Monthly Archives

June 2012

Not all accelerators are worth it

By | Startup general interest, Venture Capital | 7 Comments

Earlier this week ReadWriteWeb reported on a study into the 29 North American accelerator programmes by Aziz Gilani, a director at DFJ Mercury, one of our sister funds in the DFJ network. They found that 14 of those programmes hadn’t produced a single graduate that went on to raise venture funding.

There are a number of problems with this statistic, probably the most important of which is that raising venture capital is not the only measure of success, but it is still indicative of something I have felt for some time – that there are many startup accelerators which offer a bad deal to their companies. To be clear, I think the best accelerators offer tremendous value, Paul Graham and Y Combinator have practically re-invented very early stage financing and here in Europe Springboard, Seedcamp, StartupBootcamp and maybe now Wayra are doing a great job (let me know in the comments if I’ve missed any) – but there is only space for so many accelerators.

There are two finite resources which limit the number of quality accelerators the market can bear:

  • Inspirational founders for the accelerator programme
  • Availability of quality mentors

Accelerator programmes can only help companies accelerate if they provide good guidance and/or access to investors yet there aren’t many people like Paul Graham with the experience, coaching skills and networks required to really help startups. Access to mentors is even more problematic – I know that the folk here at DFJ Esprit have to turn down more opportunities to mentor than they can say yes to, and I’m sure the same is true for other mentors in the startup ecosystem. We choose to mentor at the programmes with the best companies and I’m left wondering what life is like for startups at the other programmes.

The ReadWriteWeb article advises that for first time entrepreneurs participating in a top programme is a no-brainer, always assuming they can get in. For the second tier they say the following:

But what about a second-tier or third-tier accelerator? “When you’re a young startup and you don’t have a lot of cash, you have one currency, your equity,” Gilani says. “So treat an accelerator like any other service provider. Be rigorous in your diligence. Or make the decision not to join one.”

Ask three questions of your accelerator:

  1. Will it help you get follow-on funding?
  2. Will it help you form partnerships with other companies or accelerate your growth?
  3. Does it have proven mentors who will help you get traction?

If the answers are no, it’s probably best to steer your startup in another direction.

I think their advice is on the money.

Conversocial on CNBC talking social customer service

By | Conversocial | One Comment

Josh March, CEO of our company Conversocial was interviewed on CNBC yesterday, 4m video embedded below. Conversocial has a SaaS product which enables companies (primarily retailers at this point) to service their customers over social media and their market is really starting to open up. In this interview Josh does a great job of explaining how an increasing number of companies (including Tesco, McDonalds, Groupon, Sephora and a large percentage of the other smart and innovative ones) have understood that their customers are demanding to be serviced on Facebook and are staffing and buying software so they can give those customers what they want. The implication, of course, is that everyone else should follow, and if they do then as the leading vendor in this market Conversocial will benefit.

I’ve written before about how Conversocial conducted primary research as a PR tool, and Josh is makes good use of statistics in this interview. My favourites comes at around 2min 40s and they show the momentum at Conversocial – accounts and comments processed have both doubled since the beginning of the year to 7,416 and 106m respectively, and the total fans of pages managed using Conversocial software is up over 50% to 545m.

“This decade will be the most innovative in history”

By | Innovation | 2 Comments

If you believe, as I do, that the pace of technological change is increasing exponentially then you would expect that most every new decade is the most innovative in history. In fact, it would be very bad decade that wasn’t.

However, looking ahead and seeing where the innovation will come from isn’t always easy, and many people think that we’re running out of ideas. Vivek Wadwha of the Singularity University wrote a great article for Forbes this week titled Why I Believe That This Will Be The Most Innovative Decade In History which sets out why they are wrong and where the innovation will come from.

