Once you're confident you're solving the right problem, which working practices help you build the solution well ("make the product right")?

Involve stakeholders (especially future users), work iteratively with prototypes, "kill your darlings", and test early and often.

These are habits for the execution phase — building the chosen solution efficiently and without falling in love with a bad idea:

• Involve stakeholders, especially future users — the people who'll live with the solution should shape it, not just receive it.
• Work agile / iteratively — proceed in short cycles and adjust as you learn, rather than betting everything on one big upfront plan.
• Work with prototypes — cheap, throwaway versions let you learn before committing real resources.
• "Kill your darling" — be willing to drop an idea you're attached to when evidence says it isn't working; emotional investment is not a reason to keep something.
• Test early, test often — find problems while they're still cheap to fix, not after launch.

The unifying idea: stay humble and empirical. You don't know you've built the right thing until real users and tests confirm it, so build in small, testable steps and let evidence — not attachment — decide what survives.