The client's Search Console screenshot landed in my inbox at 7 AM on a Tuesday. Impressions holding steady. Clicks down 34% month over month. Ranking positions unchanged in the top three for seven commercial queries we had spent eight months building. None of it adding up on the surface, and all of it adding up perfectly once you knew what you were looking at.
That's what AI Mode does. It doesn't take your rankings away. It just makes them worth less than they were.
Google announced one billion users on AI Mode in April 2026. One billion people asking questions and getting answers inside the page: summaries, comparisons, recommendations, with no particular reason to click anywhere. The SEO industry spent that week writing hot takes about what this means for traffic, for content, for agencies, for everyone. Most of what I read told me that people are afraid and aren't saying so clearly.
Here's what I actually know, from watching this unfold in real accounts, not in think pieces.
The businesses getting hit hardest aren't the ones with bad SEO. They're the ones with good, clean, well-structured content that answers questions directly. Informational pages, buying guides, FAQ content, all the stuff the industry told everyone to build for the last five years. Google's AI Mode is eating that content wholesale, surfacing the answer without surfacing the page. You did the right thing. The reward structure changed.
I've had this conversation twice in the last six weeks. Once with a US-based e-commerce client who built a hundred informational pages around their product category over two years. Traffic from those pages is down 41% since February. The pages still rank. The AI Mode panel answers the question, attributes it to nobody in particular, and the user closes the tab satisfied. Once with a local service business in Karachi that had invested heavily in local landing pages and FAQ-structured content. Same pattern, smaller scale, same explanation.
Image credit: Screenshot from "SEO in 2026: How I'd Rank in Google in the AI Era" by Ahrefs on YouTube (https://www.youtube.com/watch?v=tiW6xRYSXmM).
What I told both of them is the same thing: the advice wasn't wrong, the environment changed. That matters because blaming the strategy is the wrong lesson to take.
The harder conversation is the one the SEO industry isn't having clearly. Most of what gets sold as "AI SEO strategy" right now is surface-level repackaging. I've seen pitches for AI-optimised content restructuring that, when you look at what's actually being done, is old schema markup work with new vocabulary around it. There's nothing wrong with schema. But calling it AI Mode optimisation and charging 2026 prices for 2019 work is something different.
What actually looks like it matters, and my sample size here is still limited, is entity consolidation and cited presence. AI Mode citations aren't random. Google is pulling from sources that its systems have built strong entity associations around. Not domain authority in the old DA score sense. Actual semantic relevance, consistent entity signals across the web, structured content that maps onto how the model understands the topic. The businesses appearing in AI Mode citations I've looked at have one thing in common: Google knows exactly who they are and what they do, across multiple surfaces, and there is no ambiguity in that picture.
Building that kind of presence is slower and less legible than what people are used to buying. There's no rank tracker widget for it. There's no deliverable you can send at the end of the month with green arrows. That's why it's not what most agencies are selling, even though it's what the signal environment is now rewarding.
I tried the conventional answer first with the e-commerce client. Refined their content further, tightened their topical clusters, improved their page experience scores. Three months of clean execution. AI Mode citations up slightly. Organic traffic flat. The traffic that came back came from transactional queries: product-level, high-intent, too specific for AI Mode to answer satisfactorily. The informational traffic didn't return. That's not a recoverable situation through content optimisation alone. That traffic has moved somewhere else in the SERP, or it's gone.
Adjusting the strategy meant accepting that first, which took longer than the actual adjustment.
The SEO industry is not going to figure this out in 2026. The businesses that come out of this in decent shape are the ones who stop treating their website as the center of their digital presence and start treating it as one node in a wider entity architecture that Google can understand clearly. The ones who keep trying to win the old game harder are going to be confused for a long time about why nothing is working, because their metrics will look fine right until they don't.
That Tuesday morning call? The client asked me what the number means. I told them it means we have to change what we're building, and that anyone who already knows exactly how to build it probably learned it from a case study that doesn't exist yet.
Waleed Qamar holds a BSc in Computer Science from Purdue University and has spent the years since turning that technical foundation into something the curriculum never covered: figuring out why websites rank, why they fall, and why most businesses never find out until it is too late.
Pakistan-born and based between the United States and South Asia, he has managed search visibility for e-commerce stores, local service businesses, and SaaS startups across two continents. He started in SEO when guest posting still worked, survived the Penguin update, and has rebuilt client sites from scratch after algorithm hits more than once.
He has watched good businesses get sold packages that looked like progress and delivered nothing lasting. He has also seen the right approach quietly double a site’s traffic without a single press release about it.
His writing on SEO By Highsoftware99 covers Google algorithm updates, autocomplete optimization, semantic SEO structure, and the widening gap between what agencies promise and what Google actually rewards in 2026.
He knows what a traffic cliff looks like in Search Console on the morning you discover it.
