Monthly Archives

November 2012

Thought for the weekend: I’d rather be a missionary than a mercenary

By | Amazon | No Comments

Jeff Bezos was interviewed by his CTO Werner Vogels at the AWS: Reinvent show yesterday. I’m a big admirer of Bezos. As well as building an amazing company he has a great business brain and regularly comes up with pithy quotes and insights which help us to understand the world better. The interview was written up on wGigaOM and Techcrunch and both publications reported his money quote which I used in the headline to this post (albeit slightly adapted):

I’d pick a missionary over a mercenary every day.

He said this whilst making the point that the best businesses tackle enduring problems that their founders have a passion for solving rather than chasing the next hot thing. All sound advice.

Reading this I was reminded of a song that my kids keep singing at the moment which has the line ‘Follow your dreams, and you’ll be happy in that special way’.

A bit of passion and happiness goes a long way at a startup.

The other thing I liked was Bezos defence of having a low margin business, saying it is impossible to operate efficiently in a high margin business. Breaking that down I think he is saying that efficiency is a competitive weapon that can be wielded alongside all the others available in the market (great innovation, scale economics, proprietary content, network effects, etc.) and that companies which use all the weapons bar efficiency and maintain high margins will ultimately be vulnerable to companies that use all the weapons and are also efficient – like Amazon.

Some things that make mobile user acquisition difficult

By | Mobile | 8 Comments

Vibhu Norby, founder of YC backed Everyme and Origami has a good post up explaining why his company is pivoting from mobile-first to web-first. At the highest level they are pivoting because they were unable to acquire and retain users profitably. The interesting thing is why. Nibhu has a good list of the things that make acquiring and retaining users tougher on mobile than it is on the web:

  • iteration cycles are slow making it hard to react fast to user behaviour and test lots of things
  • non-active users only download big updates
  • shipping bugs hurts your rating
  • emails have tiny conversion rates to mobile
  • consumers need a good reason to embark on the long process of going to the app store, finding an app, downloading it, entering their password, opening the app, and going through onboarding
  • you can’t get as much data from users as you can on the web because data entry is more painful
  • small screens and less data make it hard to add value on the first impression
  • App stores take a big cut

For a host of reasons the web is, of course, much easier, e.g. bugs can be fixed ten minutes after discovery, multiple landing pages can be tested simultaneously, the journey from email to service is very quick, connecting with OAuth is much quicker.

Nibhu also points out that there have only been a handful of mobile first companies that have been successful (Instagram, Tango, Shazam…) and gives some discouraging data on Path (only 200,000 DAUs, which doesn’t get you far as an ad supported company).

For these reasons, and because he has had two bad experiences Nibhu is now developing web first.

So, is he onto something here, should other entrepreneurs follow suit?

I don’t think so. What we can say for sure is that mobile is more difficult than the web, but as Instagram and others have shown, it is not impossible. Crucially, mobile has capabilities that can’t be exploited on other platforms – e.g. camera (Instagram), GPS (maps, Foursquare, Runkeeper), and connecting to third party sensors (e.g. blood glucose monitors).

I think the takeaway is that because of all the extra friction on mobile the value proposition of mobile first apps needs to be stronger than for web first apps, probably much stronger. The good news for mobile proponents is that the unique features offer the potential for creating some very strong value propositions. I’m particularly excited about innovation using the mobile as a platform to connect to third party sensors.

Moreover, as mobiles improve all the problems Vibhu describes will get easier, with the possible exception of the App store cut…

Thinking about new market opportunities case study: 3D printed figurines

By | 3D, Innovation, Startup general interest | 4 Comments

Regular readers will know that I’m excited about the emerging opportunities that 3D printing technologies are enabling. In fact I’ve been looking for an investment in this area for nearly two years now. My first thought was that falling costs and rising quality would unlock opportunities in the consumer space, but none of the opportunities we have looked at worked out, mostly because the use cases weren’t compelling enough or were too niche.

Now that you have that background you will understand that when I saw the headline 3D printing ‘photo booths’ popping up across the world I clicked straight through. The article, which is on the excellent Singularity Hub, rightly notes that there has been a lot of talk about 3D printing this year, and that ‘tech experts are still looking for something that will signal the technology’s best bet to transition from hype to mainstream application’. I’m with them up to that point, but then they say ‘Today, one application seems poised to accomplish this. Call it 3D printed figurines, personal miniatures, or photo sculptures’.

