By
Austin Coker
Reading time: 15 min
Your Search Console report looks fine at first glance. Impressions are holding steady, and some pages even rank higher than last year. Then you look at clicks, and they have been sliding for two quarters. That same week, your CEO forwards a ChatGPT screenshot that recommends three competitors for a query you have owned for years. The question is simple: why aren’t we in there?
That moment sits at the center of the GEO vs SEO question. SEO earns your pages a place in search results and the clicks that come from them. Generative engine optimization (GEO), often sold as “AI SEO,” earns your brand a place inside the answers that AI systems like Google AI Overviews, AI Mode and ChatGPT write for their users.
In this article, you’ll find a clear breakdown of what carries over from SEO, what is genuinely new, how to measure both, and a budgeting framework you can use for 2026 planning.
GEO builds on SEO. It doesn’t replace it. Google says its AI features run on the same ranking and quality systems as regular Search, so your SEO fundamentals still decide most of the outcome. What changes is the target. You now compete to be cited and recommended inside the answer, and you need to measure that separately.
SEO is about ranking a page. GEO is about being used by an answer.
With SEO, the user sees a list of results and picks one. Your job is to rank high enough and look relevant enough to win that click. The page is the product.
With GEO, the AI system reads several sources, builds an answer and then decides which sources to cite. The user may never see a list of results at all. Your job is to be one of the sources the system trusts enough to quote or recommend. The answer is the product, and your brand is an ingredient in it.
The scale is no longer a side story. At Google I/O in May 2026, Sundar Pichai said AI Overviews has more than 2.5 billion monthly users and AI Mode more than 1 billion (source: Business Insider’s I/O coverage). OpenAI says ChatGPT is used by more than 1.2 billion people each week (source: OpenAI’s GPT-6 announcement). Your buyers are asking these systems the same questions they used to type into a search box.
The way they ask has changed too. According to Google, the average AI Mode search is triple the length of a traditional Search query, and AI Mode queries have more than doubled every quarter since launch (source: Google’s AI Mode one-year insights). People describe their situation in full sentences. A buyer who used to search “inventory software” now asks for inventory software that works with Shopify, handles three warehouses and doesn’t need an implementation partner. That longer query is much closer to a purchase decision.
So the practical difference is this. SEO answers “where do we rank for this keyword?” GEO answers “when a buyer describes their problem to an AI, does our brand come up, and how is it described?”
More than most vendors selling AI SEO would like you to believe.
Google’s own guidance is direct. In Optimizing your website for generative AI features on Google Search, Google says SEO best practices remain relevant because its AI features are rooted in its core Search ranking and quality systems. It goes further: “From Google Search’s perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO.”
Here is what that means for your team.
Your pages still need to be indexed. Google says a page has to be indexed and eligible to show with a snippet before it can appear in AI features. If a page is blocked, noindexed or set to hide snippets, it is out of the running.
Technical health still matters. Google’s guide says all existing technical SEO best practices “continue to be worthwhile,” from crawlability to page experience. Slow, broken or poorly linked pages hurt you in both places.
Search intent still matters. If someone wants a comparison, give them a comparison. If they want pricing, give them pricing. Content that misreads intent rarely ranks, and it rarely gets cited.
Authority still matters. Brands that are referenced, reviewed and linked to across the web are easier for any system to trust.
Google also lists things you can ignore for its AI features. You don’t need an llms.txt file. You don’t need to chop your content into chunks or rewrite pages “for AI.” Structured data isn’t required, and there’s no special schema for AI answers. Chasing inauthentic mentions across the web isn’t as helpful as it might seem, in Google’s words.
If your SEO program is solid, you already have most of what GEO needs. If it is weak, no amount of AI-specific tactics will make up for it.
The click is no longer guaranteed, even when you rank.
Pew Research Center analyzed March 2025 browsing data from 900 US adults. When an AI summary appeared, users clicked a traditional result in 8% of visits. Without a summary, they clicked in 15% of visits. Users clicked a link inside the summary itself in just 1% of visits (source: Pew Research Center).
Ahrefs studied 300,000 keywords using December 2025 data. The presence of an AI Overview correlated with about 58% lower click-through rate for the page in position 1. Position 2 saw about 51% lower CTR, and position 10 about 19% lower (source: Ahrefs’ AI Overviews CTR update).
That is the scenario from the top of this article. Impressions hold because you still rank. Clicks fall because the answer sits above you. We cover this pattern in more depth in our guide to zero-click search.
Clicks haven’t disappeared, though. On Alphabet’s Q2 2026 earnings call, Pichai said Google is “sending billions of clicks to websites every week through AI features in Search” (source: Alphabet Q2 2026 earnings call). The traffic still exists. It goes to the sources the answer chooses to cite.
