SEO KPIs in a World of Zero-Click SERPs

The Erosion of Traditional Success Metrics

With all of this being said, we can’t just abandon the traditional metrics. Traditional SEO KPIs: Keyword Position, Impressions, Clicks, CTR, and Conversions - remain foundational and are totally still relevant. Google processes around 14 billion searches every single day, whereas ChatGPT’s figure is at a fraction of that, at 700 million weekly active users as of mid-2025.

However, the reliability of these key performance indicators are becoming less and less relevant over time, as reliance on traditional search lessens, and mainstream adoption of LLMs increases.

The "Position One" Fallout

Previously, ranking in Position 1 was the gold standard. Today, with the integration of AI Overviews into Google SERPs, that metric can be misleading. A site ranking technically in Position 1 organically will be pushed down by an AI Overview, potentially even below the fold, resulting in zero clicks because the AI is answering the query in "Position Zero".

This shift most aggressively impacts Click-Through Rates (CTRs) for queries in the Awareness stage. Users seeking quick definitions or summaries get their needs met immediately by the interface. As such, we can no longer measure "success" strictly on clicks or sessions derived from these high-volume keywords.

The Behavioural Shift: From Search Engines to Answer Engines

The disruption isn't just in the way search results are being displayed; it’s also changing with the ways people search. In particular for problem-solving or more complex requests, customer behaviour is migrating toward AI tools and LLMs.

• The Rise of LLMs: Users are increasingly bypassing search engines to ask complex questions directly to LLMs (like ChatGPT or Google’s Gemini).

• Trust without Clicking: A user may never visit your website or “click” from a search engine for certain queries. Instead, they may simply trust the LLM’s guidance when it states, “X is the best solution for your needs”. (That said, at the time of writing, our SearchPulse Q4 survey shows 36% of respondents don’t trust results from AI tools).

• Social Discovery: Likewise, users are looking to social media platforms more frequently for inspiration, rather than trawling through ten blue links. (96% of SearchPulse Q4 respondents said they use social media for inspiration when booking their next holiday, for example)

In all of these above examples, a user may not go to a search engine at all to find your brand. As a result, if your attribution model only “starts” when a visitor lands on your website, you are missing a massive segment of the user journey in terms of brand visibility at the awareness stage.

A New Framework: The Behaviour-Led Search Mix

To adapt, we need to broaden our horizons and change how we measure success. This does not mean abandoning traditional SEO. Google still accounts for a massive percentage of global traffic and, due to familiarity bias, is likely to remain that way despite the influx of LLM tools.

However, reporting on SEO success needs to evolve to include all of the platforms where users are making their searches. With LLM use only increasing, we need to understand how we’re being perceived at the awareness stage, where zero clicks are most prevalent. 

1. LLM Tracking and Share of Model

We need to treat LLMs as a way of searching or another part of SEO. This means tracking:

• Visibility in LLMs: Or, in other words, are you appearing in the output when users ask relevant prompts in ChatGPT or Gemini?

• Sentiment Analysis: It’s not enough to just be mentioned; the AI needs to show your brand in a positive light, too.

2. Capturing “Dark” and Indirect Traffic

As clicks become less frequent in the awareness phase, the conversion action often shifts to direct entry later in the funnel. It’s more likely that the user will be more well educated before they land on your website, because of the amount of information they’ve consumed through AI results.

We therefore must also measure:

• Direct Traffic: Often a proxy for brand awareness generated by zero-click search exposures. If we can see Direct Traffic increasing, we can summarise that it’s likely to be due to better brand awareness (and therefore, better visibility in AI results)

• Referral Traffic: Visitors coming from AI citations or varied sources such as articles where we’re referenced.

• External Sources: Traffic from "Dark Social" channels like WhatsApp, where links are shared privately. Through Server Log analysis, this type of traffic can be tracked, too.

Conclusion: Measuring Everything Matters

The days of saying “Organic Search delivered 50 leads” are ending. We are moving toward a world where attribution is about the contribution of ALL your channels to ALL of your leads.

We must adopt a behaviour-led search approach that accounts for visibility in AI overviews, sentiment in LLMs, and the downstream impact on direct and referral traffic. Ultimately, every user’s journey is going to be complicated, and we’ll need to holistically track our performance across all channels in order to understand our marketing’s success.

A user journey might start with an idea from social media, then be researched with an LLM, be followed with using a search engine to narrow down their choices, and maybe finally be converted using Paid Media. We can’t ignore the social media or LLM parts of the journey, and as such, if you are not measuring success with a behaviour-led search approach, you are being left behind.

Download our latest SearchPulse report to see where search is heading, and if you're ready to look at your search strategy in the new world of AI Search, get in touch, and I'll team will be happy to help you take this to the next level.