A well-established US East Coast real estate firm lost 92% of its traffic in 14 months. Traditional SEO was not enough to get it back. Here is the full story: the hidden technical failure, the recovery, and how Generative Engine Optimization pushed the site past its own historical peak.
How Did a Real Estate Site Lose 92% of Its Traffic?
In early 2024, a well-known real estate business started losing organic search traffic, and a manageable dip turned into a sustained collapse. The company’s existing SEO provider blamed the March 2024 Helpful Content Update, which was a reasonable first guess since hundreds of thousands of sites were hit by that change.
The Helpful Content Update explanation only held up for a few weeks. Traffic kept dropping for more than a year, and by mid-2025 the site had lost over 92% of its monthly visitors. The business had to lay off staff just to keep operating. In June 2025, they brought us in to run an independent audit.
Why does a website keep losing traffic after a Google core update, even once the update has passed? Often the update exposes a deeper technical issue that was hidden by the site’s existing authority, and that issue stays until it is fixed directly.
What Was the Real Cause of the Traffic Loss?
The real cause was a property-listing widget that was completely invisible to Googlebot. We found it within thirty minutes of the audit. The widget loaded results dynamically based on user-selected filters, and it worked perfectly for human visitors. For Google’s crawler, it showed nothing.
This is a common and serious pattern in real estate and e-commerce: JavaScript-rendered content that loads after the initial page HTML. Search engine crawlers often do not run that JavaScript, or do not wait long enough for it to resolve. So Google was crawling and indexing pages that looked like they had no listings, no content, and no value, which is exactly the behavior the Helpful Content Update penalizes.
Beyond the invisible widget, the audit found a cluster of compounding problems. Contradictory structured data that sent conflicting signals to crawlers. Bad canonical tags pointing pages at incorrect versions of themselves. Crawl inefficiencies that pushed Googlebot toward low-value pages instead of high-value property listings.
None of these were as severe as the invisible content issue, but together they made the damage worse. When a site’s most valuable content is missing from the index and its metadata is contradictory, Google has no reliable signal to rank it on, even for branded queries.
How Did Technical SEO Recovery Bring Traffic Back?
The technical SEO recovery brought traffic back fast because Google re-indexes quickly once it finds content where there was none. We scoped the work at under one month and delivered on that estimate. The fix meant rendering the property widget server-side (or providing a crawlable static fallback), correcting the schema conflicts, and restructuring canonical relationships across thousands of listing pages.
Traffic grew 410% in a single month, from 85,000 to over 350,000 monthly visitors. For context, a 400% increase on a site with 100 monthly visitors is a very different achievement than the same jump on a site with 85,000. This was a meaningful, commercially significant recovery for a real business.
Growth continued over the next few months, then began to slow. By December the site had plateaued near 500,000 monthly visitors and stalled. At that point every technical issue was resolved and the site was healthy, but it was still at only half its original traffic, and it was not moving.
Technical SEO fixes recover lost ground, but they do not create a competitive advantage. Once a site is healthy and competing on equal footing with rivals, Google needs a reason to prefer it. Without fresh authority signals like links, citations, brand mentions, and content quality improvements, growth stalls. This is when the strategy has to evolve.
What Is GEO, and Why Did We Recommend It?
Generative Engine Optimization (GEO), also called AEO (Answer Engine Optimization) or AI SEO, is the practice of optimizing content and brand signals so they appear in the outputs of AI-powered search systems like ChatGPT, Gemini, Perplexity, Claude, and Microsoft Copilot. Traditional SEO targets ranked links on a results page. GEO targets citations, brand mentions, and sourced answers inside AI-generated responses.
AI models do not rank URLs. They synthesize information from sources they have indexed, training data, and real-time retrieval. To appear in those outputs, a brand needs to be present across a wide surface area of the web, not just highly ranked on a few target keywords.
We recommended GEO to this client because of the plateau. Every competitor in their market segment was roughly equal from a traditional SEO standpoint, so there was no easy technical or on-page win left. The only way to gain ground was to build the kind of broad, authoritative presence that AI systems recognize and cite, which also reinforces traditional ranking signals.
What Did the GEO Strategy Actually Involve?
The GEO strategy combined custom AI tooling with hands-on human execution across five layers. We built a machine learning system in Python, using the Claude API and OpenAI API together, to process the client’s entire website, competitor content, and existing market signals. The output was a content and signal plan based on the gaps we found in AI model coverage. The infrastructure was necessary, but the results came from the work in these five layers.
The authority layer leaned heavily on our authority link acquisition work, including the .edu domains, because those signals help with both AI retrieval and traditional ranking at the same time.
How Quickly Does GEO Produce Results, Platform by Platform?
AI platforms do not all pick up new signals at the same speed. Here is what we observed across the major platforms after deployment.
Gemini, ChatGPT, and Perplexity picked up the pages and brand within roughly a week of deployment. Claude and Microsoft Copilot followed at about two weeks. Once included, the citations were consistent rather than sporadic, which tells us the brand had crossed a relevance threshold in those models’ retrieval systems.
Citations showed up both as linked references and as unlinked brand mentions, and the unlinked ones matter for GEO. AI systems reference brand authority whether or not a hyperlink is present, and consistent brand mentions across sources feed the entity recognition that drives future inclusion.
How Did GEO Affect Both AI Traffic and Traditional SEO?
The GEO campaign produced two growth curves running in parallel: steady traditional SEO gains and a faster, more volatile AI-driven channel.
The wide range in GEO growth, 30 to 80% month over month, reflects how volatile an early-stage channel is. AI search is growing and the variance is real: some months a new platform update drove a spike, others were more linear. The direction, though, was consistent and upward across the whole campaign.
The more interesting finding was the GEO halo effect on traditional SEO. The same signals that improved AI platform presence, such as authoritative backlinks, high-quality content syndication, and consistent brand mentions, also reinforced traditional Google ranking factors. That is why GEO does not just build AI search traffic. It builds the kind of web authority Google’s ranking algorithms have always rewarded.
Yes, consistently. The activities that make a brand visible to AI retrieval systems, like broad authoritative citations, high-quality external links, and consistent entity mentions, overlap heavily with traditional off-page SEO signals. In this case, the GEO campaign contributed to the steady 8 to 10% monthly organic growth even before AI traffic was counted.
What Were the Final Numbers?
The client started the engagement at 85,000 monthly visitors, down from a historical peak of over one million. After technical SEO work and a GEO campaign running about two months (with work ongoing), the site now receives 1.3 million monthly visitors. That is 30% above its previous all-time high, with roughly 55,000 of those visitors arriving through AI-powered search channels.
The business that had been letting go of employees to survive is now growing past its own historical ceiling.
The terms are largely interchangeable in practice. GEO (Generative Engine Optimization) focuses on AI-generated search results. AEO (Answer Engine Optimization) emphasizes appearing in direct-answer formats. AI SEO is the broadest umbrella. The techniques apply across all three definitions.
In this case, 55,000 monthly visitors, roughly 4% of total traffic. AI search is growing fast but is still a secondary channel to traditional organic search for most industries. The strategic value right now is establishing presence early, while the cost of entry is still relatively low.
The property listings loaded via JavaScript after the initial HTML was served. Googlebot either did not run the JavaScript or did not wait for it to finish, so it indexed empty pages. The fix needs server-side rendering, dynamic rendering for bots, or a static HTML fallback that contains the core content.
The fundamentals, like structured content that answers specific question clusters, consistent brand citation building, and entity-level authority, work at many different scales. The AI infrastructure we built for this campaign was custom, but the strategic principles apply to any business creating content for AI-assisted search.