Late last month, Gartner made a stunning prediction, forecasting an imminent and dramatic decline in search engine traffic. “By 2026, traditional search engine volume will drop 25%,” the research firm said. “with search marketing losing market share to AI chatbots and other virtual agents.”
The prediction seemed aggressive, especially given that Microsoft Bing — with its AI chatbot — hadn’t gained much share at all in a hype-filled 2023. If it comes true, it would be an earth shaking moment in the tech world, leading to chaos within Google and the web. So I asked Gartner to talk.
This week, Gartner VP Alan Antin, who made the prediction, spoke with me via video call about what led to the prediction, and how likely it is to hold up. Though I’m still a bit skeptical, it didn’t seem so outlandish after we talked. Below is our conversation, edited for length and clarity.
Alex Kantrowitz: Don’t take this the wrong way, but when I saw Gartner’s number that search will decline 25% by 2026 I thought it was crazy. Why do you think that’s going to happen?
Alan Antin: Let me walk you through the thought process. We saw incredibly fast adoption of ChatGPT. The fastest ever to 100 million monthly users…
Right, but that has leveled off…
That has leveled off. That’s true. But think about this more broadly as answer engines. Some people are using ChatGPT, Claude, and other chatbots to answer questions like you would with search. As these bots become connected to the real time internet, the reliability of their answers is getting better. And they’re not going to be the last ones.
We don’t do the calculations of how many searches there are, so this is coming third party, but over 8 billion searches happen per day. So even with 100 million ChatGPT users, you might say, ‘Oh, you’re never gonna get there unless you see all kinds of other things come into the market.’ And the way we got to this potential decline in search traffic, is we have yet to see major companies who control a lot of the access to the internet besides Google, do anything in this space.
So if you think about it, there are over 1.5 billion Apple iPhones. All it takes is a new rollout. And suddenly, the access point to impact that giant number of daily searches happens without people having to download or subscribe to a version of ChatGPT.
Google pays Apple $18 billion a year to be the default iPhone search. So are you suggesting that Apple will forego that revenue to put a large language model that replaces search in their phone?
A shift in the marketplace might happen through new complete entry points. In which case, it might take longer for that to happen, because the behavior is going to need to evolve over a longer period of time. Or it could happen much more rapidly if a business decision was taken to say, well, we’re not going to have that same sort of platform relationship.
Right. And how did you get to the 25% number, exactly?
Just through internal debate.
Can you share a little bit about the contours of that debate?
I know you already think that 25% is a crazy number, but the first number put out there was actually even higher. The thought was you could see multiple entry points besides ChatGPT, and should you see some of the other companies that have large scale distribution come in, you could actually get there a lot faster.
So the contour of the debate was, well we don’t know what that cycle is going to look like. And so, it just became…. Within a couple of years time, the probability of one of these events happening — either an acceleration of new entry points, or a sea change in how current relationships and current platforms work — if you put some probability statement around each one, then you could say, ‘Okay, well that could get you to that point.’ So it’s not tremendously scientific. It’s probability statements that we look at to get to that.
But do you believe that search and chatting with large language models is zero sum? I have this hypothesis that search is effectively for web navigation, whereas a chatbot is to satisfy curiosity. There is some overlap, but not dramatic overlap between the two.
It depends on which platform you’re using. Using something like Perplexity, which is an AI-first search experience, where it’s giving you a nice summarized answer, but also saying, well, here’s where we drew that from… it’s giving you a blended model. Whereas, a pure chatbot may be satisfying a little bit more of a curiosity but you don’t necessarily know where it’s coming from, so you need to fact check that. You’re getting different experiences.
But from a research perspective, if you have it on your smartphone, and you are trying to get an answer, and if you believe their reliability is better, why would you want to go look at all of the links that turn up from traditional search, versus just ‘Oh, I got the answer right here in a couple of sentences.’ Depending upon the sophistication of the person doing the search, they may or may not have different considerations about that output.
So if people use Perplexity that would count as replacing the traditional search?
That’s right.
I’m curious to hear your thoughts on what this all means for Google.
They have fantastic products and they certainly are heavily invested in artificial intelligence. They have to be able to think through how they continue to deliver a great entry point and a great experience for users. And, at the same time, understand what the business model is going to be. That’s fundamentally what they have to work through and it’s a very challenging question.
If Google were to lose 25% of traffic within two years, that would be one of the greatest business tanks of all time.
I don’t know if you’re old enough to remember the earliest days of the consumer internet. There was a time where the main way to get on the internet was through a company called AOL…
Of course I remember
Things can change. We don’t have a crystal ball that says, ‘Well, we know exactly how that’s going to play out.’ But there are enough signs of change of behavior — and change of technology — that would suggest this is something that’s evolving and changing. Over the next couple years we’re going to see whether or not the incumbent players are able to figure out how to offer better experiences and not lose to new entrants.
Search replaced portals like AOL because search was a better way to access the web. But if you’re right about this 25% decline in search, it’s also a decline of the web. Because if you go chatbot instead of search, you’re not navigating to web pages.
It’s quite possible that the corollary is exactly that, which is the importance of individual web pages will also decline. But at the same time, think about what might be feeding large language models. Let’s say LLMs are the main way people are getting their answers and doing their research, there still needs to be some way for those models to be fed.
This is really a big part of what we wanted to get people starting to think about, which is you may need to be thinking about what your content strategy is and what your digital content strategy looks like. So that way, even if your site traffic goes down, which could very well be, you’re still going to be part of the conversation and part of the outputs that come from the answer engines.
I have covered internet companies for a while. In my experience, when a company comes across a formidable competitor, that tends to slow down its growth. But to lose 25% share over two years almost goes against the law of nature on the internet, where web traffic tends to expand year over year. So that being said, do you still stand firmly behind your 25% number?
Well, it’s a prediction. There are a lot of variables that are moving. We’ll probably update it next year because there are a lot of moving parts.
WorkOS, the modern identity management for B2B SaaS (sponsor)
WorkOS provides flexible and easy-to-use APIs for authentication, user identity, and complex enterprise features like SSO and SCIM provisioning. It’s a drop-in replacement for Auth0 and supports up to 1 million monthly active users for free. It’s perfect for B2B SaaS companies frustrated with high costs, opaque pricing, and lack of enterprise capabilities supported by legacy auth vendors.
This week on Big Technology Podcast: TikTok Sale Or Ban Moves Forward In U.S. — With Mike Solana
Mike Solana is the chief marketing officer at Founders Fund and editor-in-chief of Pirate Wires. He joins Big Technology Podcast to discuss the potential forced sale or ban of TikTok as the U.S. congress voted to pass a bill requiring it. Join us for an in-depth discussion of the arguements for and against banning the app, and how big money is influencing some of the key players. We also talk about Elon Musk, the state of X, Temu and Shein, the power of social media algorithms, and Devin, the first AI software engineer.
You can listen on Apple, Spotify, or wherever you get your podcasts.
Send me news, gossip, and scoops?
I’m always looking for new stories to write about, no matter big or small, from within the tech giants and the broader tech industry. You can share your tips here I will never publish identifying details without permission.
Thanks again for reading. Please share Big Technology if you like it!
And hit that Like Button and it will make Big Technology look great on the web, a limited time offer!
My book Always Day One digs into the tech giants’ inner workings, focusing on automation and culture. I’d be thrilled if you’d give it a read. You can find it here.
Questions? Email me by responding to this email, or by writing [email protected]
News tips? Find me on Signal at 516-695-8680
View Original Article
