Myth Busting – GA’s View on Marketing’s Most Talked About Questions


GA Team | GA Agency GA Team | August 17, 2026| 8 minutes

Myth Busting – GA’s View on Marketing’s Most Talked About Questions

Every CMO has a version of this list already, even if it isn’t written down. It usually starts with a question from the board about the expected impact of AI, or a performance report showing two or three channels declining year over year with no clear explanation as to why.

However, with the constant noise and updates on SEO, GEO and Paid Media coming from all directions, finding the answers to these questions can feel like an overwhelming task.

With this in mind, we put some of the industry’s trending topics to our in-house experts to get their take.

Is SEO dead?

No, and the picture is more nuanced than either the AI-panic headlines or the “nothing has changed” reassurances suggest. According to Ahrefs’ Q1 2026 AI Search Benchmark Report, based on an analysis of 146 million search results, click-through rates to top-ranking pages fell 58% between December 2023 and December 2025, specifically where an AI Overview appeared.

That sounds unambiguous, but a separate dataset tells a different part of the story. Datos and SparkToro’s Q1 2026 State of Search report, built on real desktop clickstream data across the US, EU and UK, found zero-click searches actually declined over the same period, from 24.5% to 22.4% in the US and from 22.5% to 19.6% in the EU/UK, while organic click-through rose. AI tools’ overall share of desktop events also stayed under 2% in both regions as of March 2026.

Both are true at once. Where an AI Overview appears, it demonstrably takes clicks that would previously have gone to the top result. But AI Overviews and AI tools still touch a small enough share of total search activity that, in aggregate, click behaviour hasn’t collapsed the way the CTR headline alone implies. Google still sends roughly 190 times more traffic to websites than ChatGPT does, according to the same Ahrefs study.

Technical SEO, content structure and site authority remain the foundation that both traditional rankings and AI citation depend on. What’s changing is that being findable, being cited and being the brand an AI model already trusts are now three separate outcomes, and SEO alone only guarantees the first.

Is Performance Max the only Paid Media campaign worth running?

For some, Performance Max can feel like it’s dominating their Google Ads account structures and monopolising their media spend. To some extent that’s true, but that can be an intentional trade-off. Adoption of Google’s automated, feed-driven Performance Max format rose from 60% to 71% of surveyed advertisers between 2024 and 2025, an 18% year-on-year increase, according to a 2026 survey of 170-plus agency and in-house advertisers. Automation is now a standard part of managing Paid Media. The question is whether anyone is overseeing what it’s doing.

But dominance in account structure doesn’t mean it should be the only campaign type running, and Performance Max shouldn’t be treated as a replacement for all the other available options. Search, Demand Gen, Display and Video can still play different roles depending on the objective, audience and level of intent. The question isn’t which campaign type or platform wins universally; it’s whether the account structure and platform mix gives automation enough room to find incremental demand while retaining the controls needed to protect business priorities.

Performance Max will optimise towards the goals and signals it is given. The strategic question is whether those goals and signals reflect genuine business value, or simply what is easiest for the account to measure. Left unchecked, automated bidding will spend budget efficiently against the wrong goal. This is why “we’ve automated it” and “we’re in control of it” have stopped being the same sentence.

We perform well at home. Why doesn’t that translate abroad?

Strong performance in one market reflects choices made for that market’s content, media and technical setup. Those choices rarely carry over automatically to another market. A German or Japanese version of a page or a Paid Media campaign that performs well in English inherits none of that authority; it has to earn its own, in SEO, GEO and Paid Media alike.

The most common technical failure point on the technical side is hreflang, the signal that tells Google which language and regional version of a page to serve. Hreflang errors are a well-known problem and one of the most common ways a strong domestic SEO programme fails to travel. On the content and Paid Media side, the same principle holds: translated content and campaigns consistently underperform natively adapted ones, because tone, search behaviour and purchase triggers don’t translate word for word.

How many GEO strategies do we actually need?

One, applied market by market. That’s not a contradiction. The strategy itself needs to be singular, with governance, measurement and the quality bar centralised in one place. What changes locally is the execution: content, creative and citation-worthy phrasing built natively in each language, not translated from a master version.

GEO itself isn’t optional at this point. Ahrefs’ Q1 2026 report found AI Overview coverage varies notably by country and language, with Indonesia at 37% coverage against 21% in the US and Spanish making up 11% of AI Overviews by language. If that holds up, a brand’s AI-search exposure isn’t uniform across its footprint and a single English-language playbook applied everywhere will systematically misjudge where to invest.

This means that the same team that understands technical SEO and content structure in a given market also owns how that market shows up in AI-generated answers.

Why can’t we see clear ROI across our marketing spend anymore?

The measurement stack most brands still run assumes every meaningful touchpoint produces a trackable click. Increasingly, a customer reads an AI-generated answer, doesn’t click and converts three weeks later through a branded search or a direct visit. According to a 2026 industry study reported by eMarketer, roughly three in four US marketing leaders now say their core measurement approaches, including attribution and incrementality testing, are underperforming.

This is a measurement problem before it’s a performance one, and it goes beyond the top of the funnel. If Paid Media and Organic traffic both look softer than they should, it’s worth checking the post-click experience too. A conversion rate problem can look identical to an attribution problem on a dashboard.

Do I still need an agency?

Yes, but arguably for a different reason than five years ago. Anyone can now access AI tools for keyword research, ad copy or a first-draft content brief. What’s harder to replicate is the judgement that turns AI output into a business decision and the strategic oversight that keeps automated systems pointed at the right goal.

Successful client work has never depended on execution alone. Behind every high-performing campaign sits a combination of strategic thinking, technical expertise, and experienced consultants capable of turning information into business decisions.

The agencies that will succeed in the AI era are those treating it as an enhancement to their offering, not a replacement – freeing up their specialists from repetitive production work, so they spend more time on the parts of the job AI can’t do: interpreting what the data actually means for a client, catching the mistakes automation makes quietly and connecting execution back to a measurable business outcome.

Where to start

  1. Audit what your automation is actually optimising for, in Performance Max and in AI search alike, before deciding whether to trust or distrust the results.
  2. Separate your measurement problem from your performance problem. Before cutting a channel’s budget, check whether your attribution setup can even see the journeys that channel is driving.
  3. Treat each market as its own dataset, across SEO, GEO, Paid Media and content. A single global report will hide exactly the differences that matter most.

If you’re trying to work out whether your SEO, GEO, Paid Media and measurement foundations are ready for what AI has changed, that’s the conversation to have with us.

Sources

  • Ahrefs, Q1 2026 AI Search Benchmark Report, published 27 May 2026 (146 million SERPs, 730,000 AI responses)
  • Datos & SparkToro, State of Search Q1 2026 (desktop clickstream data, US/EU/UK, January 2025–March 2026)
  • Fluency, 2026 Trends and Performance Benchmarks for Google Ads and Meta Ads (survey of 170+ agency and in-house advertisers)
  • IAB study, reported via eMarketer, 2026 (US marketing leader measurement confidence)

Subscribe to our newsletter to stay updated

"*" indicates required fields


Related articles