📚 And the AI winner is…IBM?

Who’s the leader in AI? If you ask Wall Street or the media, the answer you almost always get is either Nvidia or OpenAI. Google and Microsoft also get some mentions. But for the most part, those four vendors seem to garner all the good AI ink.

But if you ask enterprises, it’s a different story. Especially if you ask enterprises which vendor is driving the big-benefit AI projects. Then it’s a very different story. IBM gets nearly as good AI recognition among enterprises as Nvidia or OpenAI, and IBM is a walkaway winner in the key-AI-project category.

So, are we missing something here? Yes, a lot.

I’ve always tracked and measured how much strategic influence vendors have on buyers of enterprise technology. A score of 100 would mean that buyers tracked a vendor’s recommendations all the time, and zero would mean the vendor had no influence at all. Most tech vendors have scores that have hovered in the tens or twenties, but IBM has never scored less than 35. And while every other vendor is seeing a slow erosion in broad strategic influence, IBM’s influence has been rising. Among its major accounts, IBM scored well over 50 in 2023, and in its ability to influence its accounts on AI issues, IBM scores in the 70s. The reasons for that are really important to both IBM and to AI.

The first reason is simple. IBM believes its buyers are smart, and it wants them to be. Most vendors don’t. For decades, I’ve listened to sales management tell the sales force to avoid consultative selling, avoid “educating” the customer. Make your darn numbers, sales management says. For decades (six, in fact) IBM has taken another path. They used to hand out notebooks to people, emblazoned with one word: Think. I’ve sat in on CIO/CTO-level meetings that included IBM’s account teams, and I was struck by how much the team worked to draw out that thinking. And not just thinking about technology, thinking about how technology changes business.

That brings the second point, which is that IBM has focused its AI on the business case instead of trying to get customers to focus on AI. Why tell your customers to think if what you really want them to do is mindlessly write a check? That’s a good question, but the obvious follow-on is: Why isn’t writing a check good enough? It’s because a notion that AI might be valuable could fund an experiment, but not a significant deployment. AI is entertainment to most people, nothing more. All we’ve proved so far is that if you give something away, people will take it. A thinking CIO or a key planner can put all that aside if they try, and see the real question. It isn’t whether AI is interesting or sort of helpful, but how can I apply AI so it helps my bottom line?

A good way to start is telling you where it won’t help. I saw an article recently recounting how people were using AI to write employee reviews. Just how much money do you suppose that’s making companies? Even things like writing PR material, helping with documents, or answering emails, while they may offer some limited productivity improvement, don’t even nudge the bottom line. IBM is asking its customers to think about is how IBM could actually help them justify really investing in AI. And they’re getting the right answer, which is that real investment requires real benefits.

Where does a company look for real changes in profitability? It’s in areas like sales targeting, manufacturing and transportation efficiency – all parts of business operations and business intelligence. Enterprises are big, and it takes big changes to move big money. Do you think you get that by doing better press releases or writing better manuals? More importantly, do you think that it took all the AI publicity to reveal to enterprises that they needed to be thinking about the stuff that could really move that money? No chance; enterprises have known it all along, and they already had initiatives – the stuff we call “analytics” – aimed at the problems and opportunities that really needed action. When AI came along, they thought, they told a receptive IBM account team, and IBM has been focusing its AI initiatives on those problems ever since.

IBM has seen AI, its watsonx, as a business intelligence and analytics tool. They didn’t come on that as a stunning revelation in a recent sales call, they knew it because they were there thinking too. They’ve focused their AI discussions with enterprises—the same ones who’ve then told me about this—and they’re finding the real applications of AI, even what’s now called “generative” AI. Large language models, trained on a company’s own data and on public data about other companies in the same and in other verticals, can tell managers and planners where improvements can be made. This isn’t the kind of AI that could threaten your very existence unless you’re some kind of financial geek. There’s no broad public interest in this kind of AI vision, no clicks to bait, no PR mountains to move. All that’s there is the kind of value, the kind of business case, that moves financial mountains. The kind that drives AI success, and has driven IBM’s success in AI, where it counts.

Ah, but you’re probably thinking that thinking isn’t something only IBM can do. That maybe competing AI types have been resting on clicks and PR, but once IBM starts showing where the real value of AI is, those competitors will think too. They’ll run away to build their own analytics and BI around their AI stuff, and old stodgy IBM will be left in the dust, right?

Stodgy? Well, consider this. IBM had competitors in the 1950s when they started selling business computing. None are left today. The architecture that IBM announced with their System/360 in the 1960sis the foundation of the Z-series mainframes that IBM still sells. Not because it was the greatest architecture of all time, but because it’s the one that enterprises built mission-critical applications on. What’s keeping IBM in the mainframe business is the value of those applications, a value that’s endured for sixty years now. Why have those applications lasted that long? Because they weren’t the glamorous ones, they were the right ones. Their recent deal with Wipro, which now brings watsonx to key verticals, is a good example.

I’m not trying to glorify IBM here, I’m trying to recognize the approach that made them an enduring giant. PR is great, fables are entertaining, and there are endless stories we can tell about stuff like AI, stories that get more interesting and exciting with each telling. But in the end, what moves technology is what moves technology markets, and for enterprises that’s the same thing it was even a hundred years ago: the bottom line. The most important thing an AI vendor could say to a prospect is “I can show you the business case.” Who’s saying that? You guessed it. Look at their recently announced deal with Wipro, a deal that promises “specialized industry use cases” for five key verticals: banking, retail, health, energy, and manufacturing.

Does all this sound outlandish, boring, to you, like dry accounting stuff? Is it running counter to everything else you hear, everything you read, about AI? Is it challenging all that has seemed new and exciting about AI? If so, then I have a suggestion for you, and I bet you guessed that too.

Think.