI assistant channel tracking.
• Conversion rate - define what a successful visit or desired action for your website is. For informational sites, a conversion might be a resource download, a newsletter signup, or simply a second page visited. The point is to decide in advance what a successful visit looks like so you have something meaningful to measure. Many clients are getting by with avoiding this metric, but in the shifting search landscape, it's time to talk about what goals you really want your website to achieve.
• Conversion rate by channel - in Is AI Changing How People Find Content?, we shared that visitors from AI assistants are 4.4 times more likely to complete your desired action. Monitor where the visitors who convert come from and focus your communication efforts there.
• Return visitor rate - this is an important metric for websites whose goal is to inform. If visitors find your content informative and reliable, you should see a high return visitor rate.
• Engagement metrics - monitor the average number of actions (an action in Matomo is any internal link click, document download, form submission, or website search) completed per visit and the time spent on your website by content type (news, blog, how-to, program page, support page, etc.).

New questions your analytics should answer

Once you've updated your tracking goals, use your analytics to answer these questions on a regular basis:

• Are AI visitors completing goals at a higher rate than other channels?
• Is your organic search traffic declining, and is AI traffic offsetting it?
• Which content types have high AI entry traffic but low conversions? Can you determine why?

We can help you find the answers

Not sure where to start? The CALS/LAS Web Team can help:

• Channel engagement analytics dashboard - Upon request, we'll share a custom dashboard to help you identify trends in your website engagement by channel type.
• Custom reporting - if you're looking for something more specific from your analytics, we can build reports or dashboards tailored to your website's goals.
• Request a strategy meeting - if you're not sure where to start, we can walk through your data together and help you define what success looks like for your specific site.

Contact websupport@iastate.edu to get started.

Fewer pageviews doesn't mean failure

For informational websites, success has always been about whether people are finding and trusting your content, not just whether they're showing up in large numbers. The AI search era makes that more important than ever.

The goal isn't to chase pageview totals back to where they were, but to build a clearer picture of whether your content is reaching the right people, earning their trust, and bringing them back. If your analytics strategy is built around that question, you're in a much stronger position than a report full of declining volume metrics would suggest.