Ask a UAE business owner what their office costs and they'll answer to the dirham. Ask what their software costs and the answer is usually a guess — because no one signed one big contract. They signed fifteen small ones, in different months, on different cards, each individually reasonable. That's the design. Subscription software is priced to be too small to question and too embedded to cancel.
This note does the arithmetic most companies never do: what a typical mid-size operating stack actually costs, what that spend produces, and what changes when the same money is pointed at systems you own.
The anatomy of a rented stack
Here is a conservative, mid-market picture of what a UAE company of roughly 20–80 people rents each month. Your labels will differ; the shape rarely does.
| What you rent | Why it crept in | Typical monthly |
|---|---|---|
| Website platform | Fast to launch, never left | $ 349 |
| CRM & pipeline | Per-seat pricing that grew with you | $ 780 |
| Automation tools | Gluing the other tools together | $ 420 |
| AI subscriptions | Per seat, per model, per feature | $ 560 |
| Analytics & BI | The tier that unlocks exports | $ 390 |
| Internal tools | Project boards, docs, wikis, forms | $ 615 |
| Communication systems | Chat, calls, helpdesk seats | $ 240 |
| Integrations & glue | Middleware fixing the fragmentation | $ 330 |
| Total | $ 3,684 /mo |
Monthly, that's an unremarkable line item — about the cost of one junior hire. Multiply honestly: $44,208 a year. $221,040 over five years — roughly AED 812,000. And that's the static picture, before the two forces that make renting worse over time.
The two forces that make it worse every year
Price ratchets. SaaS pricing moves in one direction. Per-seat costs rise with headcount, tiers get restructured, features you rely on migrate to higher plans, and AI features arrive as paid add-ons. A stack that costs $3,700 a month today reliably costs more next year — for the same work.
Switching costs deepen. Every month, more of your data, workflow, and staff habit is shaped around each vendor. The price of leaving grows quietly alongside the price of staying. This is why businesses keep paying for tools they complain about: by year three, the exit costs more than the annoyance.
The subscription model's real product isn't software. It's the impossibility of leaving.
What the spend produces: a balance-sheet view
Now the uncomfortable question: after five years and $221,040, what does the business own? The answer is precisely nothing. No code, no infrastructure, and — practically speaking — not even the data, which lives in vendor schemas and comes out only through whatever export the plan allows. The entire spend is operating expense. Equity built: zero. Stop paying and the operation switches off — the last invoice buys the same access as the first.
Compare that with the same amount invested once in owned infrastructure: a system built around your actual processes, holding your data in one place, running on infrastructure you control. Spend of that shape ends. What it buys remains — an asset that compounds instead of a bill that renews.
When renting is still right
Honesty matters here, because "cancel everything" is bad advice. Renting is correct when you're early and your processes are still changing weekly; when the need is a commodity (email hosting, accounting software, video calls) where vendors genuinely do it better than a custom build ever could; and when a tool serves a temporary project. The problem is not subscriptions. The problem is a business whose core operations — the way it sells, delivers, and decides — run permanently on infrastructure it can never own, shape, or keep. We've written a full framework on this: custom software vs. SaaS — when owning beats renting.
What ownership changes, concretely
When Corehold replaces a rented stack, the change isn't cosmetic. Costs transform from a permanent rent into a bounded investment. Data consolidates from a dozen vendor silos into one structure that feeds every part of the operation. The software fits the business instead of the business bending around the software — the workarounds your team quietly maintains simply stop existing. And direction returns: the system evolves when your strategy changes, not when a vendor's pricing does. At handover, the client holds the system, the code, the data, the infrastructure, and the roadmap — outright.
That is the whole argument, and it fits in one sentence: software you rent is a cost; infrastructure you own is a position.