SEO in the AI era: metrics, budget, and priorities for 2027

Traffic is down: understanding if it's a real problem or a false alarm

In recent months, many companies have seen organic sessions decline without sales following the same trend. This decoupling isn't an Analytics error: it's the direct consequence of AI Overviews — the response blocks generated by Google directly on the results page — and response engines like Perplexity or ChatGPT Search, which keep users on the page before they reach your site.

The operational question is simple: was the missing traffic converting? If the answer is no, the drop is a signal to monitor, not an emergency to extinguish with additional budget. If the answer is yes, the problem is real and requires a change in strategy, not an increase in volume.

To understand how to measure this distinction, the starting point is to build a dashboard that separates sessions from assisted conversions. The article on AI visibility as a new metric that SMEs don't measure explains in detail what data to collect and how to read it.

Metrics to abandon and those to adopt

The State of Search 2027 by Search Engine Journal indicates a clear direction: stop optimizing for raw traffic volume and start measuring the quality of presence at the points where the user makes decisions.

In practice, the metrics that matter now are:

• Conversion rate by channel : how many organic sessions lead to a concrete action (request, purchase, contact);
• Citations in AI responses : how many times the brand appears in the answers generated by ChatGPT, Perplexity, Google AI Overview;
• Visibility on high-intent queries : positioning on searches where the user is ready to decide, not just to get informed;
• Direct and branded traffic : an indirect indicator of how recognizable the brand is even outside of classic organic results.

Those to stop using as primary KPIs: total sessions, page views, average position on generic keywords without commercial intent.

Where to allocate the SEO budget in the coming months

The allocation logic changes. Until 2024, it made sense to produce content in volume to capture informational traffic. Today, that traffic is absorbed by AI responses before reaching the site, and producing generic content is a measurable waste.

Investments that hold up in this scenario:

• Content with original data : internal research, real cases, proprietary data. It's the only type that AI engines cite as a source;
• Technical optimization for AI crawlers : site structure, semantic markup, content accessibility to spiders. On this point, it's worth reading how Cloudflare manages AI crawlers by category with Bot Preference Sync ;
• Local and vertical presence : for SMEs, visibility on geographic or specific sector queries is less attacked by AI Overviews compared to generic queries. The guide to local visibility in AI Search for Italian SMEs offers a practical path;
• Google Business Profile and product feeds : for those who sell products or have a physical location, these assets are still decisive. Operational guidelines are in the article on Google SERP, Business Profile, and product feeds .

What to stop doing right now

There are three habits that many companies maintain out of inertia and that consume budget without producing measurable results in this context:

• Produce generic informational articles on topics already covered by Wikipedia, vertical sites, or AI answers. No search engine will cite them, and no user will find them;
• Optimizing for keywords with high volume but low intent: they bring sessions that bounce without converting;
• Measuring SEO success only by ranking: a high position on a query covered by AI Overview brings very few clicks.

For those who manage content in a structured way, the content validation workflow for AI Search helps filter what's worth publishing before investing hours in content creation.

Three moves to make by the end of the month

Concretely, a marketing manager can do these three things without waiting for a complete strategic review:

• Segment organic traffic by intent : divide sessions into informational, navigational, and transactional, then compare conversion rates for each group. The result will immediately tell you where the drop really hurts;
• Check for brand presence in AI responses : search on ChatGPT and Perplexity for the questions customers usually ask. If the brand doesn't appear, the visibility problem is already happening;
• Review the editorial plan for the next quarter : eliminate generic topics, focus content production on proprietary data or angles that only the company can cover.

For those who want a complete picture of how organic search is evolving right now, the starting point is the section dedicated to SEO and AI visibility where we collect the most up-to-date analysis for the Italian market.