Tunnel Vision in a Fast Moving Startup
Last week was unusual. On Monday, I was bitten by a shark, and that was the least stressful day that week. On Thursday, we laid off 20 percent of our employees.
QuizUp has seen incredible growth during the two years of our existence both regarding growing the organization and the product. In this blog I reflect on some of the failures and learnings from the last two years. This is also the topic of a talk I gave at a local tech conference, the video from that is at the bottom.
After an incredible launch, the growth trajectory has continued. The QuizUp app has seen the number one spot in the App Stores of 128 countries, reaching more than 75 million users and adding on average 40 thousand new users every single day.
This rollercoaster has taught us a whole bunch; however, we’ve needed to take a step back and reflect mindfully on our progress to learn and improve.
During this blog, I’m focusing on the top three failures we’ve made.
- Allowing teams to grow too large.
- Releasing QuizUp 2.0 with a Big Bang.
- Prioritizing growth over revenue for too long.
When looking at these three failures, the first thing we realized was that all of them are pretty basic: mind-blowingly basic. Small teams are the essence of fast moving startups. Short sprints with regular releases is at the core of agile development. And anyone with a fundamental understanding of business knows the importance of revenue. Identifying these failures was easy. Digging deeper and surfacing the root causes was harder and, in many ways, a better learning exercise.
1/ Allowing our teams to grow too large
Small teams are the essence of what makes a startup move faster than the incumbents. Big teams need extensive processes to function well, and even so, they’ll routinely have to deal with miscommunications and a lack of trust.
What did we do?
The team responsible for building QuizUp 2.0 became too large. We moved members and hired new ones so we could add them to this “super-team” as the scope of the project grew. Our team meetings became infused with tension, and people felt lost. During one of our retrospective meetings, our agile coach jokingly said, “This is not a meeting; this is a conference.” Without having a voice in the matters at hand, team members felt like tools.
Why did we allow this to happen?
The twelve-person team that launched QuizUp in November of 2013 all fit around the company’s dinner table. The team members had breakfast, lunch, and dinner together. Over the course of the months that followed, team members were added rapidly. As with the infamous boiling frog anecdote, we did not see this coming.
What is different now?
We realized this in November 2014 after a missed deadline and yet another terrible meeting. We vowed to keep our teams as small as we can, usually around 6–8 members. We now take great care to nurture our small teams with proper team building, regular retrospectives, and our Friday Feelings. Each team has the autonomy to execute its work the way it sees fit. Our teams are fully stacked with the mastery of the crafts they need, negating any external dependencies. Finally, and most importantly, each team has a clear northern star that clarifies its purpose.
2/ Releasing QuizUp 2.0 with a big bang
Iterative and incremental development allows you to lower technical, business, and time risks. Moreover, enables you to understand your customers sooner and with a lower upfront investment.
What did we do?
We decided to release QuizUp 2.0 with a big bang, an aggressive marketing budget, and even celebrity endorsements. The scope of our project grew as time went on, thus increasing anticipation and expectations. We originally estimated a three-month development time and a November 2014 launch. Numerous delays pushed us back to May 2015, seven months later. We could have multiplied our initial estimates with π. Although the launch of QuizUp 2.0 had a substantial impact on our retention, it was strenuous on our team. One person burned out and was sent home for a two-week recovery, and two other engineers quit due to the stress.
Why did we allow this to happen?
QuizUp 1.0 had “once in a lifetime” growth after a big bang launch. Our tunnel vision here is that we incorrectly attributed that growth to our big bang launch rather than to the game design itself and the inherent organic virality of the product. Today, we see on average about 40.000 new users every day; those users join organically without our marketing efforts or press mentions.
What is different now?
Since the stubborn birth of QuizUp 2.0, we’ve maintained a rigorous launch schedule. The QuizUp Apps for iOS and Android are now released, on average, every 16 and ten days, respectively. This is much easier on our team, which now enjoys shipping features much more quickly.
3/ Prioritizing growth over revenue for too long
This one is a bit different; we are in working on this as we speak.
Anyone with any sense of business knows that you need to generate revenue. Companies that do become sustainable, those that don’t die.
“Eventually, all businesses are valued as a multiple of earnings.” — Sam Altman
What did we do?
We put our attention and engineering resources on user acquisition and retention. We localized and translated our app, launched in new markets, and endlessly launched features aimed at improving retention. We made great strides on those fronts, doubling our DAU, pulling D1 retention from 40% to over 60% and making massive progress on M3 retention.
In hindsight, we were never happy with our retention and kept at it for too long. As we focused on getting our retention to Facebook-like numbers, we did not get around to properly resourcing our revenue efforts.
Why did we allow this to happen?
We believe two things made us de-emphasise the importance of revenue generation. Firstly, we had unrealistic retention goals and were never happy with our progress. Second, we strongly believed our users would hate ads and leave, thus slowing our growth. However, we now know that our growth is impressive and that ads do not affect retention. Man, we wish we had known this sooner!
What is different now?
This one is different; we have not solved everything, i.e., we’re not profitable, yet, but we’ve re-organized the entire company around increasing revenue. We have been championing growth above everything else for the last two years, so this will take some re-wiring of everyone’s brains. This will be hard, but we are more motivated than ever before.
Summing up
Small teams, more frequent releases, and a newfound focus on revenue are all pretty basic stuff. When you are in the trenches, you (or at least we did) have this tendency to lose track of the bigger picture and develop a tunnel vision.
When we sit back and think of ways this will help us, we always come to the same conclusion. We cannot guarantee these things will happen again. However, what we can do is prepare our organization to deal effectively with the scenarios that will come up again and again.
The important part here is to develop a strong culture within your organization that is continuously learning. We are striving to create an organization that is unafraid to surface issues, acknowledge failure and welcome feedback with open arms.
Help others learn not only inside your organization but anywhere you get the chance. Jump at the opportunity to share your knowledge because that deepens your own understanding.
❤ Thank you for reading!
The Video Version
What originally prompted this blog was a local tech conference called UT Messan. They asked me to talk about how we are building the QuizUp Organization so we spent some time reflecting on the past two years. You can watch the video here:
Thanks!
Thanks to Kiddi, Alli, Vala, Elísabet, Arna, Gummi, Arndís, Haukur, Palli, and Jimmy for feedback. Thanks to Svenni for the Keynote help and Bobby for the smilies.
