Albert-László Barabási and the people who loathe academic popstars

The AngelList syndicate network from an earlier visualisation from this blog. Is it a scale free network? Who knows.

Some scientists really hate when other scientists become media darlings.

In my former life as a researcher for British Telecom I read and thought a lot about complexity theory and networks. We were interested in how wireless devices could form networks of their own, so called ad-hoc networks, and communicate with each other without the need for centralized control.

One of the people I followed in the field is Albert-László Barabási who has become a science popstar since I first learned of him. He wrote the best selling book Linked, which has one of the most ambitious tag lines around:

How Everything is Connected to Everything Else and What it Means for Business and Everyday Life

Everything, everything else, everyday. Phew.

The author of a blog post that surfaced recently on Hacker News, The network nonsense of Albert-László Barabási, isn’t impressed.

Barabási’s “work” is a regular feature in the journals Nature and Science despite the fact that many eminent scientists keep demonstrating that the network emperor has no clothes.

The comments on Hacker News are pretty funny as well. One of them tells of the popularity of fitting research data to scale-free models (the mathematical concept that underpins much of Barabási’s work):

[I] spent ~7 years in a bioinformatics PhD program focusing on biological networks … my advisor wanted me to SCALE-FREE ALL THE THINGS

Because Barabási has such a high profile, which makes lots of competitors loathe him, it’s hard for the layman I’ve become to determine how much truth there is in the accusations of the Network Emperor being sans clothes. I’m on the fence until someone who doesn’t have skin in the game delivers a convincing verdict. But I certainly find the exchange entertaining. Even if only as a blast from the past.

The new Seedcamp intake

Seedcamp just added 8 startups to the “family” of startups they invest in and support so Datasmoothie and the rest of February’s intake are no longer the new kids on the block. The office is buzzing because the new batch is in London for “onboarding week”. Onboarding week consists of back to back masterclasses, taught be people straight from the trenches, on everything from raising money, hiring people and building a product to marketing, selling and branding.

Findify logoWhen I scanned the list of new startups in the Seedcamp newsletter one of the companies stood out to me: Findify, “the intelligent search solution for e-commerce using data science and machine learning to increase revenue”.

Data science. Machine learning. Yummy. You plug their solution into your online store and your search becomes intelligent, increasing revenue. Also, they’re from Sweden. I’ve yet to meet a Swede I didn’t like.

I’ll get a smartwatch when they become less smart (and more sexy)

Watch and champagne
Me on Instagram demonstrating that 1.11pm is not to early to open a bottle of Moet, especially if you’re on your way to an MMA event (yes, I know). The watch is dumb but sexy.

The primary function of a smartwatch is to provide the owner with an internet enabled computer to wear on his wrist without looking like an absolute dork.

I saw my first Apple Watch in the wild last week. I was in the elevator in Campus London, Google’s startup hub. The elevator always scares me slightly, there are a few precarious seconds before it opens its doors when you’ve reached your destination where I always imagine the doors won’t actually open. So I try to find something else to think about. This time it was the Watch being worn on the entrepreneur standing next to me.

“That doesn’t look too bad on the wrist, does it?” I said, pointing to the Watch. It was the white sportsy looking one. “It’s very wearable,” he said, “but the software sucks”.

Of course it’s wearable. It’s designed by Jony Ive. He defined an era. His industrial design is in the MoMA. How can it not be wearable? So, Apple have at least achieved that much: Creating a smartwatch that doesn’t make you look like a dork (my wife disagrees, though). But the software isn’t quite there yet.

I wonder if there is simply too much software there. When I tried it in the store I couldn’t even open the right app, the “flower” home screen made all the icons too small. It looks like it was designed by the iTunes design committee, not someone with a proper vision and the guts to reduce, rather than add endlessly.

This is where I think Google might eventually have an edge over Apple. I don’t want my watch to be too smart. And it needs to be even sexier. And I’m not sure I want a watch from an electronics manufacturer. I want one from a watch maker or some sexy brand. If watchmakers like Daniel Wellington, Armani, Storm or even Rolex would create a watch that only had a very light touch of smartness, powered by Android, we might have something.

The mythical entrepreneur month

If I read another war story about a startup founder sitting in a dimly lit room, with only the soft glow of Sublime Editor on his Mac lighting up his surroundings, working for months on end on a schedule of 14 hours per day or more, six hours of sleep, a neglected dog and way too few showers, I might throw my laptop out out the window. I probably won’t though, I’d have to set up my development environment again. But I might.

The Mythical Man Month is a software engineering classic which theorizes that “adding manpower to a late software project makes it later”. In essence it advocates that measuring a project in “man months” is impossible, because adding manpower to a project doesn’t result in a linear increase in productivity.

I suggest we add to this vocabulary and introduce: The Mythical Entrepreneur Month. Doubling the number of hours you work per day won’t double your long term productivity. In fact, you might be damaging your business without noticing through your haze of sleep deprivation.

