The concrete questions to ask on the first call, what to check in the contract before signing, and the red flags worth catching early.
Confirm five things before signing: the code stays 100% yours, the price is fixed and closed before you start, they show real progress every 1-2 weeks, they have verifiable case studies running in production, and the contract clearly defines scope, warranty, and what happens if the project changes direction.
Bring these to the first call. How they answer matters as much as the answer itself.
Who owns the source code once the project is done?
Good answer: It must be explicit in writing that the code and intellectual property are 100% yours, with access to the repositories from day one.
Is the price fixed or hourly?
Good answer: A fixed price agreed before you start protects you from overruns. If they only offer an hourly estimate that 'may vary,' you're the one absorbing the risk of the project running long.
How often will I see real progress?
Good answer: A serious company delivers working software every 1-2 weeks. If they only show you the result at the end, there's no way to correct course in time.
Can I see real projects or case studies, not just a design portfolio?
Good answer: Ask to see systems actually running, not just mockups. If possible, talk to a past client.
What tech stack do they use and why?
Good answer: Modern, well-maintained technologies (React, React Native, Node.js, PostgreSQL, etc.) mean any other team can pick up the project later if you ever need to switch.
What does the warranty and post-launch support include?
Good answer: There should be a bug-fix warranty period at no extra cost, and a clear (priced) maintenance plan for after that.
Who is the actual team that will work on my project?
Good answer: You should be able to talk directly to whoever writes the code, not just a sales rep who later subcontracts the work.
What happens if I need to change scope mid-project?
Good answer: Scope changes are normal. There should be a clear process to quote them separately without stopping what's already in motion.
"We'll see how it goes" with no concrete number or range before starting.
They promise to deliver everything at the end, with no demos or intermediate reviews.
Or they avoid answering the question clearly.
Far below the market range for that type of project, with no explanation.
Timelines too short even to properly scope the project, let alone build it well.
They can't show a single real project running in production.
If they take days to reply during the sales process, it'll be worse during the project.
A freelancer works for something small and well-defined, but is a single point of failure. An in-house team gives maximum control but is expensive and slow to build. A software development company is the middle ground for most businesses: a full team from day one, without the fixed costs of an in-house tech department.
See the full comparison with a table and criteria →Who owns the source code once the project is done. It's the question with the biggest long-term impact: if the code doesn't stay 100% yours, you're locked into that vendor forever, unable to move to another team or maintain it in-house.
It depends on the size of the project. A freelancer works for something small and well-defined, but is a single point of failure. An in-house team gives maximum control but is expensive and slow to build. A software factory is the middle ground for most companies: a full team from day one, without the fixed costs of an in-house tech department.
Compare it against published market ranges (see our pricing guide) and pay more attention to what's included than to the number alone. A pricier quote that includes discovery, testing and training can end up cheaper in the final result than a lower one that skips them.
At minimum: code ownership, a closed price and scope, a timeline with partial deliveries, and what the post-delivery warranty covers. If any of these isn't clear in writing, ask for clarification before moving forward.
Not always, but understand why. If the scope genuinely can't be defined upfront (a research project, for example), it makes sense to start with a small, paid discovery phase. If the vendor simply won't commit to a price on a well-defined project, that's a warning sign.
Ask to see the system actually running, not just screenshots, and if possible talk to a past client about how the process went, not just the final result. A good vendor shouldn't have a problem arranging this.
Cotizamos con alcance cerrado en dólares. Sabés cuánto cuesta desde el día uno, sin sorpresas.
Al finalizar, el repositorio, la infraestructura y la documentación quedan completamente a tu nombre.
Trabajamos en horario de negocio de Sudamérica. Reuniones y respuestas sin delays por diferencia horaria.
Después de la primera reunión recibís una propuesta detallada por escrito, sin compromiso ni letra chica.

Damián Oliva
Fundador de Deepyze
Trabajás directo con quien construye. Años desarrollando productos digitales en LATAM, certificado en Google Cloud Machine Learning.
Conocer al fundadorBook a free call and evaluate us with this exact checklist. No obligation.