We want to be clear about something before we write this: we are not against Google.
Google is one of the most important companies in the history of technology. They organized the internet. They made the world’s information searchable. Gmail, Google Docs, Android, YouTube, Google Maps — these are products that billions of people use every day, and they’re genuinely good. We use them too.
This isn’t an anti-Google post. This is an honest post about why almost nobody leaves Google Search, even when the deal stops working — and why we think that’s about to change.
It’s fear
Pure and simple.
Google has spent 25 years making the entire internet dependent on them. Businesses built their whole acquisition strategy around Google rankings. Entire industries exist to optimize for Google’s algorithm. "If we’re not on Google, we don’t exist" has been gospel since 2001.
And here’s the thing: for most businesses, it’s still true.
A local plumber in Queens blocks Google? They’re done. No calls, no jobs, no business. A media company deindexes from Google? They lose 40-60% of their traffic overnight. An e-commerce store disappears from Google Shopping? Revenue craters.
The dependency is real. It’s not imagined. It’s not exaggerated. For the vast majority of businesses on the internet, Google is the single largest source of customers. Leaving Google is like a retail store leaving the only mall in town. It doesn’t matter if the mall is charging unfair rent — the foot traffic is there and it’s not anywhere else.
We get it. We’re not being cavalier about this. The fear is rational.
But the deal is changing
Here’s what’s shifting, and it’s shifting fast.
For two decades, the deal between Google and the internet was simple: you create content, Google indexes it, users search for it, Google sends them to your site. You get traffic. Google gets ad revenue from the search page. Both sides win.
That deal is breaking.
AI Overviews changed the equation. When someone searches "best bars in SoHo" and Google generates an AI summary at the top of the page using data from your site — complete with the venue names, the descriptions, the ratings — the user reads the summary and never clicks through. Your content trained Google’s answer. Google kept the user. You got nothing.
The data is brutal. Publisher referral traffic from Google has dropped 25% since AI Overviews rolled out. Sixty-nine percent of AI Overview answers contain word-for-word copies of publisher content. Google is literally displaying your words as its own answer and keeping the ad revenue.
This isn’t a partnership anymore. This is Google saying: "Thanks for the content. We’ll take it from here. The user doesn’t need to visit you."
And yet almost nobody is leaving.
Google can’t tell good data from old data
Here’s something that made this decision personal for us.
We built 28,000+ venue profiles with GPS-verified coordinates, current opening hours, phone numbers, structured Schema.org markup, vibe tags, insider tips, signature items, menus, and events. Every venue typed correctly — Restaurant, BarOrPub, CafeOrCoffeeShop, EntertainmentBusiness — with AggregateRating, GeoCoordinates, OpeningHoursSpecification, and FAQPage schemas. Deep, structured, verified data across 43 Manhattan neighborhoods.
We offered that to Google. Not in a meeting or a pitch — we put it on the internet, structured exactly the way Google’s own documentation says to structure it, and waited for the algorithm to do its job.
The algorithm’s job, apparently, was to rank us behind a Yelp page with 12 reviews from 2021 and a 3.5-star average from users who may or may not have actually visited the restaurant.
Google’s ranking system can’t evaluate data quality. It evaluates authority signals — backlinks, domain age, engagement history. A site that’s been online for 15 years with thousands of referring domains will always outrank a site with better data that launched six months ago. That’s not a meritocracy. That’s an incumbency protection system.
And Google’s own venue data? Google Maps listings are notoriously stale. Wrong hours. Restaurants that closed months ago still showing as open. "Popular times" based on aggregated phone location data, not actual visits. No vibe context. No insider tips. No mood matching. Just a star rating, a pin on a map, and a pile of reviews that range from genuine to fake to vengeful ex-employees.
We were offering Google a chance to present its users with data that actually matters in 2026 — structured, verified, current, and deep. They continued to value 5-year-old domain authority metrics and their own stale map pins over genuinely useful structured data.
That’s when we realized: Google’s algorithm isn’t optimizing for the best answer anymore. It’s optimizing for the safest answer. The answer from the domain it already trusts, regardless of whether that domain’s data is current, accurate, or useful.
AI models don’t work this way. AI models evaluate the data itself — its structure, its completeness, its specificity. A well-formed Schema.org entity with GPS coordinates, current hours, vibe tags, and an insider tip is more useful to an AI than a page with 10,000 backlinks and a paragraph of boilerplate. AI rewards data quality. Google rewards tenure.
We chose the system that rewards what we’re actually good at.
The abusive relationship dynamic
We’re going to be blunt here because we think it needs to be said.
What’s happening between publishers and Google has all the hallmarks of a relationship where one side has all the power and the other side keeps hoping it’ll get better.
"Maybe the next algorithm update will help us."
"Maybe if we optimize harder, we’ll get our traffic back."
"Maybe AI Overviews will go away."
"We can’t survive without them."
Publishers see the data. They know their traffic is declining. They know their content is being copied into AI summaries without meaningful attribution. They know the trend is accelerating, not reversing. The CMA literally had to step in and force Google to offer an opt-out mechanism because Google wasn’t going to do it voluntarily.
And still — two-thirds of publishers say they won’t opt out, even when they can. Because the fear of losing Google traffic is greater than the pain of watching Google cannibalize their content.
That’s not a business strategy. That’s dependency.
Why we could leave (and most can’t)
We want to be honest about our situation. We had advantages that most businesses don’t.
