The Ripple, Not the Splash: Where AI's Next Category Leaders Are Forming
Most of the AI conversation is about the splash. The lasting companies will be built in the ripples, and they’ll win on go-to-market as much as on technology.
I’ve spent about twenty years in Silicon Valley, most of it on the go-to-market side, in the room when a founder tries to convince an enterprise buyer that something new is worth changing how they work. That seat teaches you something the headlines miss: technology doesn’t create value when it’s invented. It creates value when someone reorganizes a business around it, and the customer can see the difference in their own numbers.
I saw this at Andreessen Horowitz, working with founders and enterprises directly. The companies that broke out rarely had the most impressive technology on its own. They understood a customer’s workflow deeply enough to rebuild it, and they could tell a simple story about the result: what changes, who benefits, and how you’ll know it worked.
The splash and the ripples
Today’s AI story is dominated by the splash: which lab has the best model, who raised the biggest round, who bought the most GPUs. Those stories matter, but history says they won’t be where most of the long-term value lands.
The iPhone was the splash. The ripples were Uber, Instagram and DoorDash, companies that never built a phone but couldn’t exist without one in every pocket. Cloud worked the same way. AI is following the pattern, and many of the defining companies of the next decade are being founded right now, mostly at seed and Series A, mostly outside the headlines.
Why the value moves to the ripples
The cost of a given level of AI capability keeps falling fast. When a powerful input gets cheaper, the value flows to whoever combines it with what stays scarce: ownership of a critical workflow, proprietary data, distribution, regulatory know-how and, above all, customer trust. Menlo Ventures’ 2026 State of Enterprise AI report found that enterprises now spend about $1.70 putting AI to work for every dollar spent building models. The money is already moving toward use.
Almost nothing on that list of scarce things is technology. It’s go-to-market.
The technology is ready. The go-to-market isn’t.
Enterprise leaders tell me the same thing again and again. They believe in AI and they have a budget for it, but they’re frustrated by how little has reached day-to-day operations. Madrona’s research on the ROI of AI found that most enterprises converted fewer than half their AI pilots to production last year. The reason is rarely that the model didn’t work. The tool didn’t fit how the work gets done, nobody owned the rollout, security review stalled it, or no one could show the CFO what changed.
Those are go-to-market problems, and they’re exactly where the next great companies will separate themselves. When I meet founders building in the ripples, I look for four things:
They start from the workflow, not the model. They can name the process they’re replacing, who owns it, and what it costs. Models change every six months. Deep knowledge of the workflow compounds.
They sell outcomes, not software. Buyers are tired of paying for unused seats. The best founders tie revenue to results: tickets resolved, claims processed, hours returned. That signals the thing buyers care about most, confidence that the product works.
They build for production on day one. A pilot is not a customer. Strong teams plan for integration, security and change management before the first demo. The real sale isn’t the signature. It’s the moment the old process gets turned off.
They prove value again and again. In AI, a better option is always one demo away. Companies that last build measurement into the product, so the customer sees the return every quarter.
A company that does all four isn’t selling software. It’s selling a better-run business, and that’s the strongest GTM story there is.
Swap the motor, or redesign the factory
Early in my career, at PwC and BearingPoint, I worked inside global companies and learned a lesson that still holds: large organizations don’t change because a new tool exists. They change when someone shows them a better way to run the process and makes the switch feel safe.
When factories first adopted electricity, many simply swapped the steam engine for one large electric motor and saw little change. The real gains came years later, when companies redesigned the factory floor around what electricity made possible.
Most enterprises are still swapping the motor with AI, adding a copilot to an old process. That phase favors incumbents, who already own the customer and can bolt AI onto what they sell. The redesign is where new companies break through, by showing up with the redesigned factory, a clear way to measure it, and a team that stays until it’s running.
That’s the opportunity we’re backing at Lightning. The next generation of great companies won’t win because their model is better. They’ll win because they understand the work, deliver the outcome, and earn the trust to change how a business runs.
In Part 2, we’ll look at the five areas where we see these ripple effects forming, the portfolio companies leading the way, and why the shift from pilots to production makes this moment critical.
Source links
This post is for informational purposes only and reflects the author's views and opinions. It is not an offer to sell or a solicitation of an offer to buy any security or interest in any fund, and it is not investment advice. Lightning Capital and its affiliates may hold positions in companies or sectors discussed. Forward-looking statements are subject to risks and uncertainties. Venture investing involves a high degree of risk, including loss of capital.
Comments