Good hackathon ideas solve a narrow, real problem that a small team can demo in a weekend, such as a tool for patients, tenants, students or neighbours. This guide shows where strong ideas come from and how to scope one.
The ideas below draw on the pandemic-era events covered in the covid-19 hackathon hub, plus global frameworks that give projects a clear purpose.
What makes a hackathon idea worth building
Most hackathon ideas fail for the same reason: they are too big for the time available, or they have no clear user. A strong idea has three traits. It names a specific person with a specific problem, it can be shown working in a short demo, and it does not depend on data or partners that the team cannot reach during the event.
Start with the user, not the technology. “A chatbot” is a technology. “A phone line that reads a clinic’s opening hours aloud to people without reliable internet” is an idea. The second version tells judges who benefits and how, and it points to a concrete first screen to build.
It also helps to separate the problem from the prototype. The problem can be large, such as isolation among older residents, while the prototype addresses one slice of it, such as a weekly check-in message that a volunteer can send in under a minute. Judges tend to reward a small thing that works over a grand plan that does not run.
- A named user: a tenant, a nurse, a shop owner, a student without a laptop.
- One core action: the single thing the prototype does well, shown end to end.
- Available inputs: open datasets, public interfaces or sample files, so nobody waits on permission.
- A believable next step: who could keep it going after the weekend ends.
Ideas borrowed from pandemic-response hackathons
Events built around a crisis are a rich source of ideas because the organisers had to state the problem areas in advance. A global online event on Devpost ran in March 2020 and, according to its Devpost event page, listed 18550 participants. Its organisers suggested seven themes to spark ideas: health, vulnerable populations, businesses, community, education, entertainment and an open “other” category.
Those themes translate well beyond a pandemic, and each can be turned into a buildable concept:
- Health: a symptom-diary tool that lets people record how they feel and share a summary with a clinician, or a scheduling helper for frontline workers.
- Vulnerable populations: a plain-language guide to local support services that works on a basic phone and in several languages.
- Businesses: an order-ahead page for a small shop that has no web presence, or a tool that helps a market stall move online.
- Community: a neighbour-matching board for errands, or a way for a town to publish service changes in one place.
- Education: low-bandwidth lesson packs for students who cannot stream video, or a homework tracker for teachers working remotely.
- Entertainment: a remote game night or virtual performance space that keeps audiences and performers connected safely.
The same event page notes that some challenge areas came from health partners, including the World Health Organization. Starting from a partner-stated need is a reliable habit: it keeps the project tied to a problem somebody has actually described.

Ideas anchored in the Sustainable Development Goals
Another way to find ideas is to start from a shared list of global priorities. The United Nations explains on its Sustainable Development Goals page that the 2030 Agenda, adopted by all member states in 2015, has at its heart the 17 Sustainable Development Goals. The goals cover poverty, health, education, inequality, cities, climate, oceans, land and peaceful, inclusive societies.
Using the goals as a prompt has two advantages. Judges recognise the framing immediately, and each goal can be narrowed into a local problem. A few examples of how a broad goal becomes a weekend project:
- Inclusive, safe and resilient cities: a map of accessible entrances, shaded rest stops or cooling centres, built from volunteer reports.
- Reduced inequality: a form-filling assistant that explains benefit applications in plain language.
- Climate action: a household energy comparison tool that turns a utility bill into a short list of practical changes.
- Responsible consumption: a swap-and-repair directory for a neighbourhood.
- Peaceful and inclusive societies: a tool that makes public meeting agendas and minutes searchable.
A goal on its own is not an idea. The test is whether the team can finish the sentence “this helps a specific person do a specific thing faster, cheaper or more safely” without using the goal’s name.
Ideas that grew from matching teams with partners
Some of the most useful lessons come from what happened after the hacking. The pan-European EUvsVirus programme, described on the EUvsVirus site, reports that it created 2,164 multi-disciplinary, multi-nationality teams during April, then sparked 2,235 new cross-European partnerships in May by matching the best teams with supportive partners from the public and private sectors.
For idea selection, the takeaway is practical. Projects that name a likely adopter, such as a hospital department, a city office, a charity or a school, are easier to carry past the event. When choosing between two ideas, prefer the one where you can name the first organisation that might use it.
That thinking also shapes who joins the team. An idea aimed at public services benefits from someone who understands how those services work, and a data-heavy idea needs a person comfortable cleaning messy files. The hackathon team roles guide explains how those responsibilities are usually split.
Ideas built on open data
Open data lowers the barrier to a first prototype, because the inputs already exist. The NASA Space Apps Challenge is a well-known example of the model. Its Space Apps site says that each year thousands of teams submit projects that address challenges on Earth and in space using NASA data, with multiple rounds of judging used to select 10 Global Winners.
The listed winners show how varied data-driven ideas can be. One winning project responded to a challenge to build a space biology knowledge engine, and another addressed hunting for exoplanets with artificial intelligence. Neither idea needs a rocket. Both start from a public dataset and a clearly worded question.
The same approach works for local problems. Cities, counties and agencies publish datasets on transit, permits, air quality, food inspections, road conditions and budgets. Useful concepts include:
- a plain-language dashboard for how a city spends its budget;
- an alert service for planned road works on a person’s commute;
- a searchable record of public notices, grouped by street or neighbourhood;
- a tool that compares service response times across districts.
Anything that touches personal information needs extra care. Collect only what the prototype requires, and avoid putting real people’s records into a demo. Public datasets are safer ground than private ones.
How to choose and shrink an idea
With a long list in hand, narrow it down quickly. A simple method works for most teams: write each idea as one sentence naming the user and the action, then score it on three questions. Can it be built with the skills in the room? Can it be demonstrated in a few minutes? Is there data or a partner to make it believable?
Next, cut scope until the prototype is almost embarrassingly small. Pick one screen, one flow and one dataset. A narrow prototype that runs reliably leaves time for testing and for the story, and the story matters: judges remember a clear problem, a working demo and a plausible path forward. The hackathon pitch guide covers how to present that story.
Finally, write down what the project will not do. A short list of out-of-scope features keeps the team from drifting in the last hours and gives the pitch an honest answer to “what comes next?”.
Frequently asked questions
Where can a team find hackathon ideas if nothing comes to mind?
Start with published challenge themes from past events, global goals such as the Sustainable Development Goals, or open datasets from a local government. Each gives a defined problem to narrow.
Should an idea be original?
Not necessarily. A familiar concept applied to a specific community, language or dataset can be more useful than a novel idea with no clear user. Judges usually care more about fit and execution than novelty alone.
How technical does an idea need to be?
It needs to be technical enough to demonstrate, and no more. A working prototype of a simple concept beats an ambitious design that cannot run. Teams without developers can still contribute research, design and testing.
Can a hackathon idea become a real project?
Yes, particularly when it names a likely adopter and uses data or tools that remain available after the event. Matching teams with partners was the stated aim of the EUvsVirus programme described above.
What if two teams choose the same idea?
It happens often at themed events. Differentiate through the user, the data or the approach, and focus the demo on what is distinct about the version your team built.
