Obelra Book a scope call
About

Opinions, stated up front.

Every software firm will tell you they're collaborative, transparent and quality-focused. Those aren't positions, they're table stakes. These are the actual convictions Obelra runs on — the ones that change what gets built and what it costs you.

The data model is the building. Everything else is finishing.

Almost every custom software project that goes badly went wrong in the first fortnight, when somebody decided how the information would be structured — and got it subtly wrong. Nobody notices for months. Then a reasonable request arrives, and the honest answer is "that would mean rebuilding a lot of this."

So we settle the structure before anyone draws a screen. It's the least visible week of the project and the one that determines what the system can still do three years from now.

Working software beats a status report.

A progress update is an assertion. A link you can click is evidence. Every week of a build you get the real thing, in the state it's actually in — including the parts that aren't finished.

This is uncomfortable early on and it's the entire point. Changing your mind in week two is a conversation. Changing your mind in week nine is a renegotiation.

Estimating is our job, so estimation risk is ours too.

Hourly billing quietly transfers the consequences of a bad estimate onto the client. You didn't scope the work and you can't verify the hours, but you pay for both.

Fixed price puts that where it belongs. If a build takes longer than we judged, that's ours to absorb. It also means we have to understand the work properly before quoting — which is why scoping comes first and why it's free.

Ownership is the default, not a concession.

A surprising amount of this industry depends on clients not quite owning what they paid for — the hosting is in the agency's account, the domain is on the agency's registrar, the code has never been seen by anyone else.

Every Obelra project is set up so you could replace us on a Tuesday and be entirely fine on Wednesday. Not because we expect that, but because a client who can leave and doesn't is a better relationship than one who's stuck.

Boring technology, chosen deliberately.

There's always a newer framework that's faster to write and harder to hire for. Building your business on it transfers a real cost to you later, when you need someone to maintain it and the pool of people who can is small and expensive.

We pick established, widely-supported tooling. The measure is simple: could a competent developer who has never met us pick this up and be productive? If not, it's the wrong choice regardless of how good it is.

The right recommendation is sometimes "don't build this."

Custom software earns its cost when the workflow is genuinely particular to your business. A great deal of work isn't — it's ordinary, and mature products already handle it well for a subscription fee.

You'll hear that on the call when it's true. It costs a project occasionally, and it's the reason the projects we do take on tend to be worth doing.

Fit

Who this works well for

A bad match wastes your money and our time, and it's usually obvious in advance. Here's the honest read on both sides.

Works well

  • You have a process that's genuinely specific to how your business operates.
  • Someone on your side can make decisions without convening a committee.
  • You can give it about an hour a week while it's being built.
  • You want to own the result outright rather than rent it indefinitely.
  • You'd rather hear an uncomfortable assessment early than a comfortable one late.

Works badly

  • You need it live next week. Good scoping alone takes a few days.
  • You're collecting bids and the lowest number wins. We won't be it.
  • The direction changes often and the budget is expected to stay still.
  • Nobody internally can say what the software is supposed to replace.
  • You want developers on tap by the hour rather than a defined project.
How we're set up

The practical shape of it

Where New Jersey, working remotely Based in New Jersey and working with clients across the US. On-site time is available where a project genuinely needs it — watching how work actually happens often changes the design.
Load A deliberately short list We run a small number of projects at once by choice. It's the only way weekly builds stay weekly. If there's no room, you'll be told on the first call rather than quietly queued.
After launch Available, never required Thirty days of fixes come with every project. Beyond that, support is there if you want it — a well-built system shouldn't need monthly intervention to keep running.

Recognise your situation in any of that?

Thirty minutes is usually enough to know whether there's a project here. You'll get a straight answer either way.

Book a scope call See the services