Part 2 in our BrightonSEO 2026 series. If you haven’t already, start with What BrightonSEO 2026 Tells Us About the Future of Search for the bigger picture.
Traffic and rankings have always been the bread and butter reporting metrics for all SEOs. Our goal has always been to drive increased organic traffic to our clients’ sites by improving the keyword ranking performance for relevant keywords, helping more people to discover their brand via organic search. This, in turn, would then ideally lead to an increase in conversions, whether that would be sales and revenue for an e-commerce client or contact enquiries for a service business. It’s a model that has been working well since I first started in SEO back in 2011. It is, however, a model that is changing. Or, perhaps more accurately, it’s a model that has already changed.
Whilst AI has been the biggest disruptor to this traditional model, the truth is that the organic landscape has been changing for some time. SERP features, Ads, Map Listings, People Also Ask, Shopping Carousels – all of these elements that have been introduced to Search Results Pages over the past decade have been disrupting the traditional model and impacting the CTR from traditional blue link organic search results. AI has simply made it impossible for us not to change the way we report and determine how successful the work we are carrying out is.
So, if you’ve noticed your traffic plateauing, or even dropping, despite rankings holding steady, or despite your SEO agency or internal team telling you that the underlying SEO work is going well, this post is for you. I want to walk through exactly why this is happening, with the data presented at BrightonSEO to back it up, and more importantly, what you should be looking at instead.
The Numbers That Made the Room Go Quiet
A few statistics came up across multiple sessions at BrightonSEO this year, and they’re worth sitting with for a moment rather than skimming past.
Jack Lingard opened his session with the big one: 93% of AI-driven searches end without a click. That’s not a typo. Of all the searches now happening through AI platforms and AI-powered search experiences, the overwhelming majority resolve themselves entirely within the AI answer. No click. No visit. No session in your Google Analytics. The user got their answer and moved on with their day.
Tom Capper, presenting research from STAT, brought a different but equally important angle: 60% of all searches are now classified as zero-click. This isn’t purely an AI phenomenon – it includes featured snippets, knowledge panels, and the dozens of other SERP features that have crept into Google’s results over the past decade. Tom’s data also showed that position #1 – the spot every SEO programme is built to chase – is only visible above the fold on desktop roughly 65% of the time now. On mobile, thanks to ads, map packs, and AI Overviews stacking on top of each other, it’s often not visible until the user has scrolled past several screens of other content.
James Yorke, in his session titled SEO Has a Measurement Crisis, said something Paul wrote down word for word: “We need to be looking at overall visits and traffic – can’t really look at solely SEO traffic anymore. Especially when looking at last click attribution.” It’s a slightly uncomfortable thing to hear from the stage at the industry’s biggest SEO conference, but it’s honest, and it matches exactly what we’ve been seeing in client accounts over the past 12 to 18 months.
Paul’s take:“James’s session stuck with me because he wasn’t dramatic about it. He simply said that last-click attribution is increasingly meaningless because so much of the value SEO creates now happens before the click, or instead of the click. That’s not a reason to panic. It’s a reason to change what we measure.”
Why This Is Happening: It’s Not Just AI
It would be easy, and frankly a bit lazy, to blame this entirely on ChatGPT and AI Overviews. The truth, as I said in the intro, is that this has been building for a decade. Here’s the honest timeline of what’s been eating into organic click-through rates:
- Paid ads have expanded from one or two slots to dominating the top of the page for many commercial queries, especially on mobile
- Map listings and the local pack now sit above organic results for any query with local intent
- People Also Ask boxes answer follow-up questions directly in the SERP, reducing the need to visit multiple sites
- Shopping carousels intercept retail queries before a user ever reaches a traditional listing
- Featured snippets and knowledge panels have been answering simple factual queries directly in Google for years
- AI Overviews are now the newest and most aggressive layer, synthesising a complete answer from multiple sources directly in the results page
- AI platforms themselves – ChatGPT, Claude, Gemini, Perplexity – are increasingly a parallel search destination entirely outside of Google
Each of these has chipped away at the click-through rate from a “good” ranking. None of them, individually, was catastrophic. Collectively, over ten years, they have fundamentally changed what a top ranking actually delivers.
