ChatGPT Search (CGS) is a landmark launch in the shift from traditional to AI Search.
Now, OpenAI competes with Google (Search) heads-on. Note the subtle elbow hit between the lines in the announcement:
Getting useful answers on the web can take a lot of effort. It often requires multiple searches and digging through links to find quality sources and the right information for you.
The positioning is clear: ChatGPT Search is a way to get a straight answer without digging through cluttered search results or browsing websites.
CGS, which is directly integrated with ChatGPT instead of a standalone search engine, decides whether a query benefits from web results or not, and you can rerun queries through other models like o1 preview to compare the answers:
ChatGPT will choose to search the web based on what you ask, or you can manually choose to search by clicking the web search icon.
It keeps the context of your search going in a conversation interface (bolding from me):
Go deeper with follow-up questions, and ChatGPT will consider the full context of your chat to get a better answer for you.
OpenAI has a strategic advantage, as I explained in Search GPT:
The Information reports that OpenAI loses $5 billion a year in expenses. Just capturing 3% of Google’s $175 billion Search business would allow OpenAI to recoup expenses.
OpenAI has a strategic advantage over Google: Search GPT can provide a very different, maybe less noisy, user experience than Google because it’s not reliant on ad revenue. In any decision regarding Search, Google needs to take ads into account.
CGS marks the entry to a new paradigm where traditional search engines like Google or Bing compete with AI chatbots.
They solve the same problems for users as search engines but with lower friction. But it also marks a critical event that should lead you to evaluate your strategy.
Companies face a choice to invest and “be early” to AI Search or ignore the noise and stay the course. What makes this decision hard:
- Divided opinions about Chat GPT’s chance to take significant market share from Google.
- Rapidly changing mechanisms of AI Search platforms.
- Confusion about what to do.
The first search engines didn’t represent the model (Google) that eventually won.
In the same vein, the AI Search experience we’re seeing today might be completely different in a few years. However, there is little doubt that search is fundamentally changing.
As a result, my recommendation is to invest in AI Search. It is not capital-intensive (yet), but the upside to finding a playbook is high.
If CGS grabs significant Google market share, you’re in a good position. If it fails, no harm is done.
Collision Course
In the chart above, I extrapolated ChatGPT’s and Google’s total traffic over the next two years if the trend from the last six months remains constant.
This chart will probably outrage or scare you, but the chance that events unfold exactly as depicted in this chart is low.
The reason I bring it up here is to consider the fact that many structural changes start slowly based on the saying “first gradually, then suddenly.”
It took Google about three to four years to beat Yahoo, Altavista, and Lycos. Given that new technology gets to critical mass ever faster, I’m not surprised ChatGPT could do it faster (in theory).
ChatGPT’s traffic has already passed the No. 3 search engine, Bing (YouTube is second).
When you look at comments and posts on social media, more and more people report using ChatGPT instead of Google for various purposes, but that could be availability bias.
One point a lot of people miss when looking at the traffic comparison between ChatGPT and Bing is that they’re not the same, and yet this is a fair comparison.
ChatGPT is more than a search engine. People use it for all sorts of things. But that’s exactly the point: a search engine that looks like Google never stood a chance to compete with Google or Bing.
CGS is something new, and that’s why it stands a chance. So, when you see chatgpt.com passing bing.com, the critical argument is not that both do different things but that they’re used to accomplish the same goal.
After all, search is just a way to solve problems or achieve goals, not to search for the sake of searching.
To clarify, I don’t think Google or Alphabet as a company is at risk of dying. I do think CGS has a chance to capture significant market share, and too many people underestimate how fast this can go.
Referral Traffic Skyrockets
AI Search marks a new paradigm where users get a direct answer without having to browse websites. So, how should companies think about pivoting their strategy?
Here’s what I’m telling my clients when they ask me whether they should pivot their SEO roadmap: For now, no. Reserve 10 to 20% of capacity to establish visibility in AI Search and for experimentation.
Look for signal: If you’re hesitant to invest more in AI Search right now, at least monitor traffic to and from ChatGPT. Base your decision on how long ChatGPT can keep its current traffic trend up.
Establishing visibility: This referral dashboard from Flow Agency is great for monitoring referral traffic.
With a few tweaks, you can monitor conversions in GA4 as well. You should also monitor site crawls from LLMs and your performance on Bing.
Then, experiment with content tweaks to improve your AI Search visibility. Keep investing in traditional SEO because it forms the basis of AI Search and answers.
