My role as a Tech Community Ambassador is basically to create “learning experiences” for my engineering team, and hackathons are part of the process.
Today, more and more employees decide to work for a company not only to climb the career ladder or for financial reasons, but in order to gain knowledge. At Deezer, we strongly believe that our people need to feel they are continuously learning on their professional journey, not only by doing but also by meeting experts from a variety of fields.
Taking part in an internal hackathon and meeting co-workers from other departments with whom you don’t have the opportunity to work on a daily basis, is extremely valuable for anyone in our company. It is also a first step before getting in touch with external communities.
A History of Hackathons
For a few years now, we have been organizing internal hackathons at our Parisian headquarters. These happen over two days, twice a year. They have become tradition and if you are part of the Deezer team — whatever your department — you are given the opportunity to work on a project of your choice and assemble a multidisciplinary team around you (one idea, one technical solution, one graphic identity imagined and developed by colleagues with various skills).
Our only rule has been that all participants should challenge our Deezer product and its ecosystem — in any manner.
Our core values for each hackathon are the following:
- JOIN : Anyone working at Deezer is welcome to participate, no matter their area of competence.
- TRY: Just try it, dare to tackle challenges, don’t be afraid to fail.
- BUILD: Believe in better, work on something new that has never been seen before.
- ENJOY: These are 2 days outside your everyday work, so have a good time!
- LEARN : Learn by doing and exchange ideas with co-workers you may not have met before.
- SHARE: Present your work in front of colleagues or a jury — you will always gain constructive feedback.
So far, we have tended to give teams as much autonomy as possible, allowing them to organize their time as they like before and during the hackathon.
At the end of the two days, all of the project teams — generally about 30 teams of varying sizes (1–10 people) — would pitch their work in front of a jury of managers, and 3 projects would be awarded: “Most Business-ready”, “Most Innovative” and “Most Fun”.
Time for a Revamp
We have received good feedback on our past events and our co-workers have always been happy to take some time to step back, think about the Deezer product and experience, and to suggest how we could improve. For various reasons, however, some promising innovations have not been given as much attention as they deserved and others have unfortunately not seen the light of day — at least, not outside of the company. Of course it’s always fun to spend two days hacking with your co-workers, but it would be a missed opportunity if we didn’t encourage our teams to follow through with their more promising ideas. Therefore it seemed necessary to challenge and redefine our way of organizing our internal hackathons in order to better highlight our teams’ work.
Also, after organizing and participating in several internal hackathons at Deezer, there are a few things I have noticed:
- Each hackathon needs to have its own identity: people are more likely to engage in a hackathon if your event stands out. For the last few hack sessions I tried to find a catchy tagline and create a dedicated visual that everyone could relate to.
- Clearly communicating the goal and organization of the hackathon ensures that people can prepare properly and that the event goes smoothly — the sooner, the better.
- Limiting constraints to a minimal set of rules is the key to success: people are a lot more productive and efficient when they feel free to be creative and when nothing is off limits.
For our hackathon last February, we brought in 2 robots (Nao & Pepper). Even though it sounded pretty cool, the final result — i.e. enabling the robots to play Deezer through voice control — was not so innovative. The proposition was probably too obvious, and so not challenging enough. - “We eat as we work” (French proverb from the Franche-Comté region): last but not least, it is crucial to offer participants the best hacking conditions (and quality food is compulsory!)
These observations helped us revamp our concept and deliver a new format that would hopefully better suit our teams’ initiatives, foster innovation and improve our product.
New Story, new Rules
So we organized our latest hackathon last December with a new formula:
We decided to use a grunge band — Nirvana — renowned for never accepting the music business rules as inspiration for our fifth Hack Session. We hijacked the famous track “Come as you are” and turned it into “Hack as you are”. The formulation fits particularly well with our spirit — always challenge our product, and some of our internal values — Just hack it & Believe in better.
We first teased our colleagues with a few mysterious posters inviting them to save the date, then the first official message announcing the hackathon was sent in October — more than 2 months before D-Day. Expectations were high so people were very enthusiastic!
For the first time we also changed the way we organized the hackathon. We tried out new rules with both positive and negative results:
1. A 3-week deadline
Come up with an idea and share it on a registration form before it’s too late!
