Choosing a vendorOutsourcingMVPContracts
How to hire a developer for a startup when you are not technical
9 min read
To hire a developer for a startup without being technical, define the outcome before the role, source candidates through referrals and vetted channels, judge them on shipped work and a short paid trial rather than an interview you cannot assess, and sign a contract that assigns the code to your company and keeps every account in its name. The order matters more than the channel.
Most hiring guides start with job boards. That is the wrong end. A non-technical founder who starts searching before knowing what they need ends up comparing profiles on charisma and price.
Step 1: write the outcome, not the job title
"Developer" is not a specification. Before you look for anyone, write half a page that answers four questions:
- What must exist in three months? A clickable prototype for investor meetings, a first version that ten pilot customers can use, a stable product that a growing team can extend.
- Who will use it, and on what? A web app for office staff, a mobile app for people in the field, an internal back office.
- What happens after delivery? Someone has to host it, fix it and improve it. Decide now whether that is the same person or company.
- What is fixed? Budget ceiling, a launch date tied to a trade show or a funding round, a regulation you must meet.
This half page changes who you should hire. A three-month prototype for fundraising is a project, and project-shaped work suits a freelancer or a studio on a fixed scope. A product that will change every week for years is a job, and a job eventually needs someone on your payroll or a cofounder. Our guide on finding a technical cofounder covers that second path.
Step 2: where to find candidates
Each channel produces a different kind of candidate, and the vetting effort it leaves to you differs too.
| Channel | Best for | Vetting left to you |
|---|---|---|
| Founder and investor referrals | Any role | Low to medium |
| Freelance marketplaces | Small, well-defined tasks | High |
| Vetted talent networks | Senior freelancers | Medium |
| Agency directories with reviews | Full projects | Medium |
| Local meetups and communities | Future hires, cofounders | High |
Referrals win because someone has already watched the person deliver. Ask founders who are one or two years ahead of you who built their first version, and whether they would hire that person again. The second question is the useful one.
Marketplaces are fine for a bounded task (a landing page, an integration, a bug fix) and risky for a whole product, because you inherit all the screening work. Directories that publish verified client reviews help you shortlist studios and agencies, but read the reviews for how problems were handled, not for the star rating. Our comparison of agency, freelancer and no-code explains what each kind of supplier is good at.
Step 3: vet without reading code
You cannot judge a technical interview, so do not run one. Collect evidence you can read instead.
Ask for something live. A product that real people use, with a short story: what the person built, what broke after launch, what they changed. Anyone can show a polished portfolio; fewer can describe a production incident calmly.
Ask the handover question. "If you stopped working with us next month, what would your replacement find?" A good answer mentions documentation, automated tests, a deployment process someone else can run, and accounts owned by your company. A vague answer is your answer.
Call two references, and pick them yourself. Ask the candidate for past clients who were not technical. Ask those clients whether deadlines held and whether bad news arrived early or late.
Pay for one hour of outside review. A senior engineer you trust, or a fractional CTO, can read a sample of the candidate's code and your trial results and tell you whether they are sound. It is the cheapest insurance in the whole process.
Red flags worth acting on: refusing to work in repositories and hosting accounts that your company owns, quoting a price before asking about your users, no mention of testing anywhere, and pressure to sign quickly.
Step 4: run a paid trial project
A paid trial replaces the interview you cannot run. It works when it is designed properly:
- Pick a real task from your plan, not a puzzle. Something like "users can sign up, confirm their email and see an empty dashboard".
- Write the expected result in plain language, including what you will click to check it.
- Fix the price and the deadline before starting. A few days to two weeks keeps the stakes low on both sides.
- Require your infrastructure: the code goes into a repository your company owns, deployed to an account in your company's name.
- Review three things at the end: does it do what was written, how did the person communicate during the work, and what does your outside reviewer think of the code?
Never ask for free spec work. Strong developers decline it, and the ones who accept are telling you how they value their time.
Step 5: the contract
A short agreement signed before the trial is enough to start. It must cover at least these points:
| Clause | Why it matters |
|---|---|
| IP assignment | Contractor code may not be yours without it |
| Accounts and access | Your company holds domain, hosting, repository |
| Deliverables and acceptance | Defines "done" before payment |
| Payment schedule | Milestones tied to accepted deliveries |
| Confidentiality | Protects customer data and plans |
| Termination and handover | What you receive if it ends early |
The IP clause is the one founders skip and investors check. In many countries, code written by an independent contractor belongs to its author unless rights are assigned in writing; employees are treated differently, and the rules vary by jurisdiction. Have a local startup lawyer review the template once, then reuse it.
Be careful with classification too. Treating a full-time, exclusive, long-term contractor as a freelancer can create employment law exposure in several countries. If the relationship looks like a job, it may legally be one.
When you work with a studio instead of an individual, the scope of rights transferred depends on the commercial model and is agreed case by case in the contract. Read that clause before you compare prices.
Step 6: the first month
Hiring does not end at signature. Schedule a short weekly demo where you see working software, not a status report. Keep a single written list of priorities that you own. Test every delivery yourself on the journeys your customers will use; our guide on getting an app built without a tech team shows how to do that without technical skills. If two demos in a row show nothing you can click, raise it immediately.
A decision grid
| Your situation | Usually the best first hire |
|---|---|
| Idea stage, need proof for customers or investors | Studio or freelancer on a fixed scope |
| Paying customers, weekly change requests | Employee developer or long-term studio |
| Technology is your competitive edge | Technical cofounder |
| Little time to manage, unclear scope | Prototype first, then a studio |
Day rates vary widely between countries and seniority levels; our article on software developer day rates explains why a lower rate does not guarantee a cheaper project. If you were hoping to pay in shares rather than cash, read app developer for equity before making an offer.
If you would rather not hire at all
We are a software studio, not a staffing agency: we do not place developers or sell hours. We start with a free visual prototype, presented over video within 5 working days, which remains ours: it is shown, not delivered. If you go ahead, we build on a fixed price, on a subscription where you do not pay for development, or as SaaS that we publish and run (details on our offers page). Under the subscription model LPR Studio remains the publisher of the application, which is worth discussing with future investors if the software is your core asset. If what you need is a person on your team for years, hire one: that is a different purchase.