Main content

Google Cloud President - ‘We can be the adult in the room when it comes to AI’

Derek du Preez Profile picture for user ddpreez April 4, 2024
Summary:
Recently appointed President of Google Cloud’s Global Field Organization, Matt Renner explains why the vendor is well suited to help enterprise buyers pin down their AI strategies.

Image of Google Cloud logo
(Image sourced via Google Cloud)

Matt Renner has a long history in the enterprise technology industry. Having worked at Oracle, Microsoft and Salesforce prior to joining Google Cloud at the beginning of 2023, Renner has decades of experience working with buyers as they seek to adopt - and identify use cases for - new technologies.

Upon taking his new role as President of Google Cloud’s Global Field Organization just two weeks ago, responsible for all go-to-market functions at the vendor, Renner sat down with diginomica to discuss the challenges and opportunities that AI presents for enterprise buyers - and more importantly, advise on how they can better achieve success.

Renner’s pitch ahead of Google Cloud’s annual conference in Las Vegas next week is that the vendor, thanks to its history in working with data and Artificial Intelligence (AI), and due to its recent investments in enterprise-grade cloud computing, can be the ‘adult in the room’ when it comes to AI decision making. 

At last year’s Next event we saw how Google Cloud is taking a multi-pronged approach to AI, ranging from updates to its infrastructure, to using a platform approach with Vertex for the adoption of AI models and for tuning, as well as incorporating Gemini on the front-end for Google users. Google Cloud’s philosophy (so it says) is to adopt an open approach to AI, much like it has done with multi-cloud tooling, so that users have more choice about where and how to invest. 

Renner says that when he was at Microsoft, if he were to go up against Google Cloud in a pitch, he would seek to question the vendor’s enterprise credentials, given that it was new to the Cloud and enterprise market. However, he notes that seven years into its enterprise operations, Google Cloud undoubtedly has the scale to compete and has squashed any hesitation that it isn’t a realistic alternative compared to vendors that have a long history in the market:

Google has worked its way up over the last five or six years, really checking all the boxes and building up the proper infrastructure to address the enterprise. Google has a tremendous history, there are over a billion consumer users. We took the knowledge and the infrastructure that we honed under that incredible pressure and we can now apply it to the enterprise. 

What I’ve found here is that this is an adult setup. We're profitable, we’ve got over 15,000 folks, we're in over 200 countries. You look at our scale and the capability we have: we're supporting some of the largest, most complex environments on the planet with Google Cloud. So I think we've passed all the tests of being asked the question: are you enterprise ready? And it's been a pretty amazing time to join. 

We're now just trying to compete, service to service. We always had a great product, we always had a phenomenal cloud, it was always a secure product. But the nice part about where we are today, is we have checked all the boxes on scale, profitability, size. We're here to stay and we're accelerating. 

Making AI successful

Google Cloud will no doubt make a swathe of AI announcements next week at its Las Vegas user event, but diginomica will be in attendance looking to understand more deeply what practical guidance the vendor is providing to customers on how they can successfully deploy AI. I firmly believe that the vendors that can authoritatively guide buyers through this complex landscape towards thoughtful, useful AI projects, will be the ones that see success. There are too many pitfalls with AI for it to be taken lightly and trust will be easily diminished if vendors can’t navigate this with care. 

Renner had some thoughts on this too. He notes that over the past year, throughout all of the AI hype, enterprises have been experimenting and, due to pressures from the C-Suite, have been running a number of AI proof of concepts (PoCs). However, as Renner says: 

You have a lot of customers that have had a lot of pressure in the last year to come up with some solutions. Hundreds and thousands of PoCs were spun up overnight, just because the CEO and the CIO said, ‘I've got to go get a bunch of stuff going in AI’. But 90%+ of these have failed. 

Renner’s measure of failure is that very few of these have made it into production. A proof of concept is all well and good, but if it’s not live and delivering the intended benefits, then it’s not working as intended. Commenting on why the projects are failing, Renner argues that companies are either not putting the necessary effort and resources into them to make them work, or they’re struggling with their data. He explained: 

That's one thing customers ask us - 'How do I sift through this and find one that makes sense?' 

