Google traffic in 2026
In April 2026, Cyrus Shepard, founder of Zyppy SEO, published an analysis of more than 400 websites that had gained or lost traffic during Google’s most recent Core Updates. The results are clear: Successful websites share five measurable characteristics, and they have almost nothing in common with traditional SEO techniques.
What sets them apart is that they offer something that an AI-generated response cannot replace: a proprietary tool, a genuine service, exclusive data, or an active community. Generic informational content, on the other hand, is on the decline.
What the study actually says
Shepard measured each characteristic using Spearman’s correlation coefficient, which gives an indication of the strength of the link with earnings traffic on Google in 2026. Here are the five criteria, listed in descending order of correlation.
Offering a product or service (correlation: 0.391). 70.2% of winning websites sell their own product or service, compared with 34.6% of losing websites. This is the strongest correlation in the study. Websites that guide their visitors towards a specific action – whether that be a purchase, a booking or making contact – outperform those that simply provide information without offering any next steps.
Enable the user to perform a task (0.381). Calculators, configurators, quote forms, comparison tools: websites that encourage visitors to take action hold their attention for longer and capture behavioural signals that Google values.
Owning property assets (0.357). Exclusive data, original research, and a resource base built up over time. This type of content cannot be replicated by a competitor in a matter of hours, let alone synthesised by AI from existing sources.
Have a narrow thematic specialisation (0.250). Sites that cover a specific subject area in depth outperform those that spread themselves too thinly across dozens of related topics. Google seems to identify and value topical authority more effectively when it is concentrated.
Building a recognisable brand (0.206). 32.6% of winners in Q1-Q3 account for a significant proportion of branded traffic (navigational queries), compared with 16.1% of losers in Q1-Q3. When internet users search for your name directly, Google interprets this behaviour as a sign of trust.
Websites that meet four or five of these criteria see a traffic growth rate of between 68% and 70%. Those that meet none of them: 13.5%.
Why Google now rewards this type of website
The context explains the change. Since 2024, the volume of content published on the web has skyrocketed with the adoption of generative AI tools. The result is a web awash with generic, interchangeable texts that tell no one anything new. Google responded by tightening its quality criteria, notably through the Core Updates of March 2026 and May 2026, both of which penalised content with no real added value, regardless of its source.
What Google is looking to identify is what it calls E-E-A-T: experience, expertise, authority and trustworthiness. The SISTRIX analysis of the March 2026 Core Update is clear-cut: for every winner, there are fourteen losers. Established brands and market leaders are holding their ground or making gains. Authority trumps interchangeability.
In other words: Google no longer just evaluates your web pages. It evaluates your position within your market.
What this means for service providers or micro-businesses
A quick glance at this study might lead one to believe that it is aimed at major publishers or e-commerce platforms. That is a mistake.
A consultant or independent service provider naturally meets several of the key criteria: they offer a service (criterion 1), they guide clients towards a specific action such as booking an appointment (criterion 2), and they possess expertise that is difficult to generalise (criteria 3 and 4). What is often missing is criterion 5: the brand.
For a freelancer, building a brand online doesn’t rely on an advertising budget. It relies on consistent content published over time, visible customer testimonials, and a regular presence on specific topics. Every article that answers a real question from your potential customers, every Google review, every external mention, contributes to the brand signal that the algorithm records.
The practical lesson to be learnt from this study is not to start from scratch. It is to ensure that your website doesn’t look like a site that could belong to anyone. A website that could be yours and yours alone, with a clear service offering, a specific angle and visible evidence of what you do: that is precisely the profile Google will reward in 2026.
Sources
- LePtiDigital, «Study of 400 websites: the 5 characteristics of websites that are gaining Google traffic in 2026», May 2026
- Digital Century, «How some websites continue to attract traffic despite Google’s changes», May 2026
- SISTRIX, «Google Core Update March 2026: fourteen losers for one winner», April 2026
- SPACOMA, «AI and SEO Content in 2026», April 2026
Spearman’s correlation coefficient is a good starting point, but I would have liked to see the confidence intervals. A correlation of 0.206 across 400 sites is statistically significant, but it explains only a very small proportion of the variance. This doesn’t mean that building a brand is pointless; it means that we cannot draw any firm conclusions from this data alone.