We weren’t deeply dependent. Moodap launched in early 2026. We had a few months of Google traffic — a nice honeymoon burst that peaked and then declined, exactly as the algorithm is designed to do with new sites. When we left Google, we were walking away from ~40 daily users and a declining trend line. That’s not the same as a publisher walking away from 40% of their revenue.
Our Google traffic was already dying. The honeymoon was over. Google had recalibrated us behind established players with 10-20 years of domain authority. We were competing with Yelp, TripAdvisor, Eater, and TimeOut for restaurant and bar queries — sites with millions of backlinks and decades of SEO momentum. Our GPS-verified venue profiles with structured data and insider tips were ranking behind Yelp pages with 12 reviews from 2021. The algorithm was never going to favor us, no matter how good our data was.
The cost of staying was compounding. Every day we stayed on Google, our structured data was feeding AI Overviews that answered queries about NYC venues without sending users to us. We were training our competitor’s AI for free. The longer we stayed, the more data Google had, and the less reason any user would have to click through to the source.
We had an alternative. This is the big one. If there were no AI search platforms, leaving Google would be leaving discovery entirely. But ChatGPT, Perplexity, Claude, Bing, DuckDuckGo, and Yahoo exist. They’re growing. They index our content. Some of them cite their sources. We weren’t jumping off a cliff — we were jumping to a different platform.
Most businesses don’t have all four of these conditions. That’s why they stay.
Who should consider leaving
Not everyone. Probably not most businesses. But some.
If you’re a data-rich platform — local discovery, real estate listings, recipe databases, event platforms, product catalogs — your structured data is more valuable to AI than your Google traffic is worth. AI models are built to reason over structured entities. Your Schema.org markup is their native language. When AI uses your data and cites you, that’s a better deal than Google using your data and keeping the user.
If you’re a new platform with zero domain authority — you’re never going to outrank the incumbents on Google anyway. You could spend years building backlinks and creating "SEO content" to climb from page 4 to page 2, or you could invest that same energy in structured data that AI models actually use. The ROI calculation favors AI.
If you’re a niche vertical — AI models are hungry for deep, specialized datasets. A comprehensive database of 28,000 NYC venues with vibe tags and insider tips is exactly the kind of data that makes AI recommendations useful. Broad, shallow content doesn’t help AI much. Deep, structured, domain-specific data is gold.
If you’re a startup that values independence — building on Google means building on a platform that can change the rules anytime. Every algorithm update is a business risk you don’t control. AI platforms are newer, more fragmented, and more competitive with each other. That fragmentation means no single platform has the leverage to crush you.
Who should stay
We’re not evangelists. We’re pragmatists. Some businesses should absolutely stay on Google:
Local service businesses. The plumber, the dentist, the dry cleaner. Google is how their customers find them. Google Business Profile and Google Maps are the discovery layer. AI isn’t replacing that yet. Maybe someday, but not today.
E-commerce stores with established traffic. If 50% of your revenue comes from Google Shopping and organic search, leaving is a bet-the-company decision. You can diversify toward AI discovery without abandoning Google. That’s the smart play.
Content publishers (for now). If you’re a news site or media company, you’re in a genuinely difficult position. Your content is being copied into AI Overviews, but Google still sends more traffic than any other single source. The CMA opt-out might help. Separate crawlers might help. But leaving entirely? That’s a hard call that depends on your specific traffic profile.
Anyone who depends on Google for the majority of their customers. Dependency is real. If Google is your primary channel, the right move is to diversify, not to make a dramatic exit. Add structured data. Optimize for Bing. Publish llms.txt. Welcome AI crawlers. Build alternatives. Then, when the alternatives are generating meaningful traffic, reassess the Google relationship.
The wave is coming
We think what we did will look obvious in hindsight.
Not because we’re visionaries. Because the conditions that made it rational for us are spreading. AI search volume is growing exponentially. AI platforms are getting better at citing sources. Bing’s role as the backbone of AI search is becoming more important. Google’s AI Overviews are reducing click-through rates further every quarter.
The gap between "what Google gives you" and "what Google takes from you" is widening. At some point, for enough businesses, the math will flip. The traffic Google sends won’t be worth the data Google takes.
When that happens, the exodus will start. Not all at once. Not dramatically. But steadily. Data-rich platforms first. Then niche verticals. Then new startups that never bother with Google in the first place. Then, eventually, some of the established players who realize their domain authority isn’t protecting them from AI Overviews eating their click-through rates.
We’re early. We know that. Being early means being alone for a while. It means explaining your decision in blog posts because nobody has a playbook for this yet.
But early is where the upside is. And the upside of being the platform that AI trusts for NYC venue data — because we gave them our best data when nobody else would — is worth more than page 3 on a search engine that’s cannibalizing our content.
The bottom line
Google isn’t evil. Google is a business that’s optimizing for its own interests, just like every other business. The problem is that their interests and their partners’ interests are diverging. AI Overviews are good for Google’s engagement metrics and ad revenue. They’re bad for the publishers and platforms that provide the underlying data.
Nobody is leaving because they’re scared. The fear is rational. The dependency is real. We’re not judging anyone for staying.
But we’d encourage every business — especially data-rich platforms, startups, and niche verticals — to at least do the math. Look at your Google traffic trends. Look at how much of your content is showing up in AI Overviews. Calculate what you’re giving Google versus what you’re getting back. Then look at what AI platforms could offer you if you invested the same energy in structured data and AI crawler access.
The numbers might surprise you.
They surprised us.
— The Moodap™ Team