All posts

Tools

How to choose a hackathon problem worth a month of work

Turn a broad monthly theme into a specific user problem, a small testable promise and a realistic first prototype.

HackHoster Team · · 4 min read

Modular ivory prototype pieces and a notebook on a matte charcoal workbench
Illustration: HackHoster (AI-assisted), original artwork

At a glance

  • Choose a specific person and situation before deciding on features.
  • Test the hardest assumption with a small, observable experiment.
  • Define one useful output and a clear boundary for the project.
  • Use the monthly brief and judging criteria to check your scope.

Practical guidance from HackHoster. Sources and illustration credits accompany the article. Publishing standards and corrections.

A monthly theme gives you a direction, but it does not choose a problem for you. Start by naming one person, the situation they face and the result they need. A smaller promise that you can demonstrate is a stronger starting point than a long list of features connected only by a theme.

This guide uses an imaginary student-club organizer as a working example. It is a planning exercise, not a past HackHoster entry. The organizer has meeting notes scattered across documents and messages. The useful outcome is a reviewed action list with owners and dates, rather than another place to store all the notes.

Conceptual illustration of building a focused prototype from small components
Illustration: HackHoster (AI-assisted), original artwork

Describe the moment of difficulty

Write the problem as a short scene. After a club meeting, the organizer has twenty minutes to send follow-up tasks. They search three documents, copy names into a message and still miss a deadline. This scene tells you when the tool is used and what makes the current process frustrating.

Avoid starting with a technology label such as an AI productivity platform. That label says little about whether a person needs the result. Ask what they do now, where the process breaks and what happens if they make a mistake. If possible, ask someone who actually does the task to describe their last attempt.

Separate a problem from a preference

A person saying that a dashboard would be nice is a clue, not proof of a useful problem. Look for behavior you can observe. Do they repeatedly copy information? Do they maintain a workaround? Can they show you a confusing handoff? Those details help you decide what to test without claiming market demand.

For the organizer, a good question is whether the notes contain enough information to assign tasks at all. If owners are usually missing, automatic extraction alone cannot solve the problem. Your prototype might instead highlight unanswered questions and let the organizer resolve them before exporting a checklist.

Write one testable promise

Use a sentence with an input, an action and an output. For example: paste meeting notes, review suggested tasks, then export a checklist with unresolved details marked. The sentence should describe a journey you can show on screen. Keep the benefit modest enough to verify during the challenge.

Define success before coding. You might ask whether a new tester can locate every unresolved task and export the corrected list without help. This is a prototype check, not a claim about universal accuracy or time savings. Record what you actually observe instead of inventing a percentage for the presentation.

Choose an output you can inspect. A clear checklist, a corrected record or a completed export gives you something concrete to test.

Test the assumption that could stop the project

List three things that must be true for the idea to work. The notes need to contain relevant facts. The extraction needs to preserve their meaning. The user needs to understand the review step. Pick the least certain assumption and test it with a tiny example before building account settings or a polished homepage.

You can begin with a manual workflow. Prepare three synthetic note samples and sketch the output you would want for each. Ask a tester to explain what they think the output means. If the result is confusing on paper, extra interface polish will not resolve the underlying ambiguity.

Draw a firm boundary around the prototype

Write a short list of what the project will not handle this month. The club tool might accept pasted text, support English notes and export a simple file. It might leave calendar synchronization, multiple workspaces and mobile notifications for later. These are scope decisions, not defects you need to hide.

Keep a separate parking list for interesting ideas. When a new request appears, compare it with the original journey. Does it help someone complete the promised result, or does it create another product? Remove one commitment before adding another substantial one. A focused prototype is easier to test, explain and finish.

Check the fit before committing

Read the current challenge brief and judging criteria. Explain the connection in plain language. Under an information-to-action theme, the club tool converts scattered notes into reviewed next steps. That connection should be visible in the demonstration rather than added as a sentence at the end.

Write a one-page decision note with the user, problem, input, output, hardest assumption and scope boundary. Give the idea one final test: could a teammate build the same first version from that note? If not, clarify the promise. Then use the monthly project planning guide to turn that promise into a schedule.

A short exercise to try

Try a twenty-minute selection exercise with your team. Each person writes one problem scene and a possible output, without naming a framework. Compare the scenes by asking which one you can observe, which one has accessible sample input and which one you can demonstrate within your available sessions. Choose a candidate and write the reason. If you cannot choose, test one uncertain assumption from each rather than voting on the most exciting title. Keep the discarded notes; they may become useful later, but they are not commitments for this month.

Sources