SEO Attribution in the AI Era: How to Measure What Actually Drives Growth

Search doesn't work the way it did two years ago. A lot of it now happens inside a chat window: question, answer, follow-up, decision, and the website visit that used to sit in the middle of that process is optional now. That changes what "measuring SEO" even means.

For most businesses, SEO reporting still runs on last-click attribution: whatever the customer clicked right before buying gets all the credit. That was already a rough approximation of reality. Now, with a real share of the journey happening inside AI tools before anyone reaches a website, it's missing more of the picture than it's capturing.

This guide covers what SEO attribution actually means, which model holds up better, and what's worth tracking now that a chatbot might be doing part of the job a search results page used to do.

What is SEO attribution, and why does it matter?

SEO attribution is the process of identifying and assigning credit to the organic search touchpoints that contribute to a user's journey and eventual conversion. When a customer converts, they didn't arrive out of nowhere. They probably found you through a Google search weeks ago, read a blog post, came back later, and eventually bought. Attribution is the method you use to work out how much of that sale belongs to SEO, and how much belongs to everything else along the way.

The attribution models: What’s actually available (And what isn’t)

There used to be several different ways to model attribution. Most of them don't exist in Google Analytics 4 (GA4) anymore. Google quietly retired four of them back in 2023.

Here's what actually exists today:

In short: you now have two real choices, not seven.

Last-click attribution

Last-click attribution is the practice of assigning 100% of the conversion credit to the final touchpoint a customer interacts with before they buy, typically a branded search or a direct visit.

It's the same logic as giving all the credit for a championship win to the player who scored the final goal, and ignoring every pass, tackle, and set-up play that got them there. Here's what that looks like in practice:

1. A customer searches a general question, finds your blog post, and reads it.
2. A week later, they see your brand on social media and start following you.
3. A month later, they get an email and browse your website.
4. Finally, they search your brand name directly and buy.

Last-click hands 100% of the credit to step 4, the branded search. The blog post that actually started the relationship gets nothing. Read the numbers this way, and you might conclude your content isn't working and cut the budget, when really it's the thing bringing customers in the door.

Data-driven attribution

Data-Driven Attribution is the practice of assigning conversion credit across multiple touchpoints based on their actual, measured contribution to the sale, calculated by comparing the paths of users who converted against those who didn't.

Instead of following a fixed rule, GA4 looks at your real customers and works out which touchpoints genuinely helped drive the sale. It's tailored to your business, not a generic formula.

So which one should you actually use?

For most businesses, Data-Driven Attribution is the better option. Instead of following a fixed rule, it looks at your real customers, comparing the paths of people who bought against people who didn't, and works out which touchpoints genuinely helped drive the sale. It's tailored to your business, not a generic formula.

There's a catch worth knowing about. DDA needs enough sales data to work well. If your business doesn't get many conversions each month, common for a lot of Australian small and mid-sized businesses, especially in B2B or services with a longer sales process, DDA can behave a bit erratically, and its numbers may bounce around from month to month without much real change underneath.

A simple rule of thumb:

• Good conversion volume: Use Data-Driven Attribution. It's the most accurate picture available.
• Low conversion volume: Last-click can be more stable and easier to trust in the short term. Just remember it still undervalues your top-of-funnel content, so don't use it alone to judge SEO's worth.

Either way, spend time in GA4's Conversion Paths report. It shows the actual sequences of channels that lead to a sale, and you'll usually see organic search showing up early and in the middle of the journey, not just at the end.

How AI is changing the picture

The customer journey used to be simple: search, click a result, land on your website, convert. Now it often looks more like this: someone asks an AI tool a question, gets an answer, asks a follow-up, gets another answer, and makes a decision, all without visiting a single website. They might search your brand name later. They might not visit at all.

Even the best attribution model can only credit a touchpoint that shows up in your data. A growing share of the journey now happens entirely inside AI tools and never reaches your website as a click.

A few things are worth knowing right now:

• AI Overviews reduce clicks even when you're ranking well, a shift we break down in our guide to adapting SEO for AI Overviews and zero-click SERPs. The answer sits above the search results, and for a lot of queries, it's good enough on its own."
• Your AI traffic is probably already hiding in "Direct" or "Referral." GA4 doesn't automatically separate it out. A practical fix is setting up a custom channel in your analytics for known AI sources like ChatGPT, Perplexity, Gemini, Claude and Copilot, so at least you can watch that traffic as its own group rather than losing it in the noise. Do the same in your CRM, so any leads from those sessions get tagged properly too.
• Clicks from Google's own AI Overviews still count as regular Organic Search. Google Search Console shows how often your pages appear inside AI Overviews, but only impressions, not clicks. You can see you're being shown. You can't yet see if it sent anyone your way.
• Even when someone does come back later, you often can't tell what triggered it. A customer might read an AI answer, close the tab, and search your brand name a week later. That branded search gets logged. The AI answer that actually planted the idea doesn't.

The honest version: some of SEO's value right now is simply not visible in any dashboard. That's not a reason to stop measuring; it's a reason to widen what you're measuring beyond clicks alone, which is exactly what the next section covers.

What to actually track from here

Chasing a perfect click-by-click number isn't realistic anymore, and that's fine; it's part of why we've argued for measuring marketing ROI beyond clicks in a zero-click world.  The goal shifts from proving every touchpoint to building a wider set of signals that, together, tell you whether SEO is working, even when a chunk of the journey never shows up as a session. Here's where to actually put your attention:

1. Confirm your GA4 attribution model: Go to Advertising > Attribution and check you're set to Data-Driven, or make a deliberate call to stick with last-click if your conversion volume is low.
2. Check the Conversion Paths report monthly: Look at where organic search shows up in the journey, not just where it closes.
3. Look at your Direct traffic with fresh eyes: Some of it is probably AI-referred traffic that isn't being labelled correctly.
4. Check if your AI Assistant channel is populating: If it's not, don't assume you have no AI traffic. It might just be uncategorised.
5. Check Search Console's Generative AI report: It won't show you clicks yet, but it will show you whether you're being surfaced in AI Overviews at all, useful groundwork for when click data eventually arrives.
6. If your sales cycle is long, connect GA4 to your CRM: For considered purchases, the journey doesn't end at a form fill. Tie it through to the actual sale to see SEO's real value.

Conclusion

Last-click was never telling you the full story, and AI search now means parts of the story are harder to see than ever, no matter which model you use. The goal is not a perfect number. It's giving credit to the work that actually brings customers in the door.

Pick the right model for your business size, get honest about where your AI traffic is currently hiding, and keep an eye on the tools- GA4's AI Assistant channel and Search Console's Generative AI reports- that are actively catching up to fill the gaps.

At Digital Treasury, we help businesses cut through the noise and uncover the true value of their SEO efforts. If you're ready to track results with clarity, optimise smarter, and grow faster, book a strategy session with us today.

‍