That is where I think they got it wrong. Here’s why.

When I’m assessing the market opportunity for new products I like to ask two questions:

  1. What do people do at the moment which suggests they will want the new product?
  2. What spend or activity will the new product displace?

Strong answers to these questions suggests that there will be demand for the product. There are a host of other questions that follow, not least ‘can it be profitably sold at a sensible price point?’ but with good answers here you have got something to work with.

Some examples:

  • Streaming music services:
    1. people already listened to music so they should like a new more convenient service better
    2. the activity and spend to be displaced was spending money on downloads and CDs and listening to MP3 players and traditional stereos
  • Twitter:
    1. people have always devoted significant time and money on showing their friends how cool/connected/intelligent they are and staying up to date with news and what their friends are doing
    2. the activity to be displaced was inefficient surfing of the web, writing of personal blogs, and sending of emails
  • Enterprise software products (generic)
    1. in most new product categories demand is evident because companies have hacked together custom solutions
    2. the spend to be displaced is money given to custom developers and systems integrators

I think 3D figurines are very cool, and I can see myself giving them as presents and maybe even buying one for myself, but when I apply these questions I come up blank. I can’t think of anything people do that suggest they really need 3D printed figurines and I can’t think of any spend that could be displaced. I would bet that when people are buying figurines the money is coming from their discretionary pot. There’s nothing wrong with that, but it doesn’t indicate to me that there is a sustainable market here. My guess is that at best we might see a fad where lots of us give these figurines as presents for a short while, but then the novelty will where off and we will be onto the next thing. I also don’t think the quality is great and hence the novelty value will be short lived.

Musings on the tension between trust and anonymity

By | Privacy | No Comments

Google just changed the Google Play store so that reviewers’ Google+ name and profile picture are visible, with no option for anonymity, writes Techcrunch.

I think this is a good move because identity engenders trust and the lack of trust on the web relative to offline is one of the things that holds people back from using and enjoying the web more. This is a complex issue and forcing users to reveal themselves is no panacea, but it helps with the biggest problem with reviews – people don’t know whether they are genuine.

I think this point about identity engendering trust extend beyond reviews to all online services. Offline we unthinkingly take stock of the people that we interact using a large number of non-verbal cues and identity enables us to do some of the same online. This is why interacting with people on Facebook feels better to most people than interacting with people on anonymous user groups. I think it is also why most people choose to have photos of themselves as their profile picture.

The other big benefit of forcing people to post under their real identity is that they generally behave more reasonably.

That said, this isn’t a black and white issue and there are situations where anonymity is a good thing. Whistleblowers and people reviewing products they don’t want to admit to owning are examples given in the comments to the Techcrunch post. Political rebels in authoritarian states are another obvious case. But to me these are edge cases and we shouldn’t design the architecture of the web around them. Moreover, whistleblowers and revolutionaries can find workarounds, including setting up fake accounts or using other services.

Finally, a word on the nightmare scenario of an authoritarian government using modern technology to control its population to nefarious ends. I guess my thoughts are that reducing anonymity increases the scope for abuse, but not meaningfully as for a while now there has been more than enough tech available for governments to take total control should they so wish. The right way to protect ourselves is to invest in and support our political systems rather than put controls on particular pieces of technology. I would argue that not only is this the right way, it is the only way, as any controls put in place by one government or regulator can be removed by another.

The changing dynamics of consumer internet investing

By | Business models, Consumer Internet, Venture Capital | 5 Comments

Fred Wilson wrote yesterday about changes in the consumer web and their implications for startups. Frist he observed that the large platforms (Google, Facebook, Amazon, MSFT, Twitter, etc) are ‘starting to suck up a lot of the oxygen’ making it ‘harder than ever to build a large audience from a standing start’. Secondly, he notes that the move from ‘desktop/web to mobile/app’ makes it more expensive to build a large user base, principally because of the need to develop for multiple platforms.

These changes are altering the dynamics of investing in consumer web companies. Historically the best of these businesses have been highly capital efficient and able to grow very fast on the back of the strength of their product and without dependence on a third party. This made consumer internet a very exciting investment category – this Pinterest case study shows why. Going forward the risk reward profile will be different. Higher development costs and dependence on the large platforms push the risk up and at the same time the chances of hitting a Facebook size winner have fallen (it is hard to see any of the large platforms allowing that to happen).