Running a business is a marathon, not a sprint. And absolutely everything will take longer than you think, whether it’s launching your product, closing the first deal or raising funding. At Datasmoothie we multiply all our time estimates with Pi. And stuff still takes longer than we think. So when you crash and burn because you’ve gone all out for months, there will still be so much more to do. Always. For some entrepreneurs, their business is a life’s work.

And it doesn’t even work for the period of time that you do manage to keep the pace up. An exhausted entrepreneur will write buggy code, make bad decisions and he’ll be bad company at the pub. Which means lost deals.

Everyone has to pull all-nighters every now and then. But these should be the exception, not the rule. So stop being silly. Step away from the laptop. Feed the dog. Take a shower. Your rested brain will probably come up with an important innovation for your business as soon as you give it a break.

The $3bn husband-wife data company you’ve never heard of

Screenshot 2015-05-12 09.10.15

A pioneer in the world of data science, that usually flies below the media radar, is up for sale and the price tag is somewhere north of $3Bn.

Dunnhumby was founded in west London by husband and wife team Edwina Dunn and Clive Humby. They paved the way in using statistics and data to analyse buying patterns, churn, customer segments and everything else that today is standard practice and we refer to as “business intelligence”. The BBC did a great article on them last year, titled “The couple who helped transform the way we shop“. It tells of the pioneering work they did for Tesco’s in the mid 90s:

After they had finished their presentation [to Tesco’s board], an awkward silence followed for more than a minute.

This was eventually broken by Tesco’s then chairman, Lord MacLaurin, who made a remark that has gone down in the supermarket’s folklore: “What scares me about this is that you know more about my customers after three months than I know after 30 years.

Tesco’s eventually acquired Dunnhumby but because of the supermarket’s recent change in fortune, they are looking to offload it. Interested buyers include advertising giant WPP and ratings company Nielsen.

After selling their business Ms Dunn and Mr Humby took six months off in the Caribbean and Mediterranean and then returned to work, joining Starcount, a startup that crunches data from social media and sells it to clients.



London: The fintech needs more fin

This week’s Economist has a very good special report on financial technology (fintech) startups. I jumped on this particular bandwagon last month when I opened a high-risk savings account with Nutmeg, one of the UK darlings of the industry (already down by 2.31% – thanks Nutmeg algorithm).

Screenshot 2015-05-11 10.23.21One of the things that struck me from the report is a chart that shows how investment in the sector is growing and who is doing the investing. What’s not surprising is the growth in investment. What is surprising, or at least slightly annoying for a Londoner and startup enthusiast, is that 75% of the investment is coming from the US.

London was the world’s biggest city when capitalism and modern finance was invented. Ever since it’s been one of the most important financial hubs on the globe, second only to New York. But even with all that cash slushing around in its capital’s coffers, the UK is light years behind the US in fintech investments. When TransferWise, another London based fintech startup (and a graduate from Seedcamp, I should add) raised a whopping $58M in investment to disrupt the money transfer industry, the investment round was lead by a Silicon Valley based firm.

Let’s hope that next time the Economist does a special report on the sector the London financial community will have joined me in jumping aboard the fintech ship, and will stop punching below its weight when it comes to investing in the sector’s startups. Even though it’s a much higher risk investment than my modest and as of yet unsuccessful Nutmeg portfolio, the returns it could bring would help secure London’s position as an important financial centre in the 21st century. London can not only afford it, it cannot afford not to.

Is over-segmentation creeping back into Apple?

Screenshot 2015-04-10 08.35.36

One of the first things Steve Jobs did when he came back to save Apple in the late 90s was reduce their PC product range to the “four product matrix”. Instead of having an offering for every single segment of the market (executive, geek, soccer mom, teenager, single dude … ad infinitum) they reduced it to four products: Laptop and desktop, each available for home and work. Simple. Brilliant.

When my 19 month old daughter recently started testing my iPad’s durability I decided to assess the monetary implications of any potential damage caused by her enthusiasm. When I found the page, I got confused. So many iPads. Why don’t they just have an iPad and iPad mini? Why five different products? Which one is for a Star Wars-socks wearing hipster wannabe?

When I bought my iPad 2 life was simpler. Now it feels like choosing a mobile phone plan. Are the MBAs taking over Apple again?

Datasmoothie is accepted into Europe’s premium startup accelerator


First we were lucky enough to be selected out of hundreds of applications to compete at Seedcamp week and present our startup to a group of investors and mentors from Europe’s most successful businesses and venture funds. Then we were among the 11 companies selected as winners. Datasmoothie has joined Europe’s most respected startup accelerator! (Seedcamp announcement).

Want to be in the loop? Join our mailing list for the latest news.

What is Datasmoothie?

Has IBM’s SPSS ever made you cry? We are here to relieve you of your pain.

Datasmoothie allows analysts in the “people data” industries, e.g. market research,  sociology, economics, psychology, political science etc., to analyse data and turn them into stories. It is the result of more than a year of coding for one of the world’s top market research companies, YouGov, and now we are about to release it to the masses.

Why market research?

