Google Cloud eyes bigger market share with AI

Artificial intelligence (AI) has become the new battleground for major hyperscalers, all of which have been doubling down on generative AI platforms – whether it’s Amazon’s Bedrock, Microsoft’s Copilot or Google Cloud’s Vertex AI – to stay ahead of the pack and capitalise on the growing opportunities in the space.

While it remains to be seen who will emerge as winner in the race for generative AI supremacy, Google Cloud CEO Thomas Kurian is confident the company’s generative AI capabilities will enable it to grow faster than its rivals and draw more enterprises to use Google Cloud as their primary cloud provider.

At this year’s Google Cloud Next, Google Cloud doubled down on its AI-first strategy with a slew of offerings spanning areas such as tooling, foundation models and infrastructure to help organisations harness the rapidly emerging technology.

In a wide-ranging interview with Computer Weekly on the sidelines of the event, Kurian talks up the work Google Cloud is doing in generative AI, how it’s bringing those services to market, its synergies with other parts of Google and the company’s cyber security play with the acquisition of Mandiant.

We’ve heard a lot about Google Cloud’s AI-first strategy, but what does that mean for customers from an industry perspective? Are there industry-specific AI offerings we’re looking at?

Kurian: We have a variety of customers using our services in scenarios that are unique to each industry. For example, we have retail customers with large product catalogues that contain hundreds of thousands of products.

Maintaining all that information manually is very complicated, so those customers are using our AI services to generate product descriptions from product images. There are also customers using our conversational AI agent to help their customers shop. So, you can say “I’m looking for a sofa,” and the agent can ask, “What kind of sofa would you like?” We call that conversational commerce because through conversations we’re helping you to shop.

We definitely see that becoming profitable shows the discipline we've had in growing both top line and operating income. In fact, we are the only cloud provider for four years straight that has improved both

In healthcare, customers are using our search service to help doctors and nurses find information in electronic health record systems and clinical trial systems. “I have a patient coming to see me – can you tell me if he is at risk of diabetes? If he does, can you tell me what I should be prescribing him? That’s a search problem and we have customers deploying Vertex AI Search for that purpose.

Finally, The Estee Lauder Companies is also using our AI products to help them design better products that align with their brand voice. There are many such solutions that are being used in different industries, whether it’s telecoms, banking, securities trading, industrial manufacturing, supply chain or automotive.

With more generative AI projects expected to be driven by line of business units, how are you making sure that Google Cloud is able to sell to business users who increasingly have a bigger say in the budgets of generative AI deployments?

Kurian: Historically as a cloud organisation, we’ve sold outside of IT in many different departments. For example, when we sell to car companies, the assistive features that generative AI offers in vehicles are typically implemented by the software engineering leaders of the vehicles and not the IT department. We have worked with many of them in companies like Toyota, Renault, Hyundai and Kia Motors.

With banks, we talk to heads of consumer banking and heads of private banking. We also have work going on with wealth management companies, where they use our models to summarise the things that change in financial markets every day for their wealth managers. Those solutions are typically sold to the head of private banking. Our teams are trained to do that, but we also have partners like Accenture and Deloitte who are also driving those conversations.

Google Cloud has been gaining market share, but it still trails the top two players in the market. Do you think this focus on AI is going to narrow that gap in the longer term?

Kurian: We have grown faster than the market for four years straight now. And we have gained share on everyone else. We think AI is a totally new way that people buy technology. With the quality of our products and the breadth of offerings we have, we definitely see a lot of interest from customers and time will tell. But yes, we’re very confident that we will continue to grow faster than the market and grow very fast relative to the competition.

As you know, most companies today operate a hybrid multicloud environment. Are you expecting more customers to use Google Cloud as their primary cloud provider as you gain market share?

Kurian: Yes, and to be frank, when we first announced the notion of multicloud at this conference in 2019, people said there’s no way anybody will do that. They’re just going to pick one cloud provider and it’s either Amazon or Microsoft. That is not true. Four years later, almost every large company has multiple clouds. For some, we are the primary provider and for others, someone else is the primary provider. We definitely expect that with the solutions we offer for AI, we will get a bigger share of the spend over time.

I understand that Google Cloud became profitable earlier in the year. Do you think that will assuage customer concerns about the financial viability of Google Cloud?

Kurian: We definitely see that becoming profitable shows the discipline we’ve had in growing both top line and operating income. In fact, we are the only cloud provider for four years straight that has improved both. How much does that help with customers? Customers obviously want to invest in a division and a technology that is profitable because it can be self-sustaining.

For some, we are the primary provider and for others, someone else is the primary provider. We definitely expect that with the solutions we offer for AI, we will get a bigger share of the spend over time

Let’s talk about the synergies with the rest of Google. What are some of those synergies that could benefit customers?

Kurian: The most obvious synergy is that the models we’ve built with Google DeepMind are shared across the cloud and consumer businesses. That’s a big part of our value to customers, because they see that if we’re using these models inside search, ads and Bard, we must be paying attention to those models which must be of very high quality.

We do have a lot of projects that we’ve announced at this event that work with other parts of Google. For example, we announced a solution for sustainability along with Google Earth Engine, which is part of the Geo team at Google. We’ve also announced solutions with our marketing and advertising colleagues for marketing optimisation.

There are many different levels at which we offer more integrated solutions. At the lowest level is the models themselves which are shared. But when I say they’re shared, it doesn’t mean we’re taking customer data to improve our models. It means the engineering work that goes into building the models is done once, and then we offer the models to our consumer teams, as well as customers who can then tune the models for their own needs. We’re also doing work on grounding, which is a joint project between us and our search team. There are many places where there are joint engineering programmes that deliver value to customers.

I wanted to get your thoughts on Google Cloud’s security play. What’s the longer term plan? Are you looking to become a security provider, or is it more about making sure you have the security capabilities to secure the workloads that run on Google Cloud?

Kurian: We offer, first of all, a highly secure platform to run workloads and at this event, Kevin Mandia showed the difference in quality and security between us and the other two competitors, which shows you that we are much more secure.

That said, we also offer enterprise wide security solutions. The very simple proposition we have is the following: in order to secure your enterprise, you have to know what threats are happening. Then, you need to understand if you have been compromised by those threats? If so, how did the attacker get in? What do you need to do to fix it? And then how do you validate it? Now, we have the ability to get threat intelligence from Mandiant, put it into our security operations platform and use our AI to assist you in finding the problem and fixing it. And so, we do have customers choosing us as their security partner.