Pros: By setting a time limit to register, we encouraged people to flesh out their project (with details on the idea and people to work on it) quickly.
Cons: It was quite a fast turnaround and some people who were on vacation at the time missed the boat…but the most motivated people anticipated it!
2. A proper mixed and multi-skilled jury
The jury pre-selected 10 projects before the hackathon. It consisted of 2 women and 4 men (Brand Manager, Head of Business Intelligence & Data Analytics, VP Design, VP Product, CTO and Chief Data & Research Officer) and for the first time ensured a reasonable balance between tech and non-tech managers.
Pros: With each member of the jury bringing their own skills and experiences to the table (in product, engineering, data, design, etc.), it was the perfect composition to coach and strengthen the 10 pre-selected teams before D-Day, and also challenge their project.
Cons: We encountered some resistance at first. People felt such a jury would hinder potential innovations. It was also unclear to them whether they could participate in the hackathon if they had not been pre-selected.
3. A ‘pirate’ mode for non-selected projects
Teams that were not pre-selected were still able to carry out their project and organize their own work, without having to pitch their projects.
Pros: Not having a deadline or having to pitch allowed people to enjoy the hack and work on their project without worrying about if they had put together a team with the right skillset.
Cons: The ‘pirate’ mode can be viewed as a second-class option because teams don’t get to pitch their project in front of the jury and don’t benefit from their coaching either.
4. A new voting system
In total, the jury voted 3 times:
- before the hack: a first vote to pre-select 10 projects;
- right after pitches: a second vote to rank their favorite projects (from 1 to 10);
- after the hackathon: a third and final vote to select the teams that will get extra time to deliver a production-ready project.
Pros: Voting anonymously means that every member of the jury feels free to vote however they choose. It is also a good way to keep them involved before, during and after the hackathon. Last but not least, this system allows us to easily identify relevant trends, and by extension, to improve and refine our roadmap.
Cons: Voting three times can be too repetitive and it is quite time-consuming to organize the votes well, especially the first and second times.
5. Estimation of the time-to-production
When the hackathon is over and before the last voting session, each team must estimate the extra time they need to deliver a finished feature / innovation:
- S (Small): less than 1 sprint (15 days)
- M (Medium): less than 2 sprints (<1 month)
- L (Large): more than 2 sprints (>1 month)
Pros: “Time is money” and time is key to identify strategic, innovative projects. This step also empowers teams to organize their work and pushes them to be realistic about the time that they have. Finally, it’s a very decisive metric that eases time allocation.
Cons: Pragmatism is often frustrating and it is sometimes hard to accept that some really cool projects are incompatible with our strategy and roadmap.
6. No more categories
We removed the three prize categories (“Most Business-ready”, “Most Innovative” and “Most Fun”) and decided instead to offer a unique possibility to earn extra-time to finalize the winning projects.
Pros: All kinds of projects stand a chance and a single ranking shows a vision and trends, which has a greater impact on teams.
Cons: There is no longer an awards ceremony at the end of the hackathon…but the true satisfaction will come from the public launch of hackathon projects!
Learnings & Results
When we announced our new rules in November through internal communications, there were mixed reactions. This was mainly because some people thought that pre-selecting 10 projects would prevent other projects from taking part in the hackathon.
Our main goal was to give 10 teams the opportunity to strengthen their team and put their concept to the test, but by advertising this and focusing on the 10 pre-selected teams, we overlooked communication on the projects that were not selected and didn’t properly mention that they would still be able to work on their project on their own. This meant that we had to clarify the situation very quickly through internal communications and direct face-to-face meetings with competing teams… Things eventually fell into place.
In the end, hackathons are all about learning by meeting people and doing stuff with them. It’s a win-win situation for your co-workers and your product, and my role is to ensure that all the stakeholders feel that way and enjoy every step of the journey. In order to do this, I learnt that it is crucial to be well-organized, communicate effectively with the teams and to follow up on projects.
To conclude, here is Hack-session #5 in facts and figures:
- 39 projects drafted
- 16 projects finalized, involving 83 participants
- 2 projects included in the roadmap for Q1 2019