元AWSのレジェンドが、企業がAIを本当に機能させるために必要なものを解説

Enterprise AI projects go off the rails when companies focus on the technology instead of the people.

So says Matt Domo, co-founder of AWS's database division, founder of AI consultancy FifthVantage, and also author of a tome titled “Everybody Wins”.

"The number one reason these fail is because the business and leadership, and how work gets done and decisions get made, don't change in kind for the new way things are done," Domo said in an interview with *The Register*.

Domo said he experienced this across his entire career. When AWS was new, he said, people couldn't fathom it. They couldn't see how this cloud technology might work without their own constraints. They only saw things through the lens of their existing operations.

"That's the biggest failure here," he said, before asking: "So how do we solve that?"

The first step, he argues, is taking a step back and analyzing what the organization is trying to do, who might benefit from that, how people might use the technology, and how to measure success.

The feature war is over. It’s about value

Companies, he said, need to look more closely at how technology – more than just AI – can be used to create the value people want.

"The feature war is over," Domo said. "It isn't about features anymore. It's about value. Ask how many CIOs are happy with the eight-digit forklifts they did with software packages like CRM and what value came from it."

From there, he said, it's a matter of methodically working through what has to change to deliver the desired customer experience, and working through how employees deliver that experience.

That kind of review is necessary, he said, "because you're going to speed some things up. You're going to see new signals. You have to adapt."

Organizational change, in other words, also requires a clean assessment of the signals that matter – and those that should be downplayed or ignored.

"What I see is leaders are tired of the 'robots are going to take over the world' and the negativity," said Domo. "Seventy-five percent of CEOs are panicking that they're not going to come up with an AI strategy and will lose their job.”

Domo thinks the AI component of IT budgets will rise 86 percent this year and buyers will expect real results. “We've crossed from theory to ‘Stuff's gotta work now’. We gotta get value. People have to see ROI. We have to see benefits."

• Anthropic tests how devs react to yanking Claude Code from Pro plan

• The spaghettified DBMS chart that shows Oracle's crown is slowly slipping

• OpenAI now lets you screenshot your privacy in the foot

• Another npm supply chain worm is tearing through dev environments

Automation enabled by AI must be considered, but Domo argues that business signals are the key.

"The unlock is the ability to process signals and look around corners and make predictive decisions," he said. "If you focus solely on automation, you miss the biggest unlock of the decade."

As an example, Domo recounted work done with a SaaS company that tried to address customer churn by creating a team to bring back customers who cancelled the service.

If you focus solely on automation, you miss the biggest unlock of the decade

"They'd have a couple dozen people call them and they would celebrate bringing a handful back," he said. "High stress, low success."

The problems, he said, were evident from the analytical data – fewer people logging in over time, session length declining, heated sentiment during chatbot discussions, and so on. Automating collection and processing of that data allowed the company to intervene earlier.

"It's much better curing a customer before they leave than trying to convince them to return after they're gone," he said. The data analysis also helped the pre-buy sales process by improving the company's chatbot interactions and marketing email sequences.

"They got tremendous uplift from all that by peeking around corners," he explained. "Customers want it hyper-personalized, they want it easy. They want it focused on what they're trying to do."

The days when customers would accept poorly designed products are gone, Domo insists.

"People won't accept it," he said. "And there's never been a time before where I could easily replicate a product. The price, the speed, the tools – I could step in and clone what you're doing easier than ever before. So if you don't rise to the challenge, people are going to switch."

As to the need for automation, Domo emphasized that it's not something to be done with an eye toward laying people off. Rather it's about freeing staff up to focus on high-value problems.

"At the speed all of this is going, the number one thing to focus on is reducing the delta between deciding and doing," said Domo. "And so as you figure out your pilot, start small, move fast, learn and iterate. Get a few of those under your belt that add tangible ROI to the P&L, to the costs, and then expand once you have the air under your wings. Now those pilots that have been stuck for 18 months, they're going to start going into production and adding value. And along the way, you as a leader, your team, and your stakeholders are going to be more confident." ®