Here are the highlights:

  • Genetics (1) – the cost of sequencing the human genome will be less than $100 within five years at which point genome data for millions of people will become available and correlations between DNA and disease will be used in diagnosis and to prescribe personalised medicines, bringing a revolution in medicine (DFJ Esprit portfolio company Horizon Discovery is one of a number of startups leading this revolution)
  • Genetics (2) – advances in synthetic biology are allowing “researchers, and even high-school students” to create new organisms and synthetic life forms. I don’t know where this goes, but the potential seems huge. Synthesising DNA from fireflies and trees to create trees that glow in the dark to remove the need for electric street lighting is one idea that gets mentioned a lot. There must be more/better ideas than this though.
  • 3D printing – I’ve written a lot about 3D printing here already and regular readers will have a good feel for where this technology is today. Vivek expects that within this decade we will see 3D printers doing small scale production of previously labour-intensives crafts and goods and next decade we can expect local manufacture of the majority of goods, and 3D printing of buildings and electronics. I write about 3D printing a lot because I think it will have a transformative impact on a number of industries, from toys to cars.
  • Nanotech – advances in this field are making it possible to build inexpensive sensors with a variety of applications, perhaps the most interesting of which will be healthcare – I’ve also written about this before. In the short term the most exciting of these will be consumer focused services which bypass traditional healthcare channels. Fitbit is a great example.
  • Artificial intelligence – in 2009 IBM modelled a cat brain in silicon and last year their Watson computer won the US general knowledge quiz show Jeopardy. We also have the beginnings of intelligent personal agents and self-driving cars on the roads. I expect both of these to be common place and genuinely useful by the end of the decade.

In the last fifteen years the internet has changed our lives beyond measure. Expect more even more change in the next fifteen years.

Marketing becomes media

By | Advertising | 9 Comments

Somebody recently described Red Bull to me as a media company that happens to sell beverages. I think we will soon be able to say something similar about many of the world’s most successful consumer brands.

For those that don’t know, Red Bull has grown from next to nothing to a global force over the last twenty years in energy drinks –  a category they pretty much created. Their success came partly because they have a great product, millions of people LOVE Red Bull, but partly because they have built their brand with a series of innovative events and amusing adverts. Their flagship FlugTag events around the UK are fun and memorable, as are their ‘Red Bull gives you wings‘ ad campaigns.

Turning to fashion, two British of the more successful British startups over the last five years, Net-a-Porter and ASOS, have built their businesses by blending content with commerce – i.e. a mixed media and retail business model where media offsets traditional marketing spend. ASOS in particular stands out as having achieved significant success (£1.3bn market cap, TTM revenues and pre-tax £495m and 330m) on minimal marketing spend.

Just about every other brand is now headed in the same direction, catalysed by their move into social media. As they seek to engage customers in places where they are hanging out to be entertained rather than transact they are having to balance their sales oriented communications with content that is interesting in its own right, i.e. they are having to become media companies. To get specific, social media marketeers are now offering the following sort of advice:

  •  don’t ask, “How can we convert fans into sales?”—instead ask, “How can I provide value to my fans?”.  Then, and only then, will fans actually consider making a purchase as a result of your social marketing efforts.
  • —for every 5 posts you make, only ONE post (20%) should be DIRECTLY about your brand/product (i.e. “Buy our product”, “Visit our website”, etc).  The other four posts (80%) should be RELEVENT to your audience and provide VALUE in some form—informational, funny, interesting, etc.—but they should NOT be directly about YOU
  • Find 10 blogs (do a simple Google search i.e. “popular food blogs”) relevant to your audience.  Subscribe via RSS and share the most interesting / entertaining posts with your fans
  •  you’ll need to differentiate between the ones that perform well and those that perform poorly on your Page.  There are many ways you can do this, but I’d recommend simply running the numbers in a spreadsheet

The advice in these bullets could easily have been directed at a newspaper journalist. Although maybe not at a quality paper 🙂

Voice analysis can spot Parkinson’s disease

By | Innovation | No Comments

Parkinson's Voice InitiativeOr at least that’s the hope.

Dr Max Little and his team analysed time series voice recordings from fifty patients with Parkinson’s and identified differences in voice patterns which they hope can become the basis of a tool for preliminary diagnosis. This is a big deal because there is no blood test for Parkinson’s and current tests are expensive and time consuming.

They are now collecting voice samples for further research. You can leave one via a three minute call to UK 01865 521168 (numbers for other countries here). I just left mine and it was very easy, although having to say ‘aaaaaaah’ for as long as I could two times over got me some funny looks here in the office.

