Hackathon, design sprint, accelerator, venture studio or wargame: what each one tests

"We should run a hackathon" is often the answer to a question nobody has written down. I have run hackathons, design sprints and accelerator programmes, and worked alongside venture studios, in Europe and in Asia. Each of them is good at one thing. The trouble starts when a format is chosen because it sounds right for the moment, and then asked to answer a question it was never built for.

So instead of comparing formats on cost and duration, this piece compares them on the question each one answers. Five formats, five questions. Four of them test ideas, products or ventures. The fifth tests a decision.

Hackathon0:45

The question it answers: what could we do here?

A hackathon runs for one to five days, with anything from twenty to two hundred people in small teams. It produces a lot of rough ideas, a burst of energy and connections across departments that did not exist on Monday. It is very good at getting the supply chain lead talking to the data scientist, and at showing who in the company has an instinct for building things.

What it cannot tell you is whether any of those ideas work. Nobody outside the room sees them. The energy fades within a couple of weeks unless someone has already decided what happens to the winning teams afterwards. The best hackathons I ran had that "Monday morning" plan written before the event started. Without it, the event is mostly team building.

Get the next insight by email

Email address

Company website

Unsubscribe any time · Privacy

Design sprint1:40

The question it answers: does this specific concept make sense to the people it is for?

A design sprint takes four or five days and five to eight people. It starts from a defined problem, ends with a prototype in front of real users, and leaves the team with a clear signal on one concept. Run properly, in the Google Ventures mould, it is one of the fastest ways to get evidence. It also teaches teams what fast feels like, which many corporate teams have never experienced.

It needs a clear problem going in. A team still arguing about which problem to solve will find a sprint frustrating. And it tests one concept with users, so it says little about the organisation that would have to carry the idea.

Accelerator2:20

The question it answers: which of these concepts can become a venture?

An accelerator runs for eight to sixteen weeks, usually with three to eight teams working on concepts that have already passed a first filter. Teams get structure, mentoring and milestones, and they learn the craft of venture building along the way.

Two conditions decide whether it works. The teams need real time for it, at least half of their week. And there have to be gates where the evidence can stop a concept. In an accelerator where every team coasts to demo day regardless of what they learned, the main output is a set of good presentations.

Venture studio3:03

The question it answers: can we build this into a business?

A studio partners with a company to build new ventures with its own product, design, engineering and business development people. The company brings domain knowledge, customer access and funding. Each venture takes six to eighteen months, and the commitment is the largest of any format here, in money and in time.

In return you get something the lighter formats cannot give: a business, with customers. The risk is lock-in. If the studio's method does not suit the industry, or the chemistry between the two teams is wrong, a lot of budget can go before anyone says so.

Business wargame3:42

The question it answers: what happens when the people who react to this decision react?

The four formats above test whether an idea or a product works with customers. None of them test what happens inside and around the company once a decision is made. The CFO who promised the market a margin. The sales leader whose team loses commission while a new model ramps up. The rival who answers your launch with a migration offer. The board member who reads every page.

A business wargame puts a decision in front of those people before it is real. A leadership team makes its move, the people around it respond with their own motives and their own information, and the consequences carry into the next round. What comes out is a decision that has already met its first counter-moves, and a debrief on what held and what broke.

It is the wrong tool for generating ideas or for testing a product with users. It is useful at the moments the other formats lead up to: the go or no-go after an accelerator, the investment case for a studio venture, the pricing change a new business model forces on the core.

Halyard is a scenario built around exactly that moment. You are the first Chief Innovation Officer of a listed CRM vendor whose per-seat model is being eaten by AI agents. You own neither the roadmap nor the sales force. In the first round you write a two-page innovation thesis for the board, choose up to three bets and decide what you will not do, and then the CFO, the product chief and the revenue chief each read it in their own way.

A wargame can run as a facilitated workshop, as an async run at each participant's pace, or as a programme over several months. Engagement describes the three formats.

How the formats line up5:33

Laid side by side, the formats move from breadth to depth, and from cheap to expensive:

• Hackathon: many ideas, days, very low risk, light sponsorship.
• Design sprint: one concept tested with users, a week, low risk.
• Accelerator: a handful of concepts validated, a quarter, real time commitment, sponsors who can stop things.
• Venture studio: ventures built, a year or more, the largest commitment, board-level support.
• Business wargame: one decision tested against the people who react to it, hours to weeks, sponsors who will own the decision afterwards.

The first four are often sequenced as a funnel. A hackathon surfaces ideas and champions, sprints test the most promising concepts, an accelerator turns a few into venture proposals and a studio builds the winners. A wargame does not sit at a fixed stage of that funnel. It fits wherever a decision is about to commit people and money: before the gate, before the board, before the price change.

Where the mismatches happen6:36

The mistakes I have seen most often follow from asking a format the wrong question.

A hackathon is asked to produce a venture, and everyone is surprised when nothing gets implemented. An accelerator starts without enough raw material, and teams spend the first month looking for a problem. A company with no experience of sprints or accelerators signs a studio engagement, and the organisation is not ready to receive what the studio builds.

The fourth mismatch is the quietest. A concept passes every test with customers, and then stalls inside the company, because nobody tested how the CFO, the sales leader and the board would respond to it. The evidence was right. The decision was never rehearsed.

What I am still thinking about is why that last step so rarely gets the same attention as the others, when it is usually where the venture lives or dies.