Working with Mark had been a transformative experience. He had taught me to think and to solve problems objectively. He taught me that everything is a hypothesis until you’ve been able to validate it. He taught me to connect with and learn from everyone. Fundamentally, he had taught me to be an entrepreneur, and I was itching to be in an environment where I could put it to use.
I wasn’t the only employee Mark would go out of his way for. Everything he does has an undertone of empathy: his habit of replying to every email, his insistence on great perks and a fun office space, and the amount of time he spends working to properly communicate with employees. He would have me meet people from all over the company to capture their ideas and find ways to make them heard.
Mark clearly works hard to enable and empower his employees and these quotes bring out what it means to do that, with a practical ‘what to do’ perspective for leaders and a ‘what does it feel like’ perspective for employees. Powerful stuff.
As we all know, Zynga has had its problems and all is clearly not right with its culture, but the VCs who have invested in Pincus write glowingly of his management abilities and I think we are seeing more and more of this leadership model in both small and large companies. I think it is especially powerful in startups because the founder is able to have a closer relationship with employees, because empowered employees deliver better results in highly fluid and dynamic environments (i.e. startup environments) and because fast growth creates space for people to grow into.
A couple of weeks ago I wrote about the opportunity to find value in the shadows of big data by gleaning insights that are unrelated to reason the data was collected in the first place. This morning I read Max Levchin’s transcript of his DLD talk from last month in which he also talks about drawing value from big data. Levchin sums his thesis up with the sentence:
the next big wave of opportunities exists in centralized processing of data gathered from primarily analog systems
By which he means using digitised data about non-digital items to improve efficiency in the way those resources are used and priced with value created by increased utilisation.
Collaborative consumption is one example of this phenomenon which is already playing out where often the digital data is manually entered (e.g. apartment owners tell Airbnb when their apartments are available). Going forward data will increasingly be gathered by sensors and increased utilisation will spread beyond items that we own and taxis that we use to our time and capital and benefits will grow from increasing the time in use to include dynamic pricing.
I’m going to illustrate with two of Max’s examples. Firstly better utilisation of aeroplance pilots:
There is also a neat symmetry to this analog-to-digtail transformation — enabling centralization of unique analog capacities. As soon as the general public is ready for it, many things handled by a human at the edge of consumption will be controlled by the best currently available human at the center of the system, real time sensors bringing the necessary data to them in real time. The freshest, smartest pilot, most familiar with the particular complicated airport will land your plane — via remote control.
And secondly dynamic pricing of car insurance:
On a Sat morning, I load my two toddlers into their respective child seats, and my car’s in-wheel strain gauges detect the weight difference and reports that the kids are with me in a moving vehicle to my insurance via a secure message through my iPhone. The insurance company duly increases today’s premium by a few dollars.
My keepHonest app sees this too and immediately offers me up as a customer to a few competing insurance companies in the background, but nobody is willing to charge me less right now, and the phone chirps sadly to let me know I’m now paying a higher premium. Safer, but more expensive.
But In a few hours, my car’s GPS duly reports to my insurer that I only drove two miles to the park, never sped and, and observed all traffic signs. My phone now chirps happily: not only has my rate been discounted, several companies are offering me a deal on insurance!
There will be opportunities in both the infrastructure and application layers for delivering these services. The key to successful investment will be understanding the dependencies between the two and getting the timing right.
I’m sitting at home now recovering from a knee operation I had yesterday. I had my anterior cruciate ligament (ACL) reconstructed and a tear in the meniscus repaired with a couple of stitches. The ACL is a ligament that runs from the front to the back of the knee and gives the knee stability. The meniscus is a disc of cartilage that sits on top of the tibia (shin bone).
I first injured my ACL some six or seven years ago skiing. I was transitioning from snowboarding to skiing and had just got good enough with the two planks to feel I could take on the mountain but hadn’t yet understood that it is much easier to hurt yourself on skis than it is on a snowboard. So I skied into an area I shouldn’t have been attempting and twisted my knee in a fall. I had maybe a dozen physio sessions when I got back and the knee was OK, but never great. Then at the beginning of last year it collapsed on me whilst skiing and then whilst playing football. It got a bit better but then went again in the summer when I was dancing in the silent disco at the Rewind Festival. All the coolest people injure themselves dancing in silent discos…
This collapse was worse than the previous ones because I also injured my meniscus. I had a bunch of physio during the autumn, but kept suffering from knee irriation and couldn’t play football without it hurting. So in December I went for an MRI which revealed that my ACL was fused to the posteria cruciate ligament rather than the patella (knee cap), presumably a result of incorrect healling after the ski injury of six or seven years ago.