This whole project is brilliant. Firstly because it is an amazingly elegant solution to the problem of diagnosing Parkinson’s and secondly because the the next phase of research is so efficient. They are targeting 10,000 samples, which is a small number in today’s connected world, and I imagine they will get that and more very quickly and the cost will be limited to the telephony system.

Possibly more exciting is the notion that voice tests could help us to detect Parkinson’s in patients before it does irreversible damage, and that the method might extend to other neurological conditions.

I can see a world in the not too distant future where voice samples are regularly collected and screened as part of a generally available preventative health regime. The costs could be that low.

You can read more about this on the BBC and on the Parkinson’s Voice Initiative website.

Don’t let email kill your productivity

By | Startup general interest, Venture Capital | 3 Comments

Since we had the kids a shortage of hours in the day has become one of my biggest challenges and I’m now a total sucker for a bit of productivity pr0n, and Ryan Carson’s blog post of yesterday gave me a good thrill. Unless managed well email can be a huge time sync and Ryan’s number one tip on his list of 7 do’s and don’ts for founders is to ignore email:

Email-creep is enemy #1 to your productivity. It’s a never ending list of things other people want you to do.

Never check email until you’ve completed at least 2-3 things on your priority list of todos for the day.

Once you decide to check email, follow this pattern:

  1. Quickly scan your inbox for important emails and mark them as ‘Priority’ (using a tag or star).
  2. Mark any other emails that are important, but not time sensitive, as ‘Later’. I use tags in Gmail for this.
  3. Archive them all. This is important as it keeps you from feeling overwhelmed by large amounts of email in your Inbox. Note: This means you won’t answer every email you get. This is an important realization because it frees you up to do things on your priority list.
  4. Blaze through the ‘Priority’ pile. Spend a maximum of 30 minutes on this.
  5. Ignore the ‘Later’ pile until you are stuck somewhere without an internet connection (train, plane, park bench, etc).
  6. Repeat twice a day

As your company grows, make sure you communicate this to your team and ask them to follow a similar procedure. Before you know it, everyone in the company is doing email all day instead of getting things done on the Road Map.

FYI, I’m not referring to customer support email. If you have customers, you need to get back to them as soon as possible and then hire a full-time customer support person as soon as cash will allow.

For some years now I’ve been managing my email along the lines that David Allen recommends in his great Getting Things Done: The Art of Stress-Free Productivity, and I was pleased to see that my email processes are pretty much in-line with Ryan’s. Not looking at email for long periods of time (also recommended by Tim Ferris and Marc Andreessen) is hard at first but yields real benefits.

Ryan’s last two paragraphs, however, introduced two new ideas to me. Firstly, we should all be concerned with email productivity across our whole companies, not just ourselves, and secondly customer email should be separated from other email. This second point is interesting in the context of venture capital where our customers are our LPs (the people who invest in our funds) and entrepreneurs who we are trying to sell our cash to. I always try to get back to entrepreneurs quickly, but sometimes I can be a little slow, and often the reason is that their emails are handled in a single process with all other emails, and many of them get put in the ‘process later’ bucket which takes me a long time to get to. They don’t belong in my ‘urgent’ bucket, but maybe I should create a new bucket that sits between ‘urgent’ and ‘process later’.

Rock Health’s latest crop of startups

By | Healthcare, Innovation, Venture Capital | 11 Comments

Rock Health, probably the leading healthcare focused accelerator programme announced their latest class of companies this week. They are profiled on Techcrunch here and briefly summarised below.

I’ve written a reasonable amount on here about the coming opportunity in health and as regular readers will know I think the most exciting opportunity is in consumer oriented healthcare services delivered online and over mobile. Many of these will be cheap enough that they can bypass traditional health distribution systems (e.g. the NHS here in the UK and insurance companies and doctors in the US) and many will leverage smartphone technology for diagnosis or data capture purposes. The big question for this investment thesis is unsurprisingly timing. I am confident we will see a wave of companies coming, but the key to making successful investments is to time your activity with the first burst of market growth. One way to think about that is when penetration moves from say 0.25% to 5% – that is typically when companies can grow really fast and valuations sky rocket as analysts and incumbents come to recognise that a major market shift is underway. Founders need greater prescience and will be best placed start their companies a year or two earlier, but probably not five years earlier.

