An honest guide from the other side of the table
After a decade of pitches and of inheriting projects from partners who over-promised we have strong opinions about how this decision should be made. Here's the checklist we'd use ourselves.
Signals worth trusting
They talk about your business before their stack. A partner who leads with technology choices before understanding your goals is selling inventory, not solving problems.
The people in the pitch are the people on the project. Ask directly: who exactly will work on this? Agencies that sell senior faces and staff junior benches are the industry's oldest trick.
They ship small and early. Ask what you'll see after three weeks. The right answer is working software, however modest. The wrong answer is a document.
They'll tell you 'no.' Ask about a project they turned down or descoped, and why. Partners who never push back are ones who bill for everything and warn about nothing.
Their references answer one question well: would you hire them again without a competitive process? Anything but an immediate yes is a no.
Red flags that cost real money
- Fixed-price quotes produced without discovery someone will pay for that guess, and it's usually you.
- Vague ownership terms. You should own all code, infrastructure, and accounts from day one, unambiguously.
- No opinion on maintainability. If they don't ask who will run the system in year three, they're building for handoff, not for you.
- Estimates dramatically lower than everyone else's. That's not efficiency; that's a change-order strategy.
The questions that separate finalists
- Walk me through a project that went wrong. What did you do? (You're listening for honesty and recovery, not perfection.)
- How will we see progress, and how often? (Weekly working demos, or it isn't real.)
- What will you need from us to succeed? (Good partners demand your time and decisions. Ones who say 'nothing, we've got it' are hiding the ball.)
- How do we exit this relationship cleanly if we need to? (Confident partners have a good answer, because clean exits are how they'd want to be treated.)
The bottom line
You're not buying code you're renting judgment. Optimize for a partner whose incentives, transparency, and habits you'd trust when things go sideways, because on any project worth doing, at some point they will.
Bhaibhav Baidhya
Founder & Director
Part of the Converge Solutions team writing about what we learn building technology for ambitious organizations.