[Guest post] The Great Google Search Migration
How large language model tools are 'crossing the chasm' and taking search market share from Google: guest post by Paola Santiago
Dear subscribers: This article is a guest post by exploring how AI-based tools are changing search behavior and business models, and identifying ethical concerns around these search alternatives. Enjoy! (If your email client truncates this article, see sixpeas.substack.com/p/the-great-google-search-migration-paola-santiago online to read the whole post.)
The Great Google Migration: How AI Is Redefining Search Behaviour
Google Search has long reigned as the world’s top search engine, but a quiet migration is underway. Users are slowly drifting away and turning to AI-driven tools like ChatGPT, Perplexity, and even Bing. These platforms offer something Google increasingly struggles to deliver: instant, conversational answers without the clutter of ads or the grind of clicking through irrelevant links.
In my previous article, I explored how alternatives like ChatGPT are quietly replacing Google for many users:
This time, let’s zoom in on AI’s growing role as a major disruptor of traditional search, and how Google is trying to hold onto its crown with Gemini, its AI-powered response.
So why are people making the switch? Part of it is practical: AI tools feel faster, more direct, and more personal. But there’s also something deeper: what feels like a generational shift, or maybe a quiet rebellion against Big Tech fatigue.
Search expectations are changing. The question now is: are these AI alternatives ready to take Google’s place and redefine how we find information? Or are they still missing a key ingredient?
Is Google Really Losing?
There’s no denying that Google has been the dominant Search Engine with more than 90% of the global market since 2015. But in 2025, it dipped below that threshold for the first time in a decade. While the drop may seem small at first glance, it signals a potential crack in Google’s long-standing dominance.
So why are people moving away from Google?
Is this just a novelty internet moment? Or are we witnessing the early signs of a broader migration: one that places these early migrators at the forefront of the “early market” in the Crossing the Chasm model?
Let’s start by looking at the generation leading this quiet shift. Gen Z is increasingly favoring alternatives, with 60% of them prefer searching on social media over Google [study]. In my previous video, I asked them directly: why did you stop using Google?
The responses, though anecdotal, were revealing: from “Google is a bad company” to “Bing already integrates with ChatGPT”. These weren’t just offhand remarks. They reflected deeper frustrations and a noticeable shift in search behavior. Responses fell into 3 main categories:
Ad-Filled Results Cause Growing Frustrations in Users
According to Ikius, Google’s user experience is increasingly bogged down by intrusive ads and sponsored content that dominate the top results. This isn’t surprising when 77.8% of Google’s total revenue comes from advertising. The result? Users are frustrated having to scroll past irrelevant or clickbait content just to find what they’re looking for.
Decline in Search Quality Results
Beyond ads, search result quality is also deteriorating. SEO-optimizated sites are prioritized and even AI-generated content is gaming the system, pushing low-quality websites to the top. Even Google admits it’s struggling to combat these trends. A recent poll of 2,400 people showed that 70% now append “Reddit” to their searches just to get more authentic human answers.
Privacy & Ethical Concerns
Google’s reputation has also been eroding. Through Project Nimbus, Google has reportedly provided AI and cloud services to the Israeli government, raising serious concerns over surveillance and potential human rights violations. Add to that a $5 billion class-action lawsuit for allegedly tracking users in incognito mode.
People aren’t just annoyed by the ads anymore. They feel manipulated. The top results on Google usually come from whoever can game the system with SEO or pay to show up first. And when you add the decline in answer quality, it’s just frustrating. Especially for people who are in a rush and don’t want to scroll past five irrelevant links or click through multiple pages to get what they need.
The Google Search experience is no longer about surfacing the best answers. It’s about people feeling like they’re being funneled into content farms, affiliate links, or AI-generated fluff, rather than actually getting value.
On top of that, a small fraction of users are silently protesting.They’re moving toward platforms with clearer ethics, like DuckDuckGo. In 2012, it processed around 16 million searches per year. By 2024, that number ballooned to 36 billion. That’s over 2,000x growth in just over a decade. This rise might seem small compared to Google’s scale, but the growth tells a story.