So when I saw the Rock Health list of companies I read the profiles in search of pointers about the timing of the market for my hypothesis about the opportunity in healthcare, as well as for signs of other potential investment theses.

Going through the list of 13 companies I found 4-5 that are consistent with my hypothesis (Agile DiagnosisCardiioCare At HandSessionsSano Intelligence) which I take as evidence that the timing for this market may well be now.

The Rock Health companies

Achievemint – rewarding consumers for healthy behaviour – with real world goodies paid for by insurers

Agile Diagnosis – web and mobile app for doctors to improve speed and accuracy of diagnosis

Avva – patient focused cancer management tool – helps patients and families get the information they need

Cardiio – technology that turns a smartphone camera into a sensor – e.g. to measure heart rate by looking at someone’s face

Care At Hand – monitors patients at home – e.g. to report on non-compliance

ChickRX – health Q&A site for women, will also sell products

Cognitive Health Innovations – service allows therapists to communicate with their patients via mobile and social

Docphin – health information system, initially providing doctors with access to medical news and research

HealthRally – crowdfunding platform to enable friends and family to reward healthy behaviour

Medmonk – a tool to help pharmacists find discounts for customers who cannot afford to pay

Nephosity – medical images on the iPad

Sessions – a personal exercise coaching platform

Sano Intelligence – wearable patch that tests the blood via micro-needles and sends results to doctors at regular intervals

Chris Dixon on firing people

By | Startup general interest | 8 Comments

It is an unfortunate fact of life that hiring is an inexact science and mistakes will get made, even if the recruitment process is best in class and executed with diligence. It is simply impossible to know for sure how people will work out until you have been working with them for a while. On top of that companies change as they grow and there are some people who are great at the early stages but aren’t so good when process and discipline become important.

So even the best startups and founders will most likely have to fire someone. In two of my recent investments the short company history up to the Series A included parting company with a senior member of the management team and I took that as a positive. Firing someone senior in the early stages of a company’s life is espescially difficult and I was pleased to see the founders in question had the courage and discipline to take a hard decision.

I’m writing all this because I’ve just read an excellent post on the topic of firing people from Chis Dixon, a prominent US entrepreneur and angel investor. It’s so good I’m going to quote it in full:

Firing is awful. You can try to avoid it, but even the most selective founders make serious mistakes. Here are a few things I’ve observed about firing:

1) The good people bounce up, the bad ones bounce down. I was told this by my boss once when he was firing one of my friends. At the time, I thought this just made him feel better about himself. Over time, I’ve seen the wisdom in what he said. Some people who get fired react by fixing their weaknesses. Others spiral down.

2) Do it early. If you think you’re going to fire someone over the next six months, you probably will. Don’t wait too long. Too many founders do. It’s better for management and employees if it happens fast.

3) It’s awful. You’re in control of a situation that will meaningfully hurt someone. It’s an awful place to be. The fired person will go home and tell his/her family about how terrible it was. It was your fault. Perhaps your mismanagement caused it. Who knows. You’ll question it, and perhaps you are right to do so.

4) The other choice is firing everyone. You’re the founder of the company. If you run out of money, you’re forced to fire everyone. If you don’t fire the bad employees, you risk everyone else’s jobs. It’s an impossible situation.

5) The feeling is more likely to be mutual than you think. Most of the time, the person getting fired was already about to quit. The antipathy you feel is likely reciprocated. It’s surprising how often this happens and management doesn’t see it coming.

It would be great if startups were all about growth, hiring, and success. But the reality is that founding a company is a brutal job and lots of the pain gets passed down to employees. Creative destruction sounds nice in textbooks, but in the real world it means telling friends to go home, stop getting paid, and find new jobs.

 I also want to bring out two of the comments:

If you’re doing your job firing someone should never come as a surprise to them. I’ve fired pleanty of people over the past 20 years and the vast majority are still valued contacts and many are even friends. Being proactive mitigates 90% of the issues of firing someone but that makes it no less difficult to do…particularly the first time you do it!


The amount of build-up it usually requires for you to even approach the subject of firing someone prevents you from making a mistake.Firing someone sucks. You don’t fire people “on the bubble” in your mind. You fire people as a last resort.

