It shouldn’t have come as a surprise – drawing up a ranking of startup scenes around the world was always going to cause consternation and celebration in equal measures. For Berlin, it was most definitely the former, with the Startup Genome report placing the German capital in 17th place, below such tech hubs as Vancouver, Santiago and Sao Paulo.
But is it a case of Berlin have too high an opinion of itself, or was the data genuinely flawed? Silicon Allee spoke to Startup Genome founder Bjoern Herrmann about the findings, which have been released a year after the project was founded.
SILICON ALLEE: What have you made of the reaction to the report?
BJOERN HERRMANN: I didn’t expect people to react like this over a ranking. You can structure ranking in many different ways. At this point, it’s based on activity; we are working on a more refined ranking including a number of other factors such as overall performance, number of acquisitions, volume of investment and so on.
SA: What’s the driving force behind Startup Genome?
BH: We want to give a powerful sensory system to startups so they can better interpret the signals from the market. The goal of the Startup Genome project is to decrease the massive failure rate of startups and increase the overall pace of innovation. To get there we set out to crack the innovation code. Solving this puzzle not only has the potential to raise the output of Silicon Valley to heights never seen before, but would allow the magic of Silicon Valley to be shared with the rest of the world. We believe success could mark the beginning of a new era, an era we are calling the Entrepreneurial Enlightenment.
SA: Obviously, the headline news in Berlin was the ranking you came up with in terms of tech scenes – it seems like you stress how data driven your whole project is, but some have commented that the data you used to come up with the list was flawed. Did the varying degree of access to Startup Genome have an effect on the list? Why would Berlin be below the likes of Santiago or Madrid?
BH: Generally, when looking at all these ecosystems, there is little difference in activity except for those that have been established for a long time. There are some that dominate the scene – Silicon Valley has done this the most, it’s far beyond the other ecosystems. So in terms of size, performance and general money invested and output, its 40-50 times bigger than most. All these ecosystems are young but flourishing… some are growing faster than others but all are still very much in their ‘children’s show’ stage. Every ecosystem hypes itself up and thinks it’s awesome. Berlin is certainly doing good but it just needs to be seen in perspective. Silicon Valley grew over 70 years…
SA: So Berlin needs an innovation cluster?
BH: Exactly. Berlin stands out for its consumer-focused companies, more innovative in design and more outgoing… In the Valley, you hardly see these companies, they’re sitting in the garage and working on something – they are there but you don’t hear about them. From what I’ve heard, Berlin needs more tech people – it seems like there are lots of business and creative people but not a strong scene of tech-heavy people; the engineers and scientists who come up with the groundbreaking innovations. You need to have those people in order to create a thriving technology entrepreneurship scene. Germany has been dominated by a scene of entrepreneurs who mostly come from business schools but who still show a disrepect for people who have a heavy tech background. The symbiosis between tech and business doesn’t seem to have appeared yet in Berlin.
SA: What’s the most vital piece in the jigsaw for Berlin’s tech scene?
BH: Talent is the most important thing.The talent should be strongly technical. In this regard, I know if we were more harsh on how we differentiate companies, Berlin could be lower on the list. Every ecosystem needs a soul. Berlin has this with its creativity but needs to build a better bridge with the technical side.
SA: How did you gather information on Berlin for the list?
BH: We did a lot of qualitative research through interviews (with people working in startups). Learning more about their assessment of the ecosystem and their key problems and key info that they wanted to know about the ecosystem. We have a lot of data but we needed context so that’s why we’re doing more qualitative research… that’s what (Berlin field researcher) Danny Holtschke is doing right now.
SA: You have been trying to crack the code of how to determine what specifically makes a startup successful. What inspired you to take this challenge on?
BH: After I finished my last startup, I got together with Max Marmer, who at the time was doing independent research at Stanford together with Steve Blank. He created the first step of what the Startup Genome is today. We’re both inspired by this explosion of world entrepreneurship. They have this massive impact on our economy, most of the job growth comes from highly scalable startups. However, most businesses fail not because of competition but because of self-destruction. Most of the time those businesses efficiently execute the unnecessary. That means that most of these small businesses have lots of energy but they don’t really know where the hell they’re going and why they’re doing it, who wants their solution, etc. Essentially they are missing insights into the market and their own processes. That’s something that a lot of larger businesses have been able to overcome, especially in the last five years, because there has been a big predictive analytics movement that has allowed them to leverage big data to make better decisions when they launch.
SA: What’s a good example of this?
BH: Zynga has been able to master this. They have been able to increase the success rate of new games or new startups by leveraging big data in order to know how to allocate their resources, knowing how much they spent on customer acquisition, pricing, which channel to push it, and so on. And we thought maybe we could make the same thing available to smaller companies and do this by sharing lots of data with them. That was the initial inspiration.
SA: There are around 15,000 companies currently using Startup Compass, a benchmarking tool developed by Startup Genome. Do you think it has been a success with founders around the world?
BH: That’s a question I can only answer in two to three years. Of course many people find it valuable, but specifically one of the main use cases is allocating resources… You can’t be really data driven as a business if you don’t have enough context around the data, that’s what we hope to provide.
SA: How do you prevent conflicting interests? If 6Wunderkinder and then Evernote uploaded, would they get paired up because they are in the same industry? How do you convince startups to throw in their data without the fear of revealing too much to the competition?
BH: They are grouped in a way that you can’t identify them, you can’t see competitors. And also, the clustering and algorithms are more advanced than just taking two companies that seem similar and comparing them; there is more finesse to it than that. This whole taxonomy around different parts of your business and a few other variables define how we do the benchmarking. That allows you to automate the work of an analyst. The models are inherently scalable, we call it “analyst in the cloud.” So usually, an analyst costs $150,000. Instead of paying IBM or McKenzie who do it manually, this algorithm does it automatically. This is key way in dropping the costs and delivering insights that are partially available today but just cost a lot of money.
While the Startup Genome ranking may very well be a work in progress, perhaps Berlin should focus more on reaching a critical mass than on caring what other people think about it. There will be more on Berlin itself and how it compares to other tech scenes in an interview with Danny Holtschke, field researcher for Startup Genome. Watch this space.