With all of these reasons, it’s no wonder there are so many articles right now talking about the Death of Google. What we’re seeing is a slow but meaningful behaviour change. Most people don’t switch platforms until it directly affects their experience, but that time is arriving.
More and more, people are jumping between platforms to get answers. They are seamlessly transitioning between Google, AI alternatives like ChatGPT, and social media such as Reddit and TikTok. It just depends on what gets them what they need fastest in the most convenient way. And that tells you everything. Google isn’t the default anymore. It’s becoming just one of the options among many.
The AI Disruptors: ChatGPT, Bing, and Perplexity
Let’s turn our attention to one of the biggest forces reshaping how we search: AI. While tools like ChatGPT still face limitations, like occasional hallucinations, they’re becoming viable alternatives to traditional search engines.
According to a survey by Future Publishing, 27% of respondents in the U.S. said they now use AI tools like ChatGPT instead of Google. Their reasoning often aligns with how these AI alternatives are faster, easier to use, and more personalized to what they’re looking for. A very optimistic projection is that over 90 million American users will be using genAI as their primary search tool.
So, who’s leading this AI-powered charge? While this article focuses on ChatGPT, Bing, and Perplexity, it’s worth noting there are other fast-emerging AI tools entering the space. For example, Claude (by Anthropic), Grok, and DeepSeek each offer different strengths in how they approach information retrieval and user interaction.
1. ChatGPT (OpenAI)
The most popular generative AI tool that has come to the market. ChatGPT basically disrupted every industry when it came out to the public in 2022. And it continues to rule today’s numbers with 3.8 billion visits a month. It’s known for its conversational tone that breaks down complex user questions into digestible, direct, and fast answers.
Notably, it’s superior to Google when it comes to any content gap analysis and asking follow-up questions. This basically eliminates a huge chunk of time where users don’t have to keep going back to square one, like one would on a Google results page. It remembers the context and builds on it, which makes it way more efficient for content gap analysis and deeper research. The interface also avoids all the SEO-heavy, ad-driven clutter. It doesn’t run ads, so users are not being pushed toward sponsored content or junky listicles.
Moreover, ChatGPT’s versatility outcompetes Google when it comes to what it can offer its users, especially assisting them with writing, coding, brainstorming, and summarizing information, making it a multi-functional productivity tool. And once that is done, there’s just a lot more personalisation, where users are able to instruct the system with how they want it to respond to them.
Limitation: The biggest limitation of ChatGPT is that it was never really meant to replace Google. Its rise as a competitor is more of a byproduct of how frustrated people became with traditional search and how natural it felt to use ChatGPT instead.
At its core, ChatGPT is built for content creation and idea generation. Unlike Google, it doesn’t have a real-time web crawler that indexes the entire internet. So while it can now retrieve some live information, especially for current events or financial updates, it still can’t match the scale and speed of traditional search engines like Google or Bing. Those platforms rely on massive, constantly-updated indexes of billions of web pages, giving them a broader and deeper range of results.
And with content creation as its biggest strength, and sources as its worst, I can’t help but think how prone it is to hallucinations. Citing the JMIR scientific article, they compared different AI bots and placed ChatGPT as one of the highest reference hallucination scores, with critical levels of hallucination in medical citations.
What does this mean? Right now, it simply means that ChatGPT cannot fully replace traditional search engines. There are way too many limitations that bring more detriment when it comes for people getting accurate and stable answers, especially when it comes to specialized topics.
However, let’s just dissect a statement from OpenAI’s CPO, Kevin Weil (in
by ):“The AI Models that you’re using today is the worst AI model you will ever use for the |rest of your life."
My take: The development of AI has been greatly accelerated. It feels like we’re watching a high-speed train we didn’t build, and can’t slow down, but are still somehow expected to ride.
In 2023, ChatGPT supported just over 50 languages. By 2024, it was generating images, responding to voice, and assisting with research. And in early 2025, we’ve seen trends where OpenAI has yet again found itself in hot waters by allowing users to alter images with Studio Ghibli style.
But when it comes to ChatGPT replacing Google Search, it’s not entirely out of the question, especially if OpenAI decides to pivot more intentionally in that direction.
