We build custom internal tools, so read this as a comparison written by an interested party and check it against your own numbers. That said, we tell plenty of people to use Retool. It is a good product, and for a lot of back-office work it is the cheaper answer by a wide margin. The question is not which one is better. It is which one stops fitting first, and what that moment costs you when it arrives.
If most of these describe your situation, buy the subscription and stop reading.
Per-seat pricing is fine at eight people and painful at eighty. Below about twenty internal users the licence rarely decides anything.
Retool is a front end over a database or an API. If those exist and are clean, most of the hard work is already behind you.
Standard CRUD, approvals with one or two steps, dashboards over existing tables. Nothing that needs logic a query builder cannot express.
A working internal panel in days is a real advantage, and it is an advantage no custom build can honestly match.
These are the four we see most often, usually eighteen months in rather than at the start.
Every warehouse picker and dispatcher who needs read access is another seat, and the bill grows with the part of the business you least want to slow down.
Tonnage-banded pricing, scheduling that respects vehicle capacity, an approval chain that branches. You end up writing code inside a UI that was built to avoid it.
Role models inside a builder are usually good enough internally and awkward to evidence when a customer or a regulator asks who could see what, and when.
Scale indicators, label printers, overnight imports, retries and queues. This is where a hosted builder runs out of room quickly.
Retool costs a monthly fee that scales with people and never stops. A custom tool costs more up front and then mostly stops costing, apart from support and the occasional change. Somewhere between two and four years the two lines cross, and where they cross depends almost entirely on your seat count rather than on how complicated the tool is. If you are under twenty internal users, the crossover is far enough out that buying is the rational call. If you are heading for eighty, it is close enough to matter now. The part people forget is the third line: the cost of the thing you cannot build at all, which does not show up on either bill.
Write down the three rules your operation actually runs on. Not the workflow, the rules: how a price is calculated, who is allowed to override it, what happens when a job is cancelled after dispatch. Then ask whether a query builder and a form designer can express all three without someone writing JavaScript into a text box. If they can, buy Retool. If one of the three needs real code, you are going to end up building that part anyway, and having half your system in a builder and half in a repository is the worst of both.
Regularly, and it costs us the project when we do. If your seat count is small, your rules are ordinary and the data is already in a decent database, a custom build is a worse use of your money. We would rather say that on the first call than discover it together six weeks in.
Yes, and it is often the sensible shape. Keep the quick internal panels in Retool, build the one system the business actually runs on. The thing to avoid is splitting a single workflow across both, because then nobody can say where the rules live.
Usually they keep running until the replacement is proven, then get switched off one at a time. We treat the existing apps as the specification, because somebody has already worked out what the screens need to do and that is genuine value worth carrying across.
One price for the build, agreed at discovery and written down before you sign, with a first production version typically four to eight weeks out. A single integration, or replacing one Retool app rather than all of them, is a much smaller job on both counts. The number moves with how many roles need their own permissions.
The same shape, slightly different numbers. Builders like Airtable and Softr are cheaper to start and hit their ceiling sooner, because they are further from the database. Retool sits closer to your data, so it survives longer before the rules outgrow it.
Describe what you’re building, what it has to do and when it needs to be live. You’ll get back the architecture we’d use, the trade-offs, and a realistic timeline, before anything is signed.