Datasmoothie is a tool for anyone who is in the so-called “business of evidence”. In other words:

The collection and interpretation of customer, citizen or business information for the purpose of informing commercial and public policy decisions, improving management of customer or civic relationships, or improving commercial or public management efficiency.

Why are we targeting this market? Because the global annual turnover for the market research industry is over $40Bn and in the UK it’s comparable in size to the newspaper industry and the core film industry (see Market Research Society report for more about the industry’s UK profile).

This industry also happens to be using an antiquated toolset that is largely stuck on the desktop. We plan to change that by bringing statistical analysis and data storytelling online and into the 21st century.

Can I have a look?

Not yet! The computational underpinnings of Datasmoothie are production ready and are being used to produce real-world reports for real clients. But our public facing product that will conquer the world isn’t live yet. We’ll let you know as soon as we have something to show! Until then, thanks for your interest in Datasmoothie!

Who are the founders?


Because everyone was in London for Seedcamp week (usually we’re dispersed between London and Reykjavik) we decided to pose and pretend that we were in a boy band. From the left: Geir Freysson, Agnar Sigmarsson and Birgir Hrafn Sigurðsson.

Finally, if you want to stay in the loop please join our mailing list.

Why do VCs invest in people who smell bad? (Notes from Peter Thiel’s book, part I)

Zero to One

Why are most movies crap? Why do venture capitalists ask you whether your company can become a billion dollar business or not? And why do they invest in companies run by people in hoodies (Zuckerberg) or people who don’t shower (Jobs)? The answer is: Statistics.

Peter Thiel, who the New Yorker says “might be the most successful technology investor in the world,” explains the VC part of the equation succinctly in his new book Zero to One:

The biggest secret in venture capital is that the best investment in a successful fund equals or outperforms the entire rest of the fund combined. This implies two very strange rules for VCs: First, only invest in companies that have the potential to return the value of the entire fund. [Second], because rule number one is so restrictive, there can’t be any other rules.

A super successful startup is what Nassim Taleb calls a “positive black swan” – an event that is extremely positive but practically impossible to predict. And because only one company might be responsible for an entire fund’s return and there’s no way to tell which company it will be, investors have to build a portfolio of companies where each and every one has the potential to win the lottery.

Paul Graham, who’s Y-Combinator fund was an early investor in Dropbox and Airbnb, calls this “black swan farming“. And the reason for Thiel’s rule number two is, as Graham puts it, because “the best startup ideas seem at first like bad ideas”. Ideas may seem crazy and the founders may smell like they should shower more often, but a startup’s potential to be a blockbuster is the only thing that matters to an investor.

This is also the reason why film studios make so many crap movies. Their portfolio has to be full of potential blockbusters that will make up for all the loss-makers. And there’s no way to tell what movies will succeed. They’re in the business of black swan farming as well. So they hire a big name actor or actress and make sure the script has a high-speed car chase, all in 3D, and cross their fingers.

I highly recommend Thiel’s book. It’s based on the wildly successful Blake Masters online notes from Thiel’s Stanford lectures. It’s original, entertaining and thought provoking. And importantly, it’s not too long. If Thiel can be bothered to write another one, I look forward to reading it.

How the Berlin Wall reminds me of Facebook Messenger

Screenshot 2014-08-21 09.52.01“Ask her!” the receptionist at the Feiniger Gallery in Quedlinburg told us in German, pointing towards a terrified junior colleague who was shaking her head profusely. They thought I wanted them to babysit my 11 month old daughter while my wife and I had a look around the gallery.

They were showing a photo exhibit of life in Berlin before the Wall came down. When your economic and political system is so bad you have to build a wall and guard it with guns to keep people from leaving, you might possibly have gotten something wrong. It’s far fetched, but this made me think of Facebook.

At the gallery everyone was relieved when I finally managed to make my self understandable. “Oh, I see! Of course you can take your pram into the gallery.”

The motherly lady who was in the gallery itself was easier to understand. “There was a ceremony here just the other day,” she said in German, “and I wondered what the occasion was.” Everyone seems to talk to you more when you have a baby in tow.

“Then I realized,” she continued smiling. “It was the 13th of August! When they raised the Wall! Imagine, it’s been so long since it came down I forget the anniversary!”

Facebook raised its own wall recently*. They’re forcing all of their users to download their Messengar mobile app by ripping the chat feature out of their main app. So if you have a message on Facebook you either have to download a separate app or wait until you’re home at your desktop to reply.

Facebook’s Messenger app was launched almost three years ago, in August 2011. Apparently, not enough people are using it. I removed it from my own phone after I first tried it out as I prefered the main app.

The Berlin Wall was a solution to a symptom, not the disease. The problem wasn’t that people were leaving East Germany, the problem was East Germany itself.

Facebook’s Messenger is a subpar version of WhatsApp. The problem wasn’t that people weren’t aware of the app, the problem was that the app itself wasn’t compelling enough to use. Facebook shouldn’t have given up on fixing the app rather than forcing its users to install it.


* Note: I’m not comparing nice people in Silicon Valley who I respect to nasty people in the DDR.