What Ramp Got Right That Most Companies Will Get Wrong
Ramp just published their AI adoption numbers. 99.5% team usage. 1,500 apps shipped in six weeks. Non-engineers committing code to production. Every founder reading that wants to know how to replicate it. Most will get it wrong. Not because the playbook is bad, but because they will copy the visible layer instead of the foundation that made it work.
The Part Everyone Will Copy
Leaderboards. Hackathons. Unlimited AI budgets. Slack channels where people share what they built. All Hands demos. Hiring screens that require AI proficiency.
That is the visible layer. It is also the last layer. It is what works after everything underneath it is already in place.
The Part That Actually Made It Work
Before any of that mattered, Ramp built the foundation.
They connected 30+ tools through a single authentication layer. Salesforce, Snowflake, Gong, Slack, Notion, Google Workspace, Figma. All live on day one. No tickets. No setup guides. No waiting.
They built Glass, their own AI platform, on top of that connected infrastructure. One install. One login. Everything works.
They built a skills marketplace where anyone could package a workflow and share it. 350+ skills. Every skill raised the floor for everyone else.
They built the data plumbing that made all of it possible. Unified systems. Clean connections. A single harness.
Then the culture worked.
Technology 3.0 and Why the Sequence Matters
At Dreaming Tree AI we use a framework called Technology 3.0. Three layers. Each one depends on the one below it.
Layer 1: Data Foundation. Where your data lives, how your tools connect, whether your systems talk to each other. The infrastructure nobody sees.
Layer 2: Intelligence and Automation. The workflows, the agents, the synthesis layer that eliminates manual work. Only functions when Layer 1 is solid.
Layer 3: Human Judgment. Culture, leadership, strategy, the decisions only people can make. Only reachable when Layers 1 and 2 are working.
Ramp built in sequence. Layer 1 first (connected tools, unified data, single authentication). Layer 2 next (Glass, skills marketplace, coding agents). Layer 3 last (hackathons, leaderboards, hiring changes, leadership modeling).
Most companies will read Ramp's playbook and start at Layer 3.
What Happens When You Skip the Foundation
Ramp says it directly in their own post. Before Glass existed, they had 90%+ adoption of AI tools across the company. Most people were stuck on a basic chat interface. The tools were good enough. The harness was not.
That is the condition most companies are in right now. They have given people access to AI. They might have run training. They might have encouraged usage. Nothing has fundamentally changed.
Ramp also tried the decentralized approach first. Every team building their own things. The result was redundant re-learning. They fixed it by building centrally and letting teams build on top.
They also say something that should be uncomfortable for anyone selling AI training: "Training doesn't work." The product taught people faster than any program could. But the product only worked because the foundation underneath it was already built.
The Diagnostic Question
If you are a founder or CEO reading Ramp's playbook and asking how to replicate it, the question is not which cultural initiative to launch first.
The question is: which layer are you actually ready for?
Do you know what tools your team is using today? Do those tools connect to each other? Does your data live in one place or twelve? If you pointed an AI agent at your business right now, would it find a coherent foundation or a collection of disconnected systems?
If the foundation is not there, no amount of hackathons or leaderboards or unlimited budgets will produce what Ramp produced. You will get activity without impact. Usage without compounding. The appearance of transformation without the infrastructure to sustain it.
Start With the Layer That Actually Matters
Ramp did not start with a better strategy. They say so themselves. They started with a culture that rewarded speed and a willingness to build infrastructure before building culture programs on top of it.
The lesson is not to copy the leaderboard. The lesson is to build the foundation that makes the leaderboard mean something.
That is what Technology 3.0 is about. Get the layers right. Get them in order. Everything else follows.
Sanjay Bhutiani is the founder of Dreaming Tree AI, a Technology 3.0 advisory practice helping founders and growing businesses build the technology foundation before the window to build it closes.