Once you've decided you need your own app, the next question is almost always the same: who builds it? Hire your own IT team, use a freelancer, or hand it to a software house? Each can succeed and each can fail. What makes the difference is how well it fits your company's size, the kind of app, and who will look after it once it's done.
Next stepHow we work: start with one team, first system running in 2–6 weeksSee how it worksThe three options at a glance
| Freelancer | Software house | In-house IT team | |
|---|---|---|---|
| Upfront cost | Lowest | Medium to high | High: salaries, hiring, equipment |
| Speed to start | Fast | Fast to medium | Slow, because you need to hire |
| Understanding of your business | Depends on the person | Depends on their experience in your industry | Deepest, after a few months |
| Biggest risk | The person disappears | The app isn't used in the field | One or two key people leave |
| Best for | Small tools, one function | Core apps with a clear deadline | Many apps that keep evolving |
Freelancers: cheap and fast, but fragile
A freelancer suits small tools with a clear scope: a form, a simple dashboard, or automating one report. It's the lowest cost, and work can start this week.
The risk is continuity. Company apps are rarely finished once: business rules change, bugs appear, new requests come in. If the freelancer gets busy with other projects or takes a job, the company is left with an app nobody understands.
If you use a freelancer, secure three things from day one: the source code is stored in an account the company owns, the server and domain are in the company's name, and there's simple documentation on how to run the app.
Software houses: a full team, with conditions
A software house brings a full team: analysts, designers, programmers, testers and a project manager. For a core app used by many people, that reduces the risk of one person being the only one who understands the system.
Software houses rarely fail on the technical side. More often, the app looks polished in the presentation but isn't used in the field, because the builders don't understand how a distributor works: how salesmen work, how tiered price lists are calculated, why import orders are worked out differently. That's why experience in your industry matters more than a glossy portfolio.
What the contract should include:
- the source code, data and server accounts belong to your company
- the scope broken down per stage, not one price for "all features"
- the yearly maintenance cost after the app goes live
- who to contact when something breaks, and how fast it gets handled
An in-house IT team: the deepest, the most expensive to start
Your own team understands your business best and responds fastest to change. For a company that will keep building many apps, it can be the most economical choice in the long run.
But the cost is often underestimated. The average software engineer salary in Indonesia is about Rp11 million a month according to Indeed data as of October 2026, higher in Jakarta, with senior engineers averaging over Rp21 million. A realistic small team for company apps usually needs at least:
- one senior person who can design the system and guard quality
- one or two programmers
- someone who tests and writes documentation, often doubling up
With salaries, BPJS, THR, equipment and hiring that can take months, a three-person team easily costs hundreds of millions of rupiah a year before the first app is finished. The biggest risk: if one key person leaves, the knowledge of the system leaves with them.
A fourth option: building it yourself with AI
Tools like Claude Code and Cursor now let people who understand the business build simple apps themselves. For prototypes and small internal tools, it's a very fast way to prove an idea. For an app used by dozens of people every day, you still need someone looking after security, data and maintenance.
A combined model: often the most sensible
For a mid-sized distributor, the model that works most often is a combination:
- A team that understands distribution builds the core of the app, so it's fast and fits how work is done in the field.
- One person inside becomes the app's owner. Usually someone from operations, not a programmer, who decides priorities and makes sure the app gets used.
- Maintenance and small changes are handled through a maintenance contract, or gradually moved to one or two in-house programmers once the app is stable.
This way you get the speed of a full team without building an IT department from scratch, and the knowledge stays in the company.
The most important question isn't who builds it, but who in your company owns it. Without an owner inside, an app from any builder will stop being used.
How to choose for your company
- Only need one small tool? A freelancer, or build it yourself with AI, as long as code ownership is clear.
- One to three core apps in a year? An outside team that understands your industry, with an app owner inside the company.
- Many apps that will keep evolving for years? Start with an outside team, then build an in-house team step by step.
Before you choose, make sure a custom app really is the answer, rather than an ERP module or ready-made software. The trade-offs are in ERP or a custom app for distributors. For more detailed numbers, read the cost of building a custom app for a distributor.
Where TAByte fits
A software house usually builds an app to order, then hands it over. We work a little differently: we start by mapping your process, build on apps already running in a real distributor's operations, and stay with you after the system is live. The first system is usually running within two to six weeks, starting with one department.
See how we work, or talk through your company's situation in a free consultation.