Right now, though, OpenAI isn’t trying to replace the search engine. It’s becoming part of the search process instead. ChatGPT still depends on traditional search engines in the background to fill in the gaps it can’t cover on its own. In that sense, it’s more of a layer on top of search, not a replacement. At least, not yet.
Ethics: OpenAI is no stranger when it comes to scrutiny and criticisms regarding how the company handles its data practices, training on intellectual property, and transparency. The company collects extensive user data, including prompts, responses, file uploads, account details, device metadata, and usage analytics to use for its training. And for most users, these collected data are stored indefinitely, unless manually deleted.
With all these practices, OpenAI is currently dealing with legal issues, including facing Italy's data protection authority which fined OpenAI €15 million. In addition to that, they also face lawsuits from publishers like The New York Times and Canadian media companies, alleging copyright infringement, unauthorized data scraping, and the reproduction of proprietary content in ChatGPT outputs.
For users who are absolutely intent on protecting their privacy, OpenAI is not their refuge. That would be DuckDuckGo, which you can read more about here.
2. Bing
What about Bing? A traditional search engine that spurred quickly into action by integrating with OpenAI with their search 10 months before Google ever introduced Gemini. And Microsoft’s bet paid off in many ways. Bing has seen noticeable growth, with its market share rising from 2.81% in February 2023 to 3.64% in May 2024. While it’s still a far cry from Google’s 90%+ global dominance, it's a movement. More importantly, it signals how Microsoft was able to expect what users value when they search.
So how does Bing actually stack up as an alternative?
Its biggest advantage is that it is a true hybrid: it merges the strengths of traditional search (comprehensive web results) with generative AI by using the OpenAI model to power through Copilot. This makes it more up-to-date and broader than ChatGPT, especially when it comes to having more expansive index webpages.
Bing also delivers AI-generated summaries in a conversational format, giving users fast and convenient answers up front. And unlike ChatGPT, it still holds its ground on things like image search, video previews, and maps, where it competes more directly with Google.
And with Copilot, their biggest strength relies on how it is integrated with Microsoft productivity tools such as Word, Excel, Powerpoint, etc. And it’s focused on productivity, automation, and context-aware help within the Microsoft ecosystem.
Limitations: When thinking about Bing, the limitation would be almost the opposite of ChatGPT, which is the fact that it was originally a traditional search engine. So the follow-up question steps that ChatGPT was able to eliminate continue to persist. For example, Bing does not save past conversations, while ChatGPT keeps a helpful log of the user’s activity, making it easier to revisit and build on previous chats.
There’s also a deeper UX question here, one that applies to all search tools. Should search be a focused window that answers your question in one go? Or should it open multiple tabs and rabbit holes to explore? Bing still leans into the latter. It presents you with options, links, and paths that can be useful for some, but overwhelming or cluttered for others.
And while its seeming access to more sources should eliminate the risks of hallucinations, Bing was actually on top with ChatGPT when it came to having the highest reference hallucination scores from the same JMIR scientific article discussed above. At the same time, it imposes stricter session limits than ChatGPT, which limits the user’s activity.
What does this mean? Microsoft is experimenting with bridging the gap between the new expectations that came from AI search with the traditional search capabilities that remain important, especially for users who prefer classic search experiences.
At this point, it’s unclear which direction they’ll ultimately lean into. But in many ways, Bing and Google are facing the same challenge: trying to preserve the familiarity of traditional search while also keeping up with the rising demand for faster, more conversational, AI-driven results.
However, it’s fair to say that these tech giants are the ones being disrupted the most by the new expectations brought on by AI. And it’s hard to imagine them becoming irrelevant anytime soon, especially with how much infrastructure they’ve built (like Google Chrome having 66.6% of the global market for browsers and Microsoft tools still dominating in business with 46% of office softwares around the world).
They’re being pulled in two directions. One is maintaining the systems that made them dominant. The other is trying to meet the growing demand for faster, more intuitive, AI-driven experiences. Trying to do both at once might stretch them more than they expected.
Ethics: However, in the list of alternatives we have today, Bing is pretty much on par with Google when it comes to privacy and data especially on the enterprise level. It’s generally considered better with privacy and data for enterprise and business use compared to ChatGPT, due to its enterprise-grade security, clear data retention policies, and the fact that enterprise data is not used for model training.
