Google – Code Yellow

The warning lights are flashing yellow

• Apple claims that it has seen the first sign of generative AI impacting Google’s search traffic, meaning that Q2 25 will be a crucial quarter to see whether the anecdote translates into data.
• Even if there has been an impact, Google already has a competitive product, meaning that if it can execute on Gemini, it has a good chance of coming through this trial unscathed.
• Eddie Cue, Apple SVP of Services, testified in court last week that Apple had seen a decline in Google Searches in Safari during April for the first time in 22 years.
• Mr Cue also went on to say that Apple is actively looking at revamping the Safari browser and focusing on AI-powered search, raising the possibility that Google would no longer be the default search provider on iOS devices.
• Google immediately went on the defensive with a statement that led with: “We continue to see overall query growth in Search. That includes an increase in total queries coming from Apple’s devices and platforms”.
• These are not necessarily contradictory statements as Apple is referring to Safari only, while users can also conduct searches on iOS using the Chrome browser or Google’s dedicated app.
• Hence, this could have been caused by usage shifting from Safari to Google’s iOS apps although there is no evidence for this whatsoever.
• Search currently makes up 56% of revenues and so while erosion of these revenues is bad news, it is not the cataclysm that the bears are predicting.
• Hence, the outcome depends on Gemini which is Google’s answer to ChatGPT, and here, it has some substantial advantages.
• First, Google is already the default provider of search on Apple and has the financial muscle to continue paying Apple even if the market moves to generative AI while many of the start-ups (including OpenAI) simply do not have the resources to match it.
• Second, Gemini is a good product despite what its detractors say about it and with the functionality of foundation models converging, OpenAI’s advantage has already shrunk to a very slim margin.,
• Third, Google has over 2bn users on both iOS and Android who use its services on a daily basis, giving Google an installed base of users who are very likely to give its product a try before going elsewhere.
• Hence, all Gemini has to do is be as good as OpenAI, execute a smooth transition and it is likely that its users will stick with it.
• However, ChatGPT is a global household name with over 400m users and OpenAI will be counting on these users sticking with it in order to start making inroads into Google’s business.
• The problem for generative AI is that it is far more expensive for the service provider to provide an answer to a query than it is for search and, at the moment, most search queries remain very simple.
• Hence, it is possible that OpenAI’s backers get fed up with pumping in billions and billions of dollars before the market moves decisively in its direction and it can make a profit.
• Google has the resources to withstand that pressure and already has the monetisation systems in place to ensure that it can generate revenue from whatever method it uses to deliver the search query.
• The end result is that we are at a potential dislocation point in the search market that creates an opportunity for market leadership to change.
• However, unlike Nokia, Google already has a worthy product and so assuming that it executes the transition well, it should be able to keep OpenAI and the other upstarts in their box.
• Google has not covered itself with glory so far in generative AI which is why there is plenty of scepticism about whether or not Google will succeed.
• I suspect that we will get the first signs of this at Google I/O, which starts on May 20^th, where I am expecting a relentless campaign to entice developers and users to Gemini.
• This will give us the first indication, but it will all come down to execution to determine whether Google stays on top or goes the way of Nokia.