And the most dangerous thing you can deal with, with AI, is data. Meaning this: if you don't understand your data today, you don't understand the complexities of the problem. Your challenge is that you underestimate what you have to do with your data. 

Why is that important? That's our wheelhouse. We've been data experts for 20 years. So when you come to us, we can assess is :it a good business case? How is your data? Do you have tools that can help with your data? And with this we see a lot of traction. 

Remaining open

Renner says that customers coming to Google Cloud are asking for help with prioritization when it comes to their AI projects. They are seeking help with trying to understand where success lies and how they can develop a strategy. As noted above, Google Cloud’s approach to enterprise adoption has been to promote a relatively open platform compared to its competitors. As we saw with its cloud infrastructure strategy, emphasis was placed upon multi-cloud and giving customers a portal by which they could choose other providers if they desired

It is seeking to take a similar approach to AI. And Renner believes that buyers looking for help with their AI projects should look for partners that provide choice. He explains: 

Have a strategy, right? It's funny how a lot of activities sometimes feel like a strategy, but they’re not really a strategy. How is this going to plug in with your IT strategy or your business strategy? 

Second, partner with folks that are open and give you options. From a Google perspective, we feel like we're the most open platform that's been created. We give you variety, we give you options. 

Commenting on the product itself, Renner says: 

It starts all the way down at the chip level. We've got chips, but we bring in other chips with some of our partners. At the coding level, there's variety relative to that. You get to Vertex as a platform, this allows you to plug into virtually anything, it’s wide open and was built for that. 

At the data layer, BigQuery as a tool was built to traverse your data sources, not just ours. And then last but not least you get to the model level for generative AI - we've got models, but we also have a model garden. You have variety here. 

I do think what we can do, which is different from other competitors, is we give you options in all these cases. Most of our competitors pick one path. Choose a partner who can come to you with some expertise, and ultimately make sure they're open. Make sure they're an open platform. 

In terms of industry examples, Renner identifies healthcare as an area with lots of ‘low hanging fruit’. He pointed to how hospitals don’t have enough medical practitioners to deal with demand, and yet have nurses spending hours of their time carrying out paperwork that could effectively be automated using AI. For example, an AI system could be built to record a nurse’s voice, consolidate their notes in real-time and produce a document in minutes, instead of spending hours writing up what’s needed. 

Another example in healthcare is Google Cloud’s work with Mayo Clinic, where in its oncology work it is using Artificial Intelligence to analyze and speed up the process of contouring (the critical process whereby doctors outline tumors for radiation). Renner explains: 

It takes a clinician almost eight hours to do contouring. What contouring is, is basically when you're targeting radiation, you don't want to hit the live tissue, you want to hit the problem tissue. You can sit there for eight hours with ten different charts and figure this out. 

When you apply AI, it took about 15 minutes and it was more accurate. When I use the term productivity, they're going to be able to see more patients. That may help the bottom line, but at the end of the day, their first goal is to help more people. 

But the key message Renner is hoping that Google Cloud can get across, thanks to its experience with data in the industry over the last two decades, is that it can help buyers have a grown-up conversation about what AI success looks like:

I will tell you customers are coming to us very consistently, and asking us to help, relative to prioritization and ultimately, success. If you look at the evolution of this, we firmly believe that we can take the adult slot, when we are at the AI table with a customer. The reason we can do that is because for the last 20 years we've been the innovators in AI. It's not really that debatable that between Google and DeepMind, we've had every major breakthrough for the last 20 years. 

We have the depth of experience on how you deliver enterprise grade scale with AI. But importantly, with that experience comes the ability to prioritize, and frankly, work with a customer as to how they can actually be successful with AI. 

It's a marathon, not a sprint, with AI. We have got through all the fervor of the last year, but now let's change this to an adult conversation about how we really get things to production. 

My take

Keep a close eye on diginomica next week as we will be on the ground with Google Cloud in Las Vegas, analyzing the latest news, delivering updates from executives, and speaking to customers about their AI investments. It was difficult last year to assess the impact of Generative AI announcements, given that the technology was still in its infancy. But now, in 2024, we are looking for the practical guidance, real-life advice, and will be assessing the progress made so far by customers. 

Loading
A grey colored placeholder image