I run customer success at Lovable. We make software that lets anyone - a marketer, an operations lead, a realtor - describe an application in plain language and get a working product. No code. Our customers don't buy seats. They buy credits, and they burn credits only when they build. That's consumption pricing, and it rewires the entire post-sales job: if customers stop getting value, they stop spending, immediately, mid-contract. There's no annual renewal cliff to hide behind. Value realization isn't a slide in my QBR deck. It's the revenue model.
We became the fastest company in history to reach $400 million in revenue under that model. Here's what it taught me about the transition every person in this room is living through.
Most CS orgs are still running the old motion against a new scoreboard. Adoption dashboards. Health scores built on logins. Quarterly check-ins. All of it measures the same thing: usage. And usage tells you they tried you. It doesn't tell you they need you. Those are different businesses, and only one of them survives a budget review. So we changed the question. Not "are they using it?" but "if we disappeared tomorrow, does something break?"
We call that load-bearing. Borrow the image from construction: a load-bearing wall is the one you can't knock out without the building coming down. A load-bearing app is software a customer built on our platform that their business now runs on — a tool their team opens every morning, a workflow with no manual backup, a system other systems depend on. Not a demo. Not a prototype someone showed off once. Infrastructure.
Here's the uncomfortable number: less than one percent of everything our customers build clears that bar. I tied my entire team's goals to that one percent anyway. It was the loneliest metric decision I've made, and the best one. Because load-bearing is the single leading indicator of both retention and expansion. Usage means they tried you. Load-bearing means they can't leave you.
That decision forced everything downstream to change.
You can't nurture your way to load-bearing. Relationship motions — lunches, check-ins, rapport — produce goodwill, and goodwill doesn't survive procurement. So we run deployment motions instead: designed, compressed working sessions where the customer ships something real before they leave the room. One example. eXp Realty, one of the largest real estate brokerages in the world. One structured build day, C-suite in the room. They walked out with 27 working internal products and over a million dollars in measurable value — and stood up an internal app store to distribute what they'd built to 10,000-plus realtors. That account doesn't need a save play at renewal. The renewal argument is running in production.
It also changed who we hire. Forward-deployed engineers — technical builders who sit with customers and construct alongside them — now work inside my post-sales org. Not because CSMs failed. Because the definition of success moved from sentiment to output, and output needs people who can make things. The CSM role isn't dying; it's splitting. The half that advises — what to build, why, what it's worth to the business — is more valuable than ever. The half that schedules meetings is being automated out from under us. I've never written a line of code. That's the point. Judgment is the part AI didn't compress. Judgment got more expensive.
So here's my honest read for this room: the playbook that got you hired will get you fired. Adoption, usage, nurture, relationship — that motion was built for seat-based contracts, and seat-based contracts are giving way to models where value and revenue are the same number. Leaders who wait for this to settle will manage decline. Leaders who move now inherit something better.
Because CS isn't being disrupted. It's being promoted. For a decade we asked for a seat at the revenue table. Consumption pricing handed us the whole table — and the accountability that comes with it.
"Possible" is dead as a differentiator. Every vendor can claim possible. Proof is the currency now.