AI hasn’t created this problem. It has accelerated it to the point where it can no longer be ignored or worked around with incremental tactics. It’s forced an industry-wide reckoning that, frankly, was overdue.
But Here’s the Bit Nobody’s Headline Mentions: Quality Over Quantity
This is where I want to push back gently on the doom and gloom, because the picture painted at BrightonSEO wasn’t actually one of decline; it was one of transformation, and in some respects, improvement.
Nick Beck’s session delivered the statistic that changed how I think about this entire issue: visitors who arrive at your site via AI search convert at 4.4 times the rate of traditional organic visitors.
Think about why that makes sense. A user who has had a conversation with ChatGPT, asked follow-up questions, had their objections addressed, and then clicks through to your website has already done the bulk of their research and decision-making. They’re not browsing. They’re not comparing five tabs. They’ve effectively been pre-qualified by the AI before they ever reach you. The traffic volume might be smaller, but the intent and the value is dramatically higher.
This is the central reframe of this entire post: smaller, more qualified traffic that converts at a higher rate is not a problem. It might genuinely be a better outcome for your business, even if the topline number in your monthly report looks worse.
Paul’s take:“This is the stat I bring up in almost every client conversation now. It reframes the whole anxiety around ‘traffic is down’ into ‘who is actually visiting, and what are they doing when they get there?’ Once we explain the conversion rate difference, the traffic number stops being the scary headline.”
For some time now, we have been separating out blog traffic from, say, product or service traffic, in our reporting for clients where relevant. That’s because we were already starting to spot a trend going back more than 12 months. Blog traffic was declining due to AI Overviews and the way our number one ranking posts were now being synthesised and repurposed directly in the search results.
For businesses that had been driving significant traffic numbers through their blog, this looked bad. Overall organic traffic numbers were declining, leading to some challenging questions. However, when we started to split out the reporting, we noticed that, in many cases, traffic to the key service or product pages was actually stable and in some cases, increasing.
By better understanding the overall landscape, we were able to focus on the quality over quantity debate and reassure our clients that their site was still performing well, focusing instead on conversions, sales and revenue.
This doesn’t mean that the blog content we are creating is no longer valuable. Quite the opposite, in fact. Visibility in AI Overviews and in AI Search platforms is crucial, and high-quality, original blog content continues to build brand presence and awareness, even if the clicks don’t follow.
Laura McInley’s Session: “No Clicks? No Problem!”
Paul attended a session with a title that could not be more directly relevant to this post: Laura McInley’s No Clicks? No Problem! Reporting When Traffic Declines. It was a session built entirely around the question this blog post is trying to answer for you.
Laura’s central message was that the panic clients feel about declining traffic is almost always a reporting problem, not a performance problem. If the only metric in a report is sessions, and sessions are falling because of structural changes to how search works, not because the brand is doing anything wrong, the report itself is creating a false impression of failure.
Her recommendation, echoed by multiple other speakers, was to build a reporting structure with several layers rather than relying on one number to tell the whole story. This is where the Presence, Preference, Performance framework that James Yorke also referenced becomes genuinely useful:
- Presence – are we visible where our audience is actually looking? This includes traditional rankings, but also AI Overview appearances, AI platform mentions, and SERP feature ownership (such as the local pack or a featured snippet)
- Preference – are people choosing us specifically? This shows up as branded search volume, direct traffic, return visits, and assisted conversions across multiple touchpoints
- Performance – are we converting that attention into business outcomes? This is where clicks, sessions, leads, and revenue still matter, but as the final layer, not the only layer
Paul’s take:“What I liked about Laura’s approach is that it doesn’t throw out traffic reporting altogether – it puts it in its proper place. Traffic is one input into a bigger picture, not the whole picture. Clients generally respond really well to this once it’s explained properly, because it actually gives them more confidence, not less.”
