My old friend and colleague Chunka Mui recently posted a thoughtful essay on how companies should start thinking about the next phases of the generative AI revolution — which is where the profound changes will happen.
By now, just about every company is experimenting with gen AI, and many even have gone beyond pilots and into production. But, drawing on Doug Engelbart's classic thinking about businesses' "A, B and C processes," Chunka lays out the need to go beyond AI's usage in what Englelbart would call A processes — those that a company uses to operate every day. Companies need to institute B processes, which are designed to improve those processes that run the business. And — here's the real payoff — companies need to design C processes to improve the B processes.
I know that can sound rather theoretical, but Chunka shows how it all gets practical very fast — and the approach has worked before. Drawing on Engelbart, among others, Chunka was the co-author of "Unleashing the Killer App: Digital Strategies for Market Dominance," which was a huge best-seller and a sort of bible for innovators in the internet boom of the late 1990s. Even after the bubble burst, the publisher of the Wall Street Journal lauded the insights in 2005 and labeled "Killer App" one of the five best books on business and the internet.
I'll give you the short version of Chunka's thinking and apply it to insurance, leading up to his seven questions that will help ensure that you're seeing the full potential for generative AI in your business.
To put insurance terms to the ideas in Chunka's piece, A-level processes are how agents and brokers sell, how underwriters price risk, how claims are settled and how customer service centers operate. We've all seen stories, and probably even personally experienced, how AI is being deployed at this level.
B-level processes improve those processes, and it's easy to see how gen AI can make the improvements happen even faster. The AI can, for instance, instantly spot patterns in responses to sales pitches to see what works and what doesn't work, including nuances such as time of day, day of the week, number of weeks or months before a renewal, proximity to a life event, etc. The AI can detect emotions that humans may miss as potential customers talk; the AI can then pass the information to agents, helping shape the conversations. And so on. The AI can also speed the process by which learnings are gathered, distilled and fed back to agents so the A-level processes improve and agents can be more effective next time. The same sorts of B-level improvements can happen in claims, underwriting, customer service and other parts of insurers' businesses.
From what I've seen, many companies are at least starting to think about B processes. I've published quite a few pieces about, for instance, using AI to catch fraud, to have results from claims fed back to underwriters to improve their appraisal of risks, and to help underwriters both gather data more efficiently and to highlight changes in the policyholder's situation since the last policy was issued.
But I have yet to hear about much in the way of using AI to get to the next level, to the C processes. They're a bit harder to characterize but are crucial. As Chunka writes, "C-level work isn't merely about scaling incremental improvement—it enables organizations to question and redefine their very purpose. It allows not just better performance, but different futures. C-level improvements accelerate the rate and type of change—unlocking exponential leverage."
From the initial internet boom, I'd say Amazon is the best example of C-level thinking. It started out selling books and continually worked to sell books more efficiently — showing A-level and B-level processes — but was always driven by a C-level vision that founder Jeff Bezos referred to as "The Everything Store." He wanted to sell everything to everybody, even as he founded the company more than 30 years ago.
Amazon Prime was a direct outgrowth of that vision. Once Bezos started to host enough other businesses on the Amazon site, he saw he could lock in customers by offering them fast delivery based on an annual fee — getting them out of the habit of factoring shipping costs into every purchase. That lock-in then helped him attract more merchants, feeding a virtuous circle that continues to this day.
AWS wasn't foreseeable back in the early days of Amazon but is the sort of happy accident that can happen when you set out for a C-level reinvention rather than just a B-level continual improvement. Bezos saw that many merchants needed help operating their sites, so he started a cloud service — and being in the business early let him see the huge demand before potential competitors and get a massive head start that has translated into a business that generated $108 billion of revenue last year, with an operating margin north of 35%.
For insurers, I could see a C-level approach to gen AI facilitating the move toward a Predict & Prevent model, beyond today's repair-and-replace approach to risk and losses. Gen AI can gather information — even across the silos that bedevil insurers — and analyze it instantly, then send it to whomever needs to have it, in time to perhaps prevent a loss.
A well-meaning recent attempt to get bad drivers to improve was based on a single communication to individuals with multiple moving violations, whose behavior was then monitored for the next six months. It won't shock you that driver behavior changed not at all. What we need is the sort of instant information that Nauto provides to truck fleet drivers about speeding, about tailgating, about drowsiness, about road conditions and accidents ahead, etc., based on AI analysis of images from cameras: one facing the road, one facing the driver. A C-level approach to innovation with gen AI can facilitate that sort of timely feedback — and not just for drivers. It can also help, for instance, the timely provision of information to utilities about faults in electric lines, as detected by Whisker Labs' Ting sensors in people's homes. A C-level use of gen AI could help communities monitor and encourage homeowners to harden their properties against wildfire, reducing the risks for everyone. And so on.
More generally, gen AI can be used to flesh out the sort of what-if scenarios that business leaders use to stretch their thinking and prepare for challenges and opportunities. Instead of just briefly entertaining the thought of a recession, of war spreading from Ukraine to other parts of Europe, or of even more remote possibilities, leaders can use gen AI to develop more elaborate scenarios and explore the complex interactions that may matter to a business but that are hard to see in a quick look. Even at huge companies that have planning departments, gen AI can help flesh out scenarios faster — gen AI could look at today's weak jobs numbers in the U.S. and speculate in detail on what it means for workers' comp enrollment, for employee-sponsored healthcare programs, for general economic growth, for Fed rate cuts and more.
"Killer App" explained the power of what-if analysis, in one of the many parts of the book that have stuck with me. Chunka said the invention of the electronic spreadsheet in the late 1970s led directly to the wave of mergers and acquisitions in the 1980s and 1990s. Why? While smart young financial analysts could always crunch numbers, they previously had to manually update every cell in a spreadsheet if an assumption changed. With the electronic spreadsheet, they could let their imaginations run wild. They could just set an interest rate or a sales figure or cost savings or whatever and have the assumption ripple through a spreadsheet until the analysts got the sort of result from a potential merger or breakup that they wanted. Their bosses would then sell the idea to companies or aggressive investors — and watch the fees roll in.
Chunka, boiling down his thoughts on the A, B, C processes, suggests these seven questions that you should ask yourself to make sure you get the full benefit from generative AI:
- Are we using AI only to do the same work faster, or are we also using it to design entirely new ways of working?
- What systems and processes do we have to spread AI-driven learning and improvement across the organization?
- How are we actively identifying and challenging the assumptions baked into our current workflows, products, and business models?
- Where could AI help us fundamentally reimagine our business model—not just optimize existing operations?
- Who is accountable for leading and sustaining C-level improvement—and do they have the authority and resources to act?
- How are we ensuring that AI adoption does not quietly encode and scale harmful biases, flawed assumptions, or misleading correlations?
- Do we have the culture, skills, and adaptability to continually improve how we improve?
He writes, "The real prize is using AI to redesign the road itself—not just drive faster on the old one."
Cheers,
Paul
P.S. I've told my Engelbart story before, so I'll just reprise it briefly here.
In the late '90s, I attended a cocktail party at a friend's house in Silicon Valley and struck up a conversation with an older man, who expressed interest when I told him I edited a magazine for Diamond Management & Technology Consultants that focused on innovation through digital strategy. When he asked for an example of the sort of article I published, I told him I had just edited a piece on A, B and C processes.
"But that's my idea," he said.
"That's Doug Engelbart's idea," I replied.
"And I'm Doug Engelbart," he said.
He was, too. Engelbart, one of the most celebrated of the pioneers of personal computing, lived next door to my friend.