Walk through the P&L of almost any mid-market company and you will find the same line item growing faster than revenue: software subscriptions. Forty rented tools, per-seat pricing that penalizes headcount growth, and renewal increases arriving every year. Each subscription looked reasonable when it was added. The sum is a leaky bucket that gets renewed every month.
If you own or operate one of these companies, or you sit in the COO, CFO, or CIO seat, you have probably had the uncomfortable version of this conversation already. Revenue has flattened. The software line has not. And nobody on the vendor side is accountable for either problem.
There is a name for the fix. We call it unSaaSing: replacing rented software with infrastructure you own. This article lays out why the shift is happening now, what it changes for the three problems above, and how to evaluate it for your own stack without committing to anything.
The Growth Tax Nobody Budgets For
SaaS pricing is designed to scale with your success. Add a salesperson, pay another seat. Add a location, pay another instance. Grow your contact list, move up a pricing tier. The result is a cost structure that behaves like a tax on growth: the better you perform, the more you pay for the same functionality.
That was a tolerable trade when building software in-house required a development team and a multi-year budget. It is a much worse trade today. Mature open-source cores now exist for every layer of the business stack, from ERP to CRM to marketing automation to workflow orchestration, and modern AI-assisted development has collapsed the cost and timeline of standing them up. Owned infrastructure now costs a fraction of annual SaaS rent to build, and the payback math gets better every year the subscriptions would have renewed.
Your Data, Their API
The bill is only half the problem. The more strategic issue is that the customer records, transaction history, and pipeline data inside those rented tools do not really belong to you in any practical sense. They sit behind vendor APIs, rate-limited, fee-gated, and formatted for the vendor's convenience rather than yours.
This is the quiet reason so many mid-market AI initiatives stall. Every serious AI use case, from demand forecasting to automated follow-up to margin analysis, starts with the same requirement: fast, unrestricted access to your own operating data. When that data lives in eleven different vendor silos, each with its own API limits and export fees, the AI strategy dies in the integration phase. Companies are effectively paying to rent access to their own information, and then paying again in lost capability.
Where the Revenue Plateau Comes From
Here is the connection most operators miss: the same rented stack that drains margin is usually leaking demand. Inquiries wait days for a reply because they land in a tool nobody checks. Quotes never get a second touch because follow-up depends on someone remembering. Past customers go quiet and stay quiet because reactivation is nobody's job and no single system can even see the full customer history.
None of this is a talent problem. It is what happens when demand generation depends on humans remembering to do repetitive work across a dozen disconnected tools. When revenue plateaus while software spend keeps climbing, this gap between the demand a company generates and the demand it actually captures is usually the largest single cause.
Owned infrastructure changes this because automation finally has somewhere to live. AI agents and automated workflows can answer the inquiry in minutes, send every follow-up on schedule, reactivate the customers you already paid to acquire, and hand your team only the conversations that need a human. That is difficult to build across rented silos and straightforward to build on a stack you control.
What UnSaaSing Actually Means
UnSaaSing is not a purge of every subscription, and it is not building custom software for its own sake. It is a layer-by-layer evaluation with a simple test: for each tool, is renting still the right trade, or has this become a layer you should own?
In practice, the shift has three components:
- Replace rented cores with owned ones. ERP, CRM, marketing automation, and workflow tooling all have mature open-source foundations that can be deployed, customized, and secured as company property. No per-seat pricing, no renewal letter.
- Consolidate data into infrastructure you control. Once operating data lives in systems you own, every AI use case gets cheaper and faster, because the integration barrier is gone.
- Put automation to work on demand capture. With the stack unified, the highest-payback automations are usually the unglamorous ones: speed to lead, follow-up coverage, and reactivation of past customers.
Some layers should stay rented. Commodity utilities with low data gravity and honest pricing are often fine as subscriptions. The point is that this should be a deliberate portfolio decision made with numbers, not the accumulated default of a decade of point purchases.
The Exit Table Argument
For owners thinking about an eventual sale, there is one more reason this matters. Buyers value what transfers. A subscription transfers nothing: the day after close, the acquirer is paying the same vendors for the same rented capability, and your years of fees bought no asset. Owned infrastructure transfers as property, along with the unified data and the automated demand engine running on top of it. Companies that unSaaS convert a permanent expense line into a balance-sheet asset that survives to the exit table.
How to Start: The Stack Audit
The wrong way to start is a big-bang migration. The right way is to get the numbers first. A Stack Audit inventories what you rent and what it actually costs at renewal, maps where your demand and revenue tools drop leads, and sequences the migration layer by layer, starting where payback is fastest. You see the full picture, in dollars, before committing to anything.
If your revenue has gone flat while your software line keeps growing, or you know an operator whose P&L fits that description, that audit is the honest first step. Worst case, you end up with a complete inventory of your stack and its true cost, which most mid-market companies have never actually seen.