03 · Vendor decision
Should you hire an agency, a freelancer, or build in-house?
It depends mainly on how long the product needs sustained investment and how much risk you can absorb if one person becomes unavailable. A freelancer is the cheapest way to start and the most exposed to that risk. An agency costs more per hour but delivers design, engineering, launch and marketing as one accountable team, with redundancy if someone's unavailable. In-house costs the most up front and pays off only once a product needs continuous, indefinite investment.
This decision gets framed as being about cost, and cost is part of it — but the variable that actually predicts regret afterward is continuity risk: what happens to your project if the person building it becomes unavailable, loses interest, or was never able to cover the full range of skills a shipped product needs.
None of the three models is universally right. Each is right for a specific shape of project, and the honest version of this page is helping you recognise which shape yours is.
Side by side
| Model | Agency | Freelancer | In-house |
|---|---|---|---|
| Cost structure | Team rate, but covers design, engineering, launch and marketing under one roof — no separate hires to coordinate. | Cheapest hourly rate, but usually one skill — you're likely coordinating a designer, a developer and a marketer separately yourself. | Highest fixed cost — salaries, benefits and management overhead exist whether or not there's work to do. |
| Speed to start | Fast — an existing team with an existing process. | Fastest — one person, one conversation. | Slowest — recruiting, hiring and onboarding before any work begins. |
| Skill breadth | Design, engineering, launch and marketing as one accountable unit. | Typically one specialism; broader scope means coordinating multiple freelancers yourself. | As broad as the team you build and retain — entirely dependent on your hiring. |
| Continuity risk | Low — a team has redundancy if one person is unavailable. | High — the project depends on one person's availability, health and continued interest. | Low once established, but real during hiring, notice periods and turnover. |
| Best fit | A product that needs to go from idea to shipped, across multiple disciplines, without you managing every handover. | A small, well-defined task with a clear spec and a short timeline. | A mature product with enough ongoing work to justify a permanent team, indefinitely. |
- The product needs more than one discipline — design, engineering, launch, marketing — delivered as one accountable process.
- You want redundancy: the project doesn't stall if one person is unavailable.
- You'd rather not personally coordinate a designer, a developer and a marketer as three separate relationships.
- You're going from idea to a launched, maintained product, not a single bounded task.
- The task is small, well-specified and short — a freelancer's natural strength.
- You already have most disciplines covered and need one specific gap filled.
- The product has reached a scale where continuous, indefinite investment justifies a permanent in-house team.
- You have the internal capacity to manage hiring, onboarding and retention yourself.
Buyer education
Who owns the code when the work is done?
This is one of the least-answered questions in the industry, and it belongs in this comparison because it changes with each model. Standard, defensible practice — regardless of who you hire — is that custom work product transfers to the client on final payment, in full, including source code and assets. That should be an explicit clause in any contract, not an assumption.
- With an agency — IP transfer is normally a standard contract term — ask for it in writing before work starts, not after.
- With a freelancer — The same principle applies, but enforcement is harder if the relationship ends badly — get it in writing regardless.
- In-house — Ownership is rarely in question, since the people writing the code are your employees — the trade-off shows up elsewhere, in cost and hiring risk instead.
Questions
About this decision.
Do I own the code after an agency finishes the project?
Standard, defensible practice is that custom work product — including source code — transfers to the client in full upon final payment. This should be an explicit written term in any contract, with any vendor, not an assumption based on who you hired.
Is a freelancer always cheaper than an agency?
Per hour, usually yes. Once you account for coordinating multiple specialists yourself, absorbing the risk of one person being unavailable, and covering skills a single freelancer doesn't have, the total cost gap narrows — sometimes it reverses.
When does an in-house team make more sense than either?
Once a product needs continuous, indefinite investment at a scale that justifies full-time salaries and the management overhead that comes with them — a threshold most products haven't reached at launch.
Can you start with an agency and move a product in-house later?
Yes — that's a common and reasonable path, especially if code ownership was handled properly from the start, per the point above.
Want a second opinion on your specific case?
Send the details — this decision gets easier once there's a real product to weigh it against.