My consultant gave me the choice of fixing the meniscus tear and reconstructing the ACL so it was fused in the right place or just fixing the meniscus tear. It was a difficult choice because most of the time my knee was fine. I could run and cycle without problems, and skiing was fine most of the time. I even considered not having surgery at at all. In the end I evaluated the options like this:
No surgery – benefit – knee works fine under most circumstances and no risk of an imminent step backwards due to complications in surgery, con – couldn’t play football with the kids very much, knee seems to be getting more unstable over time, risk of early arthritis
Fix the meniscus tear – benefit – would be able to play football again, relatively minor surgery compared with an ACL reconstruction, con – doesn’t address the stability or arthritis issues
Fix the meniscus tear and reconstruct the ACL – benefit – prospect of a fully working knee, con – ACL reconstruction is a majory surgery with accompanying chances of complications, including major loss of knee function (small chance)
The consistent advice I received was to go for the third option if my knee was unstable and I didn’t trust it. As I saw it the answer to this was unclear, making it a marginal decision whether to go for the full reconstruction. I went for it in the end because I’m the sort of person who likes to get the fundamentals right and struggling along with the ACL fused in the wrong place with the accompanying increased risk of further knee collapse and early arthritis just didn’t feel like the best thing to do. The decision was also linked to my desire to stay fit in mind and body and keep exercising for as many years as possible. Not having a full operation would have meant giving up on that a little, something I’m not ready to do.
Happily, one day after the operation it looks like I made the right choice. At six weeks in a leg brace and on crutches the recovery time is much longer than anticipated, but paradoxically my chances of a fully functioning knee are better than we all expected going into the operation. The reason for the paradox is that they were able to repair the tear in the meniscus rather than just tidy the frayed edges. That should lead to a better end result, but for the time being the stitches are fragile and need to be protected.
I’m not yet sure what the longer recovery time means for my ability to get out and meet people. I had planned to work from home this week and be back in the office next week, but it may not be wise to leave here so soon. Certainly I think it will be unwise to travel on the Underground for a little while yet. My phone and email are working though.
I’m going to see how my knee heals this week and make a decision on Thursday or Friday about whether to spend next week working from home as well. After that I will be in touch with any of you who I’m due to meet, either to re-arrange or change the meeting to a Skype call.
As I sat waiting to go into surgery yesterday I wished I’d already posted about my decision to go for the operation. Writing afterwards like this inevitably has an air of post event rationalisation, and it would have been fun to post my reasons first and then see if I got it right afterwards. Still, I wanted to capture my thinking for posterity. I hope it’s been interesting.
I just read this quote from Jeff Weiner, CEO of LinkedIn:
I’ve come to learn there is a virtuous cycle to transparency and a very vicious cycle of obfuscation. People have an insatiable curiosity, and if they’re officially denied access to information, they’re going to dig for it on their own. And if they find it, they’ll become resentful and want to leak it. That’s when executive management says, well, clearly we can’t trust our employees with this information. So, we’re going to have to buckle down and release even less information.
I couldn’t agree more. Seeking to control people by withholding information was probably never a good idea and these days it is downright dangerous. Good corporate cultures encourage respect between employees and transparency breeds respect. Obfuscation breeds suspicion and sometimes contempt.
At startups I advise sharing as much information as possible, partly for the reason given above, but also because it helps people to do their jobs better.
There is, of course, some information that it isn’t appropriate to share. Most companies choose not to share it widely when they are in acquisition talks because it is destabilising for employees who will start worrying about whether they will keep their jobs and what it might mean for their careers. All that for a deal that probably won’t happen anyway (most companies that exit successfully have multiple discussions with different suitors before a deal is consummated). Similarly most companies choose to keep their balance sheet information within a small group as it is easy for people to see a low cash figure and over-estimate the chances of the company running into problems. Most successful businesses run low on cash at some point and they wouldn’t have been successful if staff had panicked and looked for jobs elsewhere. Sharing revenue information is helpful though because it helps explain why decisions are being taken in both upside and downside scenarios, moreover, most people have a reasonable feel for how the business is doing anyway.
Subscription ecommerce has become quite a hot business model over the past couple of years, perhaps reaching peak hotness 6-12 months ago. Notable examples of companies in this market include our portfolio company Lovefilm (acquired by Amazon for a rumoured £200m), our portfolio company Graze that sells healthy snacks, shoe subscription businesses ShoeDazzle and StylistPick, and startup coffee subscription business Kopi.
Phil Wilkinson, the founder of Kopi, is an old friend and we were discussing characteristics of good ecommerce subscription businesses over lunch yesterday. This is the list that we came up with:
Product that is well curated and premium, different, or hard to get elsewhere
Large number of potential subscribers (often hard to establish this at the beginning – who knows how many people will want to subscribe to DVDs, food boxes, shoes or coffee?)