This last comment is important. Firing people is a last resort. The right thing to do for the individual and the company is always going to be to try and find a way to make things work. In some cases it won’t be possible and then the right thing to do is to quickly recognise when that is the case and take appropriate action.

Simple charging models work best

By | Business models | 4 Comments

I’m a big fan of simple charging models. I’m at LeWeb London right now listening to the CEO of Evernote who describes his business model as direct – he only makes money off his users and then only if they use the product (for those that don’t know Evernote operates a freemium model where the free version of the software is quite rich and advanced users can pay for a variety of advanced features – I use the free version). This has a beautiful simplicity that is easy for customers to understand and works well if your product is as fantastic as Evernote’s.

Other simple models that have been very successful include Google’s search advertising and Salesforce’s subscription software licensing.

There are plenty of successful companies that don’t have simple business models. Sometimes that is because there isn’t an obvious simple way to go – Facebook springs to mind as a great example. However, there are other companies, and sometimes industries, that choose to have opaque business models because they can make more money when it is harder for their customers to understand exactly what they are getting and what they are paying. These companies can be successful in maximising their revenues over the short term, but confusion stifles innovation and is rarely the best long term answer for them, and it certainly isn’t good for society at large.

The mobile industry has long used bundles of minutes, data, texts, handsets, contract tie in and free gifts or cash back on sign up to maximise revenues and innovation was stifled as a result. We didn’t see meaningful innovation in handsets until Apple used its clout to make carriers adopt the iPhone and confusion over data charges slowed adoption of the mobile web. In order to encourage adoption carriers then flipped to all you can eat data plans, which were great for consumers, but resulted in a dislocation in value received and cost of service provision, and carriers started looking to politicians to force providers of web services to subsidise their operations and we had the whole net-neutrality debate.

The simple answer is for carriers to charge on the basis of network usage and I was pleased to see that last week Verizon announce a plan that operates in exactly this way – wish users paying more if they use more data. They still have a proportion that is bundled as there is a significant line rental element and unlimited free calls and texts, but this is a step in the right direction. When cost of provision is aligned to value derived then the market operates efficiently and appropriate investments get made.

Microsoft and the future of software pricing–when free isn’t quite free

By | Business models, Open Source | 2 Comments

The excellent, but paywalled, Lex column in the FT has an interesting article today which speculates on how Microsoft will price its forthcoming releases of Windows 8 (tablet edition) and Microsoft Office. Software prices are dropping and the importance of local operating systems and having locally installed productivity software is declining which makes the future for Microsoft’s consumers software a challenging one. The interesting question is how challenging.

The biggest competitor in mobile, Google’s Android is free, and Google docs is getting better every year (although in my opinion still not up to the standards of Office for anything complicated or difficult). To keep their burn low many startups I know eschew Microsoft entirely and run on Google’s free software. For these reasons I’ve long thought that Microsoft will lose their monopoly positions in OS and productivity software and is in for a tough time.

I still think that is true but they may be able to hold pricing better than I had previously assumed.

The interesting new news, as reported in the FT, is that Bernstein Research have analysed the hidden costs of working with Android (software integration, support costs and royalties for the parts of Android that are based on Microsoft’s intellectual property) and estimate that if Windows 8 is priced at $40 per tablet it could actually be cheaper for tablet makers who ship fewer than 9m units a year, rising to $50 if Office is bundled in. Apparently that is broadly comparable with what Microsoft charges PC makers now.

If Microsoft can maintain profitability whilst their market share declines then they still have an interesting paid for software business.

The centrepiece of the ‘free’ argument is that the price of everything drops to its marginal cost of distribution. In theory software delivered over or downloaded from the web has a marginal cost that is very close to $0 and a lot of people believe that software companies of the future will derive their revenues more from support, services and hosting than licenses.

The Microsoft analysis highlights that in practice their are other costs (software integration, royalties) which can keep the marginal cost significantly above $0 hence allows for successful license based business models – this is a rehash of the total cost of ownership argument that software companies like Oracle and Microsoft have been making against open source for some time now. $40-50 isn’t that much though, and software companies with license business models will need to keep their costs low and find efficient routes to market if they are to make good profits and become very valuable.