Show ChatGPT the $$$: Monetizing AI Chat
Bing's testing ads on chat responses, and they're rough
The Deload educates you about how technology is changing the world to make smarter growth investments. If you haven’t subscribed, join more than 1,500 rational optimists:
Show ChatGPT the Money
How will Google and Bing monetize chat?
Investors have pondered this question since ChatGPT roared into existence late last year. One answer is starting to become clear and probably isn’t that surprising: Ads.
At least Bing is starting to show ads in chat responses. I got these ads in a Bing chat about how to train your dog not to bark (h/t to Mike Olson from Deepwater for the query).
Obviously Microsoft is just experimenting here, but dumping existing product ads at the bottom of a chat response is pretty ugly. It’s not relevant to the conversation, it’s not integrated into the content, and it’s a bad user experience. The chat response didn’t suggest I use any of the promoted tools to train my dog not to bark, which was the initial prompt. Should I now ask Bing how I might use one of these products? Do I ask it which product might be best for my dog?
Product ads in AI chat like this feel like display ads on Facebook in 2008.
If ads are ultimately part of AI chat monetization, we should assume they feel more like News Feed ads than poorly targeted, legacy product ads.
An ad within the context of the chat experience would feel more natural. I prompted Bing to tell me about cleaning tile and told it to include an ad in the response. Here’s what Bing came up with:
While this ad experience makes much more sense, it’s a form of hallucination (although I did ask it for an ad). The phone number offered isn’t real (perhaps obvious given the digits), and tilemasters.com takes me to a website for a tile company in Kansas with landing page copy that promises customers will “EXPERIENCE THE BEST LAY IN TOWN.” They chose all caps, not me. Could have been a great match, but sadly I live in Pennsylvania.
As a proof of concept though, a plug for a company to help you with whatever your query was about seems like an obvious and useful way to integrate ads into AI chat responses.
Given Google’s DNA as a company that makes free products available to as many users possible, and Bing’s direct competition with Google, it seems inevitable that ads will be a part of the AI chat experience of the future. Maybe there are more creative ways to insert promotions into chat responses, or maybe Google and Microsoft need to reconsider ads as the basis of monetization for AI products.
The Bigger Monetization Picture
The major bear case on Google has been that AI chat is going to cannibalize search queries; however, there are key differences in AI chat use cases that may determine different maximal paths to monetization.
I use a simple mental model to frame search vs AI chat use cases:
Search has two core use cases. People use it to look up facts and find things to buy. Both use cases are things that AI chat can also do, although it’s debatable whether AI does it any better than Google. At least for now.
For fact searches, Google does a good job of building answers into results, and it doesn’t suffer from hallucinations. For commerce, Google gives you a set of options, which may be preferred by users who want to make a final decision themselves vs outsource it to a single recommendation from AI.
Most people assume that AI will improve in time to surpass any advantage Google might have today in fact and commerce searches. I’d argue it doesn’t matter because AI does things tied to those two use cases that traditional search can’t do.
The bars connecting the left and right hand side of my mental model represent loose (dotted) and strong (solid) ties to AI-specific use cases.
When we search for facts, it’s often in preparation to create something — like writing an article — or to complete some task — find directions to go to the DMV. Many tasks include a commercial angle, and all engagements in commerce result in the task of buying something.
Where AI chat has super powers compared to search is that it can perform acts of creation for you. You could ask Bing/ChatGPT to write an article for you based on some information it found for you. And with AI agents, you could engage in the commercial act of hailing an Uber to get to the DMV to finish your task of getting a new driver’s license.
The distinction between fact and commerce searches vs creation and task completion opens the possibility for new mechanisms of monetization for AI chat.
Fact-based searches are unmonetizable by Google today. They’re a loss leader to keep users engaged to perform commercial queries that are monetized by search ads.
For AI chat, fact-based queries could tie to a paid offering to use advanced creation tools for writing, audio, video, etc. If the AI chat bot is good at creating, it should save the user time, and saving time is always worth money. Commercial requests could be sent by the AI to preferred vendors that pay fees for conversions like in-app purchases on the App Store. When I ask my AI chatbot to book a trip for me, there are possible integrations with hotels, airlines, car services, and more that could be monetized.
The more the AI knows about you, the better it will be at creating things in your voice and performing tasks to your liking. This is the holy grail of Personal AI we’ve talked about previously. As a paid opportunity, I think it’s a trillion dollar market, but I don’t know whether Google has the DNA to embrace the paid services model as its core business. We already know Microsoft does with Windows/Office/Xbox/etc.
Facebook News Feed is the Guide
Google has never experienced a meaningful shift in how consumers search for and engage with information in it’s 20+ year history, but we do have another guide in Facebook.
Facebook survived a paradigm shift in how consumers engage in social — the move from PC to mobile. And the company came out stronger for it. Facebook innovated with News Feed ads to capture the opportunity presented by the mobile era that brought new avenues of usage and growth. The reason News Feed ads worked so well then, and still do now, is because they’re integrated into the user’s experience.
Great products, even ad products, survive the test of natural evolution. Products that resonate with users and drive engagement and willingness to pay survive and thrive. Products that don’t die.
Whether Google or Microsoft or any other AI chat company wants to use ads as a monetization mechanism or paid or an App Store model is less relevant than how the monetization mechanism supports the user experience. In the end, users should win, even if that means Google monetizes us even more efficiently than they do today.
Disclaimer: My views here do not constitute investment advice. They are for educational purposes only. My firm, Deepwater Asset Management, may hold positions in securities I write about. See our full disclaimer.
Intelligent Indices: ChatGPT is Beating the S&P 500
Here’s where ChatGPT is really seeing the money: In trying to beat the S&P 500.
About a month ago, I created a number of Intelligent Indices powered by ChatGPT, Bard, and Claude to see if the world’s best AI tools can beat the market. Now that we’re a month in, we can look at inception to date performance, and the results are promising so far.
The flagship Intelligent Select beat the S&P 500 by 50 bps since inception, and the Intelligent Tech Select outperformed the Nasdaq 100 by 90 bps. Most impressive so far is the Intelligent Select Equal which beat the S&P 500 Equal Weight index (RSP ETF) by 170 bps.
Yes, it’s one month of data. Yes, it’s only a back test. But the experience with Intelligent Indices has led me to a few conclusions:
AI is already capable of creating usable stock indices with intelligently engineered prompts, and it’s only going to improve over time.
As AI influences investment processes more, we need intelligent benchmarks as barometers for AI-enhanced investment strategies.
Like many other industries with entrenched dominant players, AI will create the potential for something new and dynamic to disrupt legacy indices like the S&P 500.
It’s hard to imagine a future where there isn’t an AI-powered index scrolling on CNBC all day next to the S&P.
The future of indexing is intelligent.