This doesn’t mean that the opportunities in consumer web are over. It does mean that going all out to build a huge audience without building in a solid business model from day one has become a riskier bet where success is largely predicated on one of the platform players acquiring the company. For me the sweet spot has now shifted to businesses that are able to able to extract a high ARPU from a focused community. Often times that will mean the companies are ‘close to the transaction’ or ecommerce related. These businesses can be capital efficient in spite of mobile development costs because they are able to generate meaningful revenues fairly early on and they can have great operational leverage which combined with high ARPUs allows them to generate significant profits from audiences in the tens of millions.

Lyst and MoviePilot are good examples of businesses from our portfolio that fit into this category.

Indicators that a startup idea is a good one

By | Startup general interest | 21 Comments

I always enjoy reading Paul Graham’s essays and his latest one How to get startup ideas is another great read. If you are an entrepreneur looking for ideas then I highly recommend you read the whole thing, particularly taking note of the common errors (building for a problem nobody has and trying too hard to come up with startup ideas).

I’m going to focus on one of the by-products of Paul’s essay which is a list of indicators that an idea might be good or an idea might be bad. Note that none of these are definitive evidence, they are simply signs. There will be great startups with some of the indicators of a bad idea and bad startups with some of the indicators of a good idea, but on balance good startups will have more of the good indicators and fewer of the bad ones.

Indicators of a good idea

  • It is something the founders want
  • The founders can build it
  • Few people realise it is worth doing
  • A small number of people really want it, right now, even if it’s a crappy version
  • It is on the leading edge
  • Looking back from the future the idea will seem obvious
  • The idea came because a problem was noticed, rather than because it was thought up

Indicators of a bad idea

  • Lots of people want it a little bit, but nobody wants it a lot
  • Lots of people want it a lot – if this is true someone would have built it already
  • Is in a domain the founders don’t know well

Involve developers in the design process: Another reason to have development in-house

By | Startup general interest | One Comment

Having development in house is really important for startups, principally because it is really hard to iterate quickly when using third parties. This is why technical co-founders are so sought after.

Being able to involve developers in the design process is another reason to have your developers on staff. David Royer at UIrrational gives three good reasons why:

  • developers have great ideas – they are typically early adopters of other services which gives them a lot of relevant experience
  • understand development trade offs earlier – at startups speed of execution is paramount and tight resources and deadlines make trade-offs a necessity – so make them early
  • developers will be bought into the design – and then they will work harder and better

Three great reasons.

London ranked best startup ecosystem in Europe

By | Uncategorized | 9 Comments


A new report out from the Startup Genome and Telefonica Digital ranks London as the seventh best startup ecosystem in the world. As you can see from the table above Silicon Valley comes out top and Tel Aviv is the only city outside the US that comes ahead of London.

There are always questions over methodology with reports like this, but the research takes data from 50,000 startups and seems pretty robust, although it would be good to know how the different components were weighted to produce the overall ranking (apologies if I’ve missed this, I did look). The ‘Startup output index’ is weighted for population size and maturity of startups in the region which I guess explains why some small cities rank highly on the list.

If you take out the smaller startup ecosystems then London would rank third behind Silicon Valley and New York, but my main interest is in what we need to do to improve. This research tells us that performance, talent, and staying up to date with trends are the areas to focus on. This rings true to me – better results from startups and their investors will bring more money into the ecosystem and help us raise the game generally, talent is the big enabler of performance and leading the world by setting trends is the raison d’etre for startups.

The good news is that I think we are improving in at least two of these key areas. Performance stats will look a whole lot better if/when companies like Mind Candy and Wonga go public, the quality and experience of the entrepreneurs we see increases all the time.

Diving into the detail of the report surfaces two other issues. We have a funding gap in London relative to the Valley and our entrepreneurs are less ambitious as evidenced by the size of their goals and the extent to which they consult on the side rather than commit fully to their startups. I think these are two sides of the same coin – if more cash was available for investment entrepreneurs would aim for bigger outcomes and wouldn’t have the same need to supplement their income.

For further reading check out this article on CityAM.

Some thoughts on leadership and excellence

By | Startup general interest | 2 Comments


This morning I went to a breakfast seminar hosted by PER, a private equity recruitment firm that I have known for a long time. Business speaker Jeff Grout gave a talk on leadership and excellence with a focus on sport and business. It was very interesting, and if you get a chance to hear him apeak I recommend you take it. There are a couple of takeaways I want to share with you.

