Every decade in software has a dominant verb. The 2000s were about installing — software was a thing you bought and ran. The 2010s were about Entbündelung — every job got its own cloud tool, and the app store mentality won. The 2020s have a verb too, and it's the opposite of the last one. This is the consolidation decade, and it isn't a fashion. It's the bill for the decade before it coming due.
Why unbundling happened — and why it stopped.
Unbundling was rational while two things were true: software was hard to build, and capital was cheap. Hard-to-build meant a focused team could win a category by being meaningfully better at one job. Cheap capital meant customers could afford to assemble forty subscriptions and investors could afford to fund forty companies chasing forty categories.
Both reversed. Software got dramatically easier to build, so single-feature advantages evaporated — your clever tool is a weekend project for someone else now. And capital got expensive, so the customer who once happily paid for forty tools started counting them, and the investor stopped funding the forty-first. When the conditions that produced unbundling disappear, the unbundling does too.
Unbundling was a feature of cheap money and hard software. Both are gone. The strategy built on them is going with them.
What consolidation actually means.
Consolidation doesn't mean fewer companies, though that's happening. It means the unit of value shifts from the feature to the platform. Customers stop asking “what's the best tool for this one job” and start asking “what's the fewest number of systems I can run my business on.” The winning answer to the second question is almost never a point tool.
The number above is the engine of the decade. When more than a third of the bill is redundant, consolidation isn't a nice-to-have — it's the highest-ROI project most teams have available. And unlike the cost-cutting of past downturns, it doesn't reduce capability. Consolidating a stack usually increases what a team can do, because the pieces finally work together.
This is structural, not cyclical.
The easy dismissal is that this is just belt-tightening that reverses when money gets cheap again. It won't, because the second force — software getting easy to build — is permanent and accelerating. Even if capital floods back, a single-feature tool can't rebuild a durable moat in a world where features are commodities. The only durable position left is the integrated one: own the data, own the workflow, own the relationship across jobs.
We started building Mewayz on exactly this thesis: that the 2020s reward the platform that does many jobs coherently over the tool that does one job brilliantly. Five years in, the bet looks more right every quarter. The consolidation decade isn't coming. You're already in it.