That shifts your work in four ways.
Being cited becomes a goal of its own. A page can rank fifth and still be the source an AI answer quotes. A page can rank first and get skipped.
Your brand has to exist off your own website. When an AI system recommends products or vendors, it draws on what the wider web says about you. Reviews, comparisons, industry coverage and forum discussions all shape whether you’re recommended and how you’re described.
Commodity content loses value. If your page says the same thing as twenty others, an AI system has no reason to cite you specifically. Google’s guidance puts “non-commodity content” with a unique point of view at the top of its list. Your own data, your own process and your own opinion give a system something worth quoting.
Your keyword list becomes a prompt list. You still research keywords. You also need to map the full questions buyers ask AI tools, especially the ones that signal intent to buy. Our GEO guide walks through how AI engines pick sources and what that means for your pages.
AI search engines retrieve first and write second.
Google explains that its AI features use retrieval-augmented generation, also called grounding. The model doesn’t answer purely from memory. It relies on core Search ranking systems to pull relevant pages from the index and builds its answer from them. That’s why being indexed and snippet-eligible is the entry ticket.
Google also uses query fan-out. One user question gets broken into several related searches behind the scenes. Google’s own example: “how to fix a lawn that’s full of weeds” might fan out into “best herbicides for lawns,” “remove weeds without chemicals” and “how to prevent weeds in lawn.” Pages that rank for those sub-questions have more chances to be pulled into the final answer.
This explains a pattern many marketers notice. A brand that covers a whole topic well tends to get cited more than a brand that ranks first for one head term. Depth across the cluster matters.
Research also points to what helps a page get used once it’s retrieved. The paper that named the field, “GEO: Generative Engine Optimization” by Aggarwal and colleagues at Princeton and IIT Delhi, presented at KDD 2024, tested how content changes affected visibility in generative answers. Adding citations, quotations and statistics boosted visibility by up to about 40%. Keyword stuffing, a classic SEO habit, didn’t help (source: the GEO paper on arXiv).
The most interesting finding was about lower-ranked pages. In the study, citing sources lifted the visibility of fifth-ranked pages by 115%, while first-ranked pages lost 30% on average. Ranking first doesn’t lock in your place in the answer. A well-sourced, specific page further down can take it.
In practice, the pages that get cited tend to share a few traits. They answer the question in the first lines of a section. They back claims with named, linked sources. They include something the reader can’t get elsewhere, like original numbers or a firsthand explanation of how something works. And they come from a brand the rest of the web already recognizes.
You measure SEO with the metrics you already trust. You measure GEO with a newer set that sits next to them.
For SEO, keep tracking rankings for your priority keywords, organic clicks, organic conversions and revenue from organic traffic. None of that goes away.
For GEO, you now have first-party data from Google. On June 3, 2026, Google announced a generative AI performance report in Search Console, and it rolled out to all websites worldwide on August 31, 2026 (source: Google’s announcement for website owners). It shows how often your URLs appear in AI Overviews, AI Mode and generative features in Discover, broken down by page, country, device and date. Google added a control at the same time to include or exclude your site from those features. Google says opting out won’t affect your rankings elsewhere in Search, but you lose all traffic and impressions from AI features, so treat it as a last resort.
Beyond Search Console, build a short set of GEO measures:
Start your prompt list with the questions closest to a purchase. Comparison prompts, “best tool for” prompts and pricing prompts tell you far more about revenue than broad educational questions. Get visible there first, then move up the funnel.
Then report both sets of metrics in one view. If organic clicks fall while AI impressions and branded demand rise, you aren’t simply losing. Your visibility is changing form, and your reporting should show it.
Budget for one search program with two sets of outcomes. Don’t create a separate GEO line item that competes with SEO for the same work.
Most of what drives GEO results is SEO work done well. The parts that are truly new, like prompt research, off-site brand building and AI visibility tracking, are real costs. They’re also smaller than many agencies pitching AI SEO packages suggest.
Here’s what the published data shows.
Conductor’s 2026 State of AEO/GEO report surveyed more than 250 digital leaders at enterprise companies. On average, those enterprises allocated 12% of their digital budgets to AEO in 2025, and 94% plan to increase their investment in 2026 (source: Conductor’s CMO investment report).
Digiday reported in March 2026 on a Scribewise survey of 205 US marketers. 55% said they have dollars set aside for GEO in their marketing budgets, and 70% said GEO accounted for 11% to 20% of that budget. In the same piece, Noise Media Group CEO Joseph Levi said a brand spending $10,000 a month on SEO would now expect at least half of that budget to also cover GEO (source: Digiday).