The only difference is that Microsoft has not faced the same level of backlash as Google, especially when it comes to employee walkouts and working with the Israeli government. They’ve also faced some criticism when it comes to eliminating their AI Ethics Team which is a concerning issue with any oversights regarding AI development.
3. Perplexity AI
A very powerful alternative, and certainly with my seal of approval. Perplexity AI is another breed of search engine that is also dubbed as an “answer engine”, launched back in 2022. It’s built on top of advanced language models, including GPT-4, Mistral Large, Claude-3 and its own models designed to deliver concise, conversational answers with real-time citations from across the web. And by 2025, it has 15 million monthly active users, from just 2 million back in 2023, an impressive 650% increase over two years.
Now this is the reverse-Bing or reverse-Google, where Perplexity AI considers itself to be a hybrid model, except it’s focused on being the traditional search engine. Therefore, it is clutter-free from ads and provides a conversational tone. It blends traditional search (links, images, maps) with AI summaries, offering a “best of both worlds” experience, yet allowing users to maintain their history and not necessarily needing the tabs open as we’ve discussed in the Bing section.
But perhaps its most notable differentiator is its citation-backed responses. While many other AI-competitors such as Claude and Grok are going in this direction, Perplexity is best used for real-time facts and authenticity, especially as it gets their real-time sources from the web, academic articles, and social. Meanwhile, Claude grounds its citations from documents that the user provided and it excels referencing exact passages from those sources, but it does not natively perform real-time web search for citations. On the other hand, Grok has introduced web and social media citations, but independent studies show it struggles with citation accuracy and it generates far more incorrect citations than Perplexity.
Each answer Perplexity provides is backed up with sources that build user trust, as this allows for them to do easy fact-checking. From the same JMIR scientific article we’ve mentioned so far, Perplexity had a lower hallucination score than both ChatGPT and Bing. From the point of view of looking for accurate information that is easily digestible and comprehensive, Perplexity holds the torch from ChatGPT, Bing, and even Google.
Limitation: At present, due to its focus on citations and sources, Perplexity often lacks the creativity and personalisation that ChatGPT offers to its users. And without the ability to customize instructions or integrate with 3rd parties, Perplexity can provide overly long, dense, or comprehensive answers, making it hard to quickly extract what’s most relevant.
Also, while it tracks the user history and refers back to their sequence of questions, it lacks the ability for complex image generation or code tasks that can be found in ChatGPT or even Google Gemini, making it a primarily text-based search engine only.
What does this mean? Perplexity positions itself as the “ad-free answer engine” for users frustrated with the ad-heavy, cluttered, and SEO-gamed results that dominate Google and Bing. Its focus on transparency, speed, and citation-backed accuracy has built real momentum and attracted a growing user base.
But it didn’t ride the wave of OpenAI’s popularity, and it doesn’t have the default status that Google and Microsoft hold across browsers and operating systems. Like every other emerging player, its long-term challenge will be scaling, especially when going up against the massive infrastructure and distribution advantages of the tech giants.
As for where Perplexity goes from here, that’s still up for debate. But if current patterns hold, it’s likely to stay intentionally small, yet powerful. They would be tailored to a community of time-sensitive users who value depth, research integrity, and clean answers over mass-market appeal.
Ethics: Because Perplexity is considerably the least popular out of all the companies, there’s not much public knowledge around their controversies. However, their biggest controversy surrounds their web crawler scraping content from websites that explicitly prohibit it via robots.txt. It has also faced accusations of presenting content without proper attribution, sometimes closely paraphrasing or copying original sources. Since the company’s main ethos is to stay true to the sources, it’s interesting how it will evolve its ethics in the future.
How Google is Keeping Pace
It would be a disservice to dismiss Google’s efforts to evolve alongside these shifts. They’ve created their own AI models such as Gemini and Gemma, completely removing OpenAI from their flagship products, unlike Bing and Perplexity.
Their own models have been used to fuel Google’s AI Overview to present a concise, AI-generated summary on top of the search page results that they’ve rolled out back in 2023. And by 2025, AI Overviews went live for over a billion users in more than 100 countries. These summaries aim to answer complex questions directly, reducing the need to click through multiple links.