Gross margins of 30%+
Attractive customer acquisition dynamics (for high value services this could just be paid search and TV , for lower value services member get member campaigns and other low cost channels are likely to be important)
Form factor that allows for convenient delivery (in the UK that means fitting through a standard letterbox)
A customer acquisition model that makes customers comfortable with taking on the commitment of a subscription (e.g. free first month, option to purchase before subscribing)
Content strategy that educates the customer and takes the customer on a journey
It isn’t (of course) necessary for a business to have all of these characteristics in order to be successful, but high potential ecommerce subscription businesses will have most of them. If you can think of any others that should be on the list please let me know in the comments.
I’ve started putting interesting Youtube videos in my ‘watch later’ folder and watching them on my Nexus 7 whilst I stretch and do physio in the mornings. This morning I finished watching a great Seth Godin talk on what’s wrong with our education system (embedded below). There’s lots of very thought provoking content and if education interests you I heartily recommend taking 15 minutes to watch the whole video. I will return to education briefly at the end of the post, but the thing I loved the most was the bit around 8mins 40s when Godin said:
Someone who is making art doesn’t say ‘Can I make one less canvas this month?’. They don’t say ‘Can I write one less song this month?’. … It’s art. They want to do more of it. But when it’s work, when it’s your job … of course you want to do less of it.
Hearing that I immediately thought about our portfolio companies. As you would expect, some of them are places where everyone loves their work and does a lot of it. They put a lot of hours in. For people there work is almost their first commitment. Whereas others have people who don’t love their work quite so much (sometimes they don’t like it at all) and they look to balance their work life and other interests much more evenly. When it comes to recruitment the first set of companies generally have an easier time – they are more able to find good candidates from their networks and people take salary cuts to join (at least in the early stages). The second set end up using headhunters more often and find themselves paying higher salaries.
These differences between companies where people love to work and those where they just work are not the be all and end all. We have had great exits from companies that definitely fall into the latter camp, and we have had our fair share of failures with companies where all the staff have been very passionate. Moreover, whilst I’ve painted a simple picture here with two types of company the reality is more complex with most companies sitting somewhere on a continuum between the two extremes, with different teams in the company being more or less passionate about their work.
However, with these caveats, I believe that companies with passionate people who love their work are more successful than the average.
Which begs the question: what makes people love their work?
Coming back to Godin – one of the big contributors is the feeling that they are making art, i.e. making something important. Companies create that feeling, or rather founders create that feeling, by combining vision and culture and by working in interesting areas.
Our portfolio company Lyst is a good example of a company where people love to work. They are aiming to build the worlds premier online destination to discover and buy fashion. That’s a big goal in an industry that many people find fascinating, and there are many interesting challenges that need to be solved to realise the vision. The company is still fairly small in headcount terms but it’s already clear that Chris and Seb (the founders) have created a positive and empowering culture.
This post is long enough already and I want to come back to education so I won’t comment more here on what constitutes a great vision or a how to build a strong culture, other than to say that for anyone thinking about these topics it is helpful to be mindful of the the distinction between art and work and how the former motivates whilst the latter demotivates. This is a confusing notion for many of us in the west who have been brought up as good capitalists with a protestant work ethic, but it will be increasingly important going forward as the higher value endeavours become more and more about creating something magical.
Coming back to education. The main point of Godin’s talk is that our educational systems were designed to create good factory workers who did what they were told. Too many of our schools teach us how to learn stuff by rote and follow plans without asking difficult questions. Those skills are less useful now. As Godin puts it, learning the dots isn’t enough now, we should be teaching our kids how to join them. The real disruption in education then is changing the goal of school to teaching our kids to create interesting things by joining new dots, and maybe to have an ethical agenda that is more socially responsible and less about industry and production.
Last week I wrote about my frustration with passwords. A number of you chipped in saying you shared the feeling and a few of you suggested services that can help with the problem.
These services break down into two groups.
The first group run client software (maybe in the browser) and automatically enter passwords so users don’t have to remember them. They can also generate complex passwords that are hard to hack. I haven’t seen their numbers but my impression is that they are doing ok, but haven’t penetrated the mainstream. They have been in the market for a while. I started using one of them myself a year or so back but didn’t really stick with it. I think they are good services, but to use them effectively takes a bit of work from the user, who has to install the software on multiple devices (and it might not work on all their phones) and may have to pay for a premium subscription. My guess is that the money and effort is worth it for the people who most worried about password security, maybe because they have been hacked or because they place a very high value on peace of mind, but not for the mainstream. I haven’t named these services because I don’t want to be critical of specific companies (at least not when they are startups).