Read those numbers carefully. Conductor’s sample is enterprise. A company doing $1M to $50M in revenue has a different cost structure and less room for experiments. Levi’s point is the most useful one for most mid-size brands: the expectation is that existing SEO spend also covers GEO, not that the budget doubles.
Use these benchmarks to sanity-check your plan. Don’t copy them.
Answer these four questions before you set a number.
Your answers tell you where the money should go. They also tell you whether you need more of it.
Ecommerce brand. Your category pages rank, but AI answers now name specific products for buying questions in your category. Keep funding technical SEO and product page quality. Shift content spend from generic buying guides toward pages with your own testing, specifications and customer data. Add budget for earning reviews and coverage on the sites AI systems draw product recommendations from.
SaaS company. Your blog traffic is down, mostly on top-of-funnel posts. Your demo requests come from comparison and alternatives searches. Protect and expand those high-intent pages first. Track the prompts buyers use when they ask AI tools to compare software in your category. Fund original content, like product benchmarks or usage data, that gives AI systems a reason to cite you over a review aggregator.
B2B services firm. Your buyers ask AI for vendor shortlists, and you rarely appear. Your site is thin on proof. Fund detailed service pages that explain how you work, who you work with and what makes your approach different. Invest in expert-led content from your partners or founders. Put real budget into being mentioned in the industry publications and directories your buyers trust.
Reallocate first. Most teams can cover a meaningful part of GEO by cutting work that no longer pays off:
Add budget for the parts that are actually new:
If your foundation is healthy and your search revenue is stable, reallocation may cover most of it. If AI answers already shape your category and you’re absent, adding budget is the faster path back.
Splitting SEO and GEO into separate teams or vendors creates duplicate work and conflicting reports. The same page has to rank, get cited and convert. The same team should own it.
Running them as one program looks like this. You build one research process that maps keywords and buyer prompts together, starting with the ones closest to a purchase. You create content that ranks for the topic cluster and gives AI systems something original to quote. You build your brand off-site so you get recommended, not just indexed. And you report on revenue, with rankings, clicks, AI impressions and citations as the signals that explain it.
Paid search belongs in the same conversation. As AI answers take more space on the results page, the balance between organic, paid and AI visibility changes query by query. Managing them together lets you put money where the clicks still are.
At 95 Projects, we run SEO, PPC and GEO as one program for $1M to $50M ecommerce brands, SaaS companies and high-value B2B services firms. Our founder, Austin Coker, works on every account, so the person setting your strategy is the person doing the work. We measure everything against revenue, not traffic for its own sake.
That approach is how Buoy became the #1 ChatGPT recommendation for “best electrolytes with no sweeteners” in 3 weeks. You can learn more about our AI search services and SEO services, or book a call to see where your search budget should go in 2026.
Austin Coker is the CEO and founder of 95 Projects. He has worked in SEO since 2020, when he built and sold an ecommerce brand that grew mainly through search. He founded 95 Projects in 2022 to help SaaS, ecommerce, and B2B service companies show up wherever their buyers search, from Google to ChatGPT and AI Overviews.
Have questions about working with us? Book a 30-minute strategy call to discuss your goals and see if we’re a good fit.
No-pressure conversation
Clear next steps (if we’re a fit)
Talk directly to an expert
No. GEO depends on SEO. Google says its AI features are rooted in the same ranking and quality systems as regular Search. If your pages aren’t indexed or don’t rank for the topic, they’re very unlikely to be cited in AI answers.
SEO aims to rank your pages and earn clicks, while GEO aims to get your brand cited and recommended inside AI-generated answers. The underlying work overlaps heavily. The difference is in what you target and how you measure success.
Mostly, yes. AI SEO and GEO are usually used to describe the same work: improving your visibility in AI-powered search experiences like Google AI Overviews, AI Mode and ChatGPT. Google itself frames optimizing for generative AI search as still being SEO.
Not for Google. Google’s guidance says Google Search ignores llms.txt files, and structured data isn’t required to appear in its AI features. Your page does need to be indexed and eligible to show with a snippet.
Use the generative AI performance report in Google Search Console. It shows how often your URLs appear in AI features, broken down by page, country, device and date. Pair it with regular checks of your most important buying prompts to see how your brand is described.
It depends on how much revenue comes from search and how exposed your key queries are to AI answers. In a Scribewise survey reported by Digiday, 70% of marketers said GEO accounted for 11% to 20% of their budget, and Conductor found enterprises put an average of 12% of digital budgets into AEO in 2025. Many teams cover a large part of it by reallocating spend from low-performing content.
No. Google says the Search Console control isn’t used as a ranking signal outside its generative AI features. You do give up all impressions and traffic from AI Overviews and AI Mode, which is where a growing share of searches now happen.