And they are putting in work to see how to design their next steps with their search engines. They’re running what seems to be the largest A/B tests in search history in order to ensure that they can keep users engaged on their page. Finding themselves with a similar position as Bing, they would need to assess user preferences and refine their search experience as they balance innovation with familiarity and reliability.
On the surface, this seems like a win for users: faster answers, less friction. But it introduces a major tension for the company: Google’s core business model still relies heavily on ad revenue and SEO-driven click-throughs. If users get what they need from the AI summary at the top of the page, they’re less likely to visit the actual websites that bid for attention through ads or search rankings.
This creates a strategic dilemma for Google. They have to balance improving the search experience with protecting the ecosystem of publishers and advertisers that keep their revenue engine running. Every time they move closer to a seamless, AI-powered answer box, they risk cannibalizing the very business model that made them dominant.
And while Perplexity lags behind the multi-modal capabilities of image and audio, Google Gemini (AI Mode), ChatGPT (GPT-4o), and Bing are pushing forward with increasingly sophisticated tools. These platforms allow users to upload images, interpret visual data, and in some cases, generate or analyze audio. Google’s AI mode in particular gives users the ability to ask complex, multi-part questions, follow up with clarifying queries, and receive synthesized, source-linked answers. Yet, like most AI models out there, all of these AIs are not immune to hallucinations with Perplexity having the lowest rate compared to the others. And depending on Google’s priority, they can either improve their citation sources or focus on other features that generate more revenue and retain more users.
Conclusion
As AI technologies become more sophisticated and people’s expectations continue to be shaped by these advancements, Google plays defense to ensure that they continue to sit on their throne as the dominant search engine. Throughout this piece, we explored how behaviors are shifting, which competitors are rising, and how Google is attempting to hold its ground by adapting its design and product strategy.
Now, is this enough to lessen the frustration around the company’s controversial positions, SEO-heavy websites and ad results? Time will tell. Its biggest advantage continues to be that it already was so dominant to begin with that their efforts are more ways to keep their users engaged in their sites rather than attracting new users. And while this article heavily focused on the AI alternatives, we also know that more younger users are shifting their habits to find more authentic voices through Social Media sites, Reddit, Instagram, or TikTok. But that’s for another story.
Google may not be dethroned overnight, but for a growing segment of users, it’s already being pushed to the sidelines. To end this one, I want to invite you to try out these different alternatives and see for yourself whether this quiet migration is just a passing trend, or if Google can plug the leak before more users quietly slip away.
Join the conversation: Which search tools (Google, ChatGPT, Bing, Perplexity, or other) are YOU using more often lately, and why?
Readers, Karen here: if you enjoyed this guest post, please check out ’s newsletter (PS Product Person Founder Series) and consider subscribing!
(manage your subscription here, or subscribe here)
Disclaimer: This content is for informational purposes only and does not and should not be considered professional advice. Information is believed to be current at the time of publication but may become outdated. Please verify details before relying on it.
All content, downloads, and services provided through 6 'P's in AI Pods (AI6P) publication are subject to the Publisher Terms available here. By using this content you agree to the Publisher Terms.
 | A guest post by
|
Thanks, Paula, for writing this!
This is a fascinating topic. A few points I thought would be interesting to discuss,
1. You’ve probably seen Google’s recent push to innovate new products and explore what the future of search and AI-driven interfaces looks like. It’ll be interesting to watch how this trend unfolds over the next year or two.
2. On the economic side, I think we’ll see consolidation among AI players. Startups like Perplexity or OpenAI may be well funded, but they are lost making, unlike Google or Meta are players with sustainable revenue streams. Think of it as another “railroad war”—definitely a drama worth following.
I did a follow-up research after I wrote a similar piece last year. I looked for evidence that companies like OpenAI or Perplexity have already crossed the finish line...
But the numbers tell a different story. They’ll probably need tens of millions more users before they really get there. Though I haven't checked the latest data ... :D
Thanks for another great post, Paola & Karen! 🤗 You’re definitely onto something, especially with ChatGPT and Perplexity rolling out shopping features. I also really appreciate the ethics commentary for each tool.