The second group can be classed as new approaches. PixelPin is one, Nok Nok Labs is another. Rather than work with existing text based password systems they seek to replace the ‘password’ dialogue box with another form of authentication. PixelPin asks users to remember specific points on a photo whilst Nok Nok uses smartphones to identify via voice recognition and fingerprints. Like the first group they are primarily focused on solving the memory problem, but they have chosen a different go-to-market route. Instead of going after consumers they are going after websites and services that ask their users to enter passwords. If the products are designed well this has the merit of potentially being simultaneously less effort for consumers and much more secure, which would be a massive step forward. They will, however, have to convince website operators to implement their systems. To do that they will have to on the right side of the cost/convenience equation for them too, although the calculations of the costs of existing systems should include the cost of replacing passwords and sorting out people whose accounts have been hacked.
So it all comes down to ease of use – either from the consumer side or the perspective of the site operator. That tells me that the lack of security which comes with users using the same password across multiple services doesn’t worry most consumers or site operators that much.
I’ve long been a fan of Carlota Perez and her work on innovation and and financial capital. I think it brings an important historical context that helps us understand what is happening today and to better predict what might happen over the next few years. Over the weekend Chris Dixon wrote the following in a post about her work and his view on where the opportunities for innovation will most likely come going forward:
Technological revolutions happen in two main phases: the installation phase and the deployment phase. Here’s a chart (from this excellent book by Carlota Perez via Fred Wilson) showing the four previous technological revolutions and the first part of the current one:
Each revolution begins with a financial bubble that propels the (irrationally) rapid “installation” of the new technology. Then there’s a crash, followed by a recovery and then a long period of productive growth as the new technology is “deployed” throughout other industries as well as society more broadly. Eventually the revolution runs its course and a new technological revolution begins.
I think it might actually be more useful to think of the PC and the internet as separate revolutions with the PC related bubble crashing in the late 1980s and the PC deployment period thereafter overlapping with the installation period for the internet. It doesn’t make that much difference though because whether their has been one ICT revolution or two we are thirteen years into a turning point/crash that has been going on since 2000 and hopefully we will soon enter a deployment period of steady growth as internet technologies and network based business models drive productivity improvements and economic growth. My reading of the tea leaves is that there is a good chance that the financial instability that has dogged us since 2008 is in its final phases.
Innovation in the installation period is in the core technologies (e.g. routers, search engines) and in the deployment period it shifts to applications that are higher up the stack and to utilising the new technology across all industries. Note that realising the maximum benefits usually means changing the structure of an industry rather than simply making its existing structure more efficient. Netflix’s recent release of high budget drama House of Cards is a case in point – their decision to make all thirteen episodes available from the first day could herald the end of linear TV. New ecommerce models where web retailers focus on providing great search and discovery and drop ship goods direct from manufacturers without taking stock is another good example, as is the whole collaborative commerce industry and the transformation of education by free to use online services.
I think the opportunities come in both directly disrupting what I will call traditional industries and in providing the picks and shovels for the disruptors.
I’ve just read Betaworks 2012 Shareholder Letter written by the CEO John Borthwick. At 43 pages its a long read. Its pretty dense too. But there’s an amazing amount of great insight there – from developments in the investment market to interesting startup themes.
One of the themes John talks about is drawing value from the ‘shadows’ of big data – i.e. gleaning insights from big data sets that bare little relation to the reason the data was collected in the first place. It is easy to think of some obvious examples – including traffic data from people using Google maps (see picture above) and opening hours for restaurants and bars from Foursquare check-in data – but the value will be in finding the larger number of smaller bits of value. Call it making big data useful, if you will. This applies to consumer and enterprise, and I think we will see startups succeed both in making enabling infrastructure and tools, and in making the end user applications.
I’m going to close with a fun example of drawing insight from the shadows of data, also from the Betaworks shareholder letter. This time the original dataset was links shared on Bitly (quote from Hilary Mason, Bitly Chief Scientist):
“People in Brooklyn are more likely to read about food than people in any other part of NYC. People in Manhattan are more likely to read about business. People in Queens are more likely to read about sports.
The top Wikipedia article of 2012 was "Lunch"
People who read about fashion read about physics, but people who read about chemistry don’t read about anything else (which is behavior we see in two other categories — religion and adult).
Sports was the one topic for which people actually only want to consume fresh information. Every other topic was roughly equal, with religion being the slowest.“
That made me laugh on a number of levels. I know people think with their stomachs, but it’s amazing that ‘lunch’ is the most popular article in the world’s most read encyclopedia, and it’s amusing to read about chemists… (not that I draw any conclusions).
I don’t post many infographics here, but this is a great one. It covers concepts like the viral co-fficient and time through the viral loop and lists out a bunch of tactics for increasing virality. None of this stuff is new, but this list is comprehensive and it’s great to have it in one place.
Additionally, not many businesses have the characteristics that make viral growth a possibility and the other thing I like about this infographic is that it can be used as a kind of check list to establish whether viral growth is a possibility.