Google’s parent company Alphabet lost $100 billion (approximately Rs. 8.3 lakh crore) or nearly a tenth of its market value after its new AI chatbot Bard failed to respond to a query in an ad promoting its launch. He claimed that the James Webb Space Telescope took the first pictures of planets outside Earth’s solar system, when in fact it was the European Southern Observatory’s Very Large Telescope.
At the same time, Microsoft saw its shares rise 3% as it announced it would integrate ChatGPT into Bing, Office and Teams. Microsoft is a significant shareholder in OpenAI, creators of this highly acclaimed AI chatbot.
Many are asking if we are witnessing Google’s Kodak moment, in reference to the famous demise of the American camera giant at the hands of digital photography. That might be an exaggeration, but we certainly think there’s some merit to investors’ concerns about Google’s future as a search engine company.
How Bard Disruption Happens Making a mistake is not a problem in itself. ChatGPT has been known to provide the wrong answers to queries with disturbing confidence. The big market reaction against Alphabet was more because the launch disaster broke the proverbial camel’s back. If Google can’t even run a convincing launch announcement about its new technology, he thought, can it really defend its search business? In our experience, companies are generally not stopped by a lack of technology or resources. More commonly, it’s because they lack imagination or struggle to reinvent themselves – often out of fear that developing a new business will undermine an existing one (known as cannibalization).
A lack of imagination is mostly the problem with longtime holders. Kodak, for example, could not imagine a world without photographic film and prints, and it paid a heavy price. Likewise, hotel groups were completely blindsided by Airbnb. They got little response, except to pressure government officials en masse against the service.
On the other hand, Google is at the forefront of developing the technology behind AI like ChatGPT. Known as large language models (LLM), they essentially work by assembling arrays of very powerful computers and “training” them on massive amounts of information from the Internet and elsewhere.
Google search scientists wrote the groundbreaking paper in 2017 in this area called “Attention is all you need”. Google incorporated LLMs into companies like Google Translate very successfully, though never into its mainstream search business. It seems likely that she fears cannibalization and the difficulty of reinventing her search business. Unfortunately, the status quo doesn’t seem viable either.
Google totally dominates search, with 84% of global traffic, and derives 70% of its revenues from this and other related markets. Having created a business on such a scale, he effectively has a monopoly (outside of certain countries like China, which do things their own way).
The problem is that AI chatbots like ChatGPT circumvent the need for a search engine, providing accurate and in most cases correct and creative answers to complex human queries. ChatGPT has become the fastest consumer app ever with over 100 million users as of November. In addition to Bard, several other companies, including Chinese search giant Baidu, are well advanced in developing their own LLMs. If there’s a better way to find out what’s on the internet, why bother searching Google anymore? Making Money with AI Chatbots For now, the business model for AI chatbots is unclear. Search is free for end users thanks to advertisers paying the other end for the customer traffic they receive from valuable search terms. It’s a predictable, high-margin business.
AI chatbots, on the other hand, are tricky. Would ads need to be inserted into responses to convince users to click through to certain advertiser sites? Would that seem inauthentic and cause consequences? How many ads would be too much? There’s no telling how far this would cannibalize Google’s search business, which must make it terrifying for management. Again, consider Kodak. She bought photo-sharing platform Ofoto in 2001 and could have developed it into a social media platform. Instead, it tried to protect its business by encouraging users to print more photos instead of sharing them with others.
This is how the main capabilities of successful companies end up becoming their main rigidities. Microsoft doesn’t have this problem precisely because it has never been able to successfully compete with Google since the launch of Bing in 2009. It makes only about 6% of its revenue from search, so it has much less to lose from industry disruption. He’s already given investors an idea of how he’ll incorporate ChatGPT into Bing’s ad model.
The Current Problem Google’s innovation has already been atrophiing in recent years. He closed promising deals such as the Stadia gaming platform and Duplex web-based automated booking tool. Elsewhere, it was too late, catching up with Amazon’s Echo smart speakers with Google Home.
Management mistakes are partly to blame, but so are impatient requirements to make a comeback. The stock market rewarded Google’s sharp focus on revenue growth and profitability, encouraging management to be less patient with its investments. Kodak’s market valuation was the highest in its history in 1996, before the global shift to the internet led to a notable collapse. Maybe we’ll say the same about Alphabet/Google in 2021.
In our experience, companies would rather have competitors kill their golden goose than do it themselves. That’s the pitfall Google must avoid. The only option is to start cannibalizing your search business.
Google could copy Microsoft’s approach with Bing and present Bard’s results as just one of the answers to search queries. This can decrease your ad sales as there are no actual bids for a bot’s response and no clicks that could take searchers to monetizable partner sites. But release this in beta, making it accessible only to those who pre-register, and you’ll at least contain the impact. Learn from experience, test different monetization models, and only scale when you see what works best.
Above all, Google cannot continue to thrive or even survive by thinking like an incumbent. It needs to reinvent itself. That means leaving something on the table, not trying to carry everything you own into the future. The sooner he realizes this, the more likely he is to survive.