What strikes me is criterion 2: enabling the user to carry out a task. We’re talking about calculators, configurators and forms. But how many micro-businesses have the resources to develop this sort of tool? In practice, this is a criterion that favours websites with a development budget.
I work at an SEO agency and we read this report as soon as it came out. What struck me most was the winners-to-losers ratio for the March Core Update: 14 losers to every 1 winner. That’s brutal. But in my experience, many of the sites that lost out had simply been churning out content en masse using AI tools, without any proofreading or a unique angle.
@Luc: Yes, and the problem is that these same sites are now going to start using the same methods all over again, hoping for a different result. The market for cheap SEO still has a bright future ahead of it.
I’m a self-employed therapist and I run a small WordPress website on my own. Reading this article, I feel that criteria 1, 3 and 4 apply to me quite well: I offer a specific service, I have a specialised focus (brief therapy), and my content is drawn from my actual practice. What’s clearly missing is a brand. My name doesn’t generate any direct searches. How do you build that when you’re working on your own?
@Isabelle: Google reviews are probably your quickest way to gain traction. Every review that mentions your name is an indirect brand signal. Next, maintain a regular presence on a single platform using your first name and your area of specialism. It takes time, but that’s what builds navigational traffic.
@Isabelle: I’d add listings on specialist directories (Doctolib, ThéraPsy, etc.), which are starting to build brand awareness. These aren’t direct brand searches, but Google recognises that your name appears alongside your line of work, which helps build thematic authority.
Thank you @Luc and @Karima for your practical feedback. I’ll start with Google reviews and perhaps sign up to one or two specialist directories. What I’ve taken away from this article is that my narrow thematic focus is an advantage, not a disadvantage. That was far from obvious to me.
I’ve read that bit about generic, interchangeable content three times. I run a personal finance blog and I know full well that 80% of my articles could have been written by anyone. I thought consistency was enough. Clearly not.
What the article says about narrow thematic specialisation is exactly what I’ve experienced. I used to have a general-interest blog about interior design, gardening and architecture. I stopped everything except interior design. The result: my traffic dropped for four months and then shot up again. It took Google a while to recognise my repositioning.
@Sébastien: Four months of losses is a long time to stick it out when you’re not sure of the outcome. Did you have any indicators that gave you confidence during that period, or were you just ploughing on blindly?
@Romain: I was using Ahrefs, which showed an increase in Domain Rating and a few rankings climbing for architecture-related search queries. That’s what gave me confidence. Without a tracking tool, I’d probably have given up.
What interests me about this study is that it goes some way towards debunking the ‘content is king’ myth. We’ve all spent years being told that all you need to do is publish regularly. This data shows that it’s not about how much you publish, but what you offer that nobody else does. That changes the nature of the work.
@Marie-Hélène: I’d qualify that a bit, though. Posting regularly is still important for freshness signals. But regularity without differentiation – yes, that’s clearly not enough since the 2024 Core Updates.
Question about proprietary assets: does data from your own customers (anonymised) count as a proprietary asset for the purposes of this study? I manage an HR SaaS platform and we have some interesting statistics on recruitment trends that we don’t publish because we don’t know how to capitalise on them.
@Florent: Absolutely. Data from your own database – even if it’s based on a limited sample – is exactly the sort of information that Google cannot synthesise from other sources. It might well be your best SEO tool if you present it properly. A quarterly report is enough to get started.
The point about the website being one that could only belong to you – that’s a phrase that really struck a chord with me. I took a look at my own website after reading the article and realised that, even with a new logo and a different name, nothing would actually change. It’s uncomfortable to admit.
@Victor: That’s the right test. And it applies to web writers too. If your writing could have come from anyone without anyone noticing, that’s a problem. It forced me to rethink how I present my work and which examples I highlight.
I’m 64 years old and have been running a small bookshop with an online shop since 2017. I read a lot about SEO without fully understanding it all. What I’ve taken away from this article is the importance of offering a genuine service and having a specialism. The bookshop specialises in regional Occitan literature. I think that fits criterion 4. That’s encouraging.
@Paul-Antoine: Occitan regional literature is a very specific area; it’s exactly the sort of specialism that holds its own against the big generalist publishers. If you’ve written your own book descriptions based on your knowledge of the subject – rather than copying them from publishers’ blurb – you probably have an advantage you’re not even aware of.