Place a bet: The big question in this is whether you’re willing to take a bet or play it safe.
Being a first-mover to SEO had massive benefits as the incumbents tend to stay incumbents, mainly caused by strong backlink profiles, robust user signals, and brand familiarity.
For now, ChatGPT uses Bing search results to ground and weigh answers, which means sites with strong visibility on Bing also have a high chance of being very visible in CGS.
However, there is a chance that using Search for RAG (grounding) is just a jumping-off point until AI Search platforms have gathered enough of their data (queries and user behavior).
Early in this transition period, not much changes. Content that ranks well in traditional search engines, specifically Bing, gets a higher weighting in CGS, which means traditional SEO has a big impact on visibility in AI Search.
AI Chatbot referral traffic is skyrocketing, and ChatGPT’s new search capability could accelerate that growth even more.
Outgoing referral traffic from chatgpt.com increased massively in August and September, according to Similarweb.
Noticeable call-outs:
- YouTube’s referral traffic increased from .17% in July to 3.9% in September.
- Bing grew from 0% in April to 1.8% in September.
- Amazon grew from 0% in July to 1.1% in September.
If referral traffic keeps growing at the same rate, it will get interesting in the next six to 12 months. It’s not just the volume but also the quality of traffic.
People use longer and more complex questions when they engage with AI answers, according to Sundar Pichai. Length is a way to be more specific.
Longer questions allow search engines, LLMs, and marketers to better understand and serve users what they want.
Based on conversations and observations, referral traffic from AI chatbots isn’t consistently higher than search traffic in every case, but in most.
Looking Forward
I’m leaving you with two interesting questions:
1. Is it a coincidence that ChatGPT Search came out three days after Apple Intelligence launched publicly?
Apple launched Apple Intelligence, which uses ChatGPT in certain situations:
Apple is integrating ChatGPT access into experiences within iOS 18, iPadOS 18, and macOS Sequoia, allowing users to access its expertise — as well as its image- and document-understanding capabilities — without needing to jump between tools. Siri can tap into ChatGPT’s expertise when helpful.
Users are asked before any questions are sent to ChatGPT, along with any documents or photos, and Siri then presents the answer directly.
Additionally, ChatGPT will be available in Apple’s systemwide Writing Tools, which help users generate content for anything they are writing about. With Compose, users can also access ChatGPT image tools to generate images in a wide variety of styles to complement what they are writing.
We also know how valuable Google’s exclusive search deal with Apple is.
From Monopoly:
Apple’s impact on Google Search is massive. The court documents reveal that 28% of Google searches (US) come from Safari and make up 56% of search volume. Consider that Apple sees 10 billion searches per week across all of its devices, with 8 billion happening on Safari and 2 billion from Siri and Spotlight.
“Google receives only 7.6% of all queries on Apple devices through user-downloaded Chrome” and “10% of its searches on Apple devices through the Google Search App (GSA).” Google would take a big hit without the exclusive agreement with Apple.
Since Search is part of ChatGPT, any API request could trigger the new Search feature.
As a result, ChatGPT has a direct line to searches and actions on Apple devices whenever Apple Intelligence uses ChatGPT. Is that integration the new version of Google’s deal with Apple?
I speculated that OpenAI could work on a browser in Search GPT:
If the main benefit or Search GPT for OpenAI is a revenue stream and access to more user data, the next logical step for OpenAI is to build a (AI-powered) browser.
Browser data is incredibly valuable for understanding user behavior, personalization and LLM training. Best of all, it’s app-agnostic, so OpenAI could learn from users even when they use Perplexity or Google.
We’ve seen the power of browser data in the Google lawsuit, where it turned out Google relied on Chrome data all along for ranking. The only layer that’s more powerful is the operating system and device layer.
OpenAI seems to be very aware of the importance of being the default when we look at how hard it pushes its Chrome extension, which changes the default browser search engine to ChatGPT.
hey i’m really sorry to keep hyping our own product but you really should get chatgpt plus and install the chrome extension for search
i am cheerfully the first to admit when we ship something that isn’t very good, but this time it’s…really good
— Sam Altman (@sama) November 1, 2024
2. As it’s likely that more users don’t browse the web but get answers from ChatGPT, Gemini, Perplexity, etc. directly, will the open web become a place primarily for bots instead of humans? And how would that change the purpose and look of websites?
Featured Image: Paulo Bobita/Search Engine Journal