Firstly a story he shared about Michael Johnson (pictured above). Michael won four Olympic golds as a 200m and 400m sprinter and he still holds the 400m world record. He is considered by many to be one of the greatest long sprinters in the history of track and field. He was a winner. To become one he crafted his dreams into ambitions, crafted his ambitions into goals, and crafted his goals into plans. The message is that winners are planners.

Translating this into a business context great leaders and high performing companies have a clear answer to the question to which every employee knows the answer: ‘what’s the plan?’. They know where they want to take the company and translate that into a smalll number of priorities which are clearly articulated to the staff. 3-4 is a small number. 6 is too many. Jeff cited some research from Cranfield Business School which found that having a small number of priorities was the biggest indicator of a high performing company. Steve Jobs was famous for keeping Apple laser focused and every company benefits from that discipline. Everybody kind of knows this, but it is surprising how many people choose not to implement it. I think the reason is that focusing on 3-4 priorities means choosing not to do a lot of stuff, which many find hard, particularly if they aren’t sure which areas of their business they should focus on to drive success. Additionally, I think that over confidence in their own execution skills often contributes – many people find it hard to believe that they can’t get their business to focus on six things at once.

Do me a favour now and take a few seconds to think about how many priorities your company has. Is it too many?

Next Jeff told us about Duncan Goodhew, the British Olympic gold winning swimmer. He realised his dream by articulating an outcome goal – to win gold, translating that into a performance goal – swim 100m breast stroke in XX seconds, and then translating that into a series of process goals that helped him train – eat well, improve stroke, etc. Goodhew set himself the target of winning gold in the 100m breast stroke when he was twelve years old. He knew he had to learn to swim much much faster and process goals were the tool he used to break down that big challenge down into manageable chunks he could focus on day by day. 

The lesson here is that many businesses are good at setting outcome goals – e.g. become the market leader, and performance goals – e.g. profit of £XXm this year, but fail to translate those peformance goals into appropriate process goals, and their people are left foundering.

This all links to yesterday’s post about culture. Johnson’s ‘dreams and ambitions’ motivated him in the same way as missions motivate companies and his plans are the equivalent of a company’s priorities. The difference, of course, is that Johnson is one man, whereas a company has many people. Keeping them all pulling in the same direction and working well together requires clear communication and commonly held values.

Conversocial’s company culture

By | Startup general interest | 3 Comments

Josh March, CEO of our portfolio company Conversocial has been thinking hard about company culture for the last couple of months and today he put up a great post describing the journey he has been through, and published the Conversocial company culture presentation (embedded below). I think they had around 7 people when we invested and one of the attractive things about the company was how committed everyone was and how much they liked working together, and it’s been great to see those strengths maintained as the company has grown to over 30 people.

I’ve blogged about company culture from time to time and regular readers won’t be surprised to read that I think it is a very important topic. I observe that great companies almost always have great cultures and that that investing early on in having a strong culture smooths the growth path for many high performing startups.

Here are some tips, generalising from the Conversocial experience and what I’ve seen elsewhere:

  • If company culture is going to be a management tool should be articulated early in a company’s life. Once there are more than 30-40 employees there will be a culture already that is hard to change –either one that management has chosen to help them succeed, or one that has arrived by default is most likely less helpful.
  • The culture as articulated should reflect the culture and values as is and not be too aspirational – i.e. the primary purposes of articulation are clarification, reinforcement and onboarding new employees.
  • The CEO and management must live and reward the values or they will be ignored.
  • A good test of whether a culture is meaningful is that it says as much about what people won’t do as what they will do (this is also a good test for company strategies). Anodine statements like ‘Be the best’ are empty because nobody wants to ‘be the worst’, but Conversocial’s first value ‘Put customers first’ has meaning because it tells employees not to put Conversocial’s or their own interests first. Josh says that thinking about the behaviour of people he wants to keep and promote helped him find Conversocial’s values.
  • Building a strong culture takes work and half measures are a waste of time – if the CEO doesn’t believe deeply in the power of a good culture then s/he is unlikely to put the time and effort required to get results. Josh personally takes every new employee at Conversocial through the presentation below so they can see how important it is.

Good investors understand all this. They are attracted to companies that are going to scale well and a good culture helps scaling. Remember that half measures don’t count.