Brittany Deller: Making Measurement Impossible to Ignore
The other session from Paul’s notes that deserves a mention here is Brittany Deller’s, titled How to Make Measurement Impossible to Ignore. Her core argument was about communication, not technique and it’s one I think a lot of agencies, ourselves included, need to take on board.
Brittany’s point was simple: measurement must use business-impact language – revenue, cost, risk – not technical SEO jargon. Vague reporting stifles action. If a client receives a report full of impression counts, average position changes, and crawl statistics, they have no real way to judge whether the work is succeeding, and more importantly, no way to act on it. If a client receives a report that says “this work reduced your cost per lead by 12% this quarter, and here is the risk of pulling back,” they can make a real decision.
This is exactly the gap we are working to close in how we report to you. A declining traffic number, presented on its own, is alarming and not actionable. A declining traffic number presented alongside a stable or improving conversion rate, a growing branded search volume, and a clear view of where you’re appearing in AI platforms is a completely different and much more useful conversation.
So What Should You Actually Be Looking At?
Pulling all of this together, here is what we believe should replace “traffic” as the headline metric in your reporting, and what we’re actively building into how we report to our clients:
- Conversions and revenue, not sessions. The number that matters is what happened after someone arrived, not how many people arrived. A smaller, higher-converting audience is a win, not a loss.
- Branded search volume and direct traffic. If more people are searching for your business by name, or typing your URL directly into a browser, that’s a strong signal that your visibility and brand-building work is paying off, even if it doesn’t show up as “organic” traffic in the traditional sense.
- Return visits and assisted conversions. Someone who discovers you via an AI platform, doesn’t click through immediately, but returns directly a few days later to make a purchase, represents exactly the kind of value that last-click attribution misses entirely.
- Presence in AI platforms and AI Overviews. Are you appearing when people ask ChatGPT, Claude, Gemini, or Perplexity about your category? Are you appearing in Google’s AI Overviews for your key terms? This is becoming as important as a traditional ranking, and arguably more important for high-intent categories.
- Citations and brand mentions beyond your own site. Are you being referenced by third-party sources – review platforms, industry publications, social platforms – that AI tools pull from when constructing an answer? This ties closely into the entity and brand visibility themes we covered in the hub post, and we’ll go much deeper on this in our next post in this series.
- Cost per acquisition, calculated holistically. Rather than looking at organic traffic volume in isolation, what matters is the total cost of acquiring a customer across every channel that contributed, including the AI-driven and zero-click pathways that don’t show up neatly in a single platform.
Conclusion
If your traffic has dipped this year, it doesn’t automatically mean something has gone wrong. It might mean exactly the opposite: that the channels feeding your business have shifted, that the visitors you are getting are more qualified, and that the old yardstick simply isn’t measuring the right thing anymore.
That said, a genuine decline in business outcomes – fewer leads, fewer sales, fewer enquiries – is a different conversation entirely, and one we’d always want to have directly and honestly with you. The goal of this post isn’t to wave away every concern with statistics. It’s to make sure that when we talk about performance, we’re talking about the metrics that actually reflect whether your business is growing.
We’re in the process of evolving our own reporting frameworks based directly on what we learnt at BrightonSEO this year, and we’ll be rolling these changes out across client accounts over the coming months. If you’d like to get ahead of that and have a conversation now about what your reporting should look like, get in touch with us – we’d love to talk it through.
Coming Up Next
In the next post in this series, we’ll tackle the question that naturally follows from this one: if traffic isn’t the goal, what is, and how do you actually get your brand mentioned by ChatGPT, Claude, and Google’s AI? We’ll cover the practical steps, the tools worth exploring, and the three levers – your website, third-party citations, and user-generated content – that determine whether AI platforms know who you are.
