Monthly Archives

July 2011

A developer and his app leave the Twitter ecosystem due to poor treatment

By | Twitter | 4 Comments

I have long been a user of the Topify service which provides email alerts to events on my Twitter account.  My primary use case has been to quickly see who has started following me and to efficiently follow some of them back by simply replying to the Topify email.

Unfortunately the service is being shut down on August 5th.  The email below explains why (highlights mine).


A week ago, without any prior notice, Twitter changed their backend resulting in removing headers from their emails which we used to provide you the Topify service. Once I discovered about this change, I asked on their official developers support forum about it and twice sent a mention to @TwitterAPI account. All of these were unanswered until today. Today they finally posted a clarification:

  • Many of the emails we send have X-Twitter* headers in them, with pieces of information about the event which triggered them. You might have noticed we’ve started decommissioning these headers.

    If for some reason you were using these headers programmatically in order to detect / process events, you should stop doing it and switch to one of the means supported by the API. For example, the Streaming API. Please let us know if you needed help or if you had questions!


I considered switching to using the Streaming API in the past, but the only option for Topify is to use the Site Streams version of it. But Site Streams are still in beta, and according to the documentation there is no estimated date for it to exit beta. Considering this last episode and other actions by Twitter in the past year, I have no desire to expriment with their beta offerings. Not only this can result in unstable service for you, they might just shut it down one day.

If Twitter (or any other company) wants third party developers to leverage their platform then they need to offer as much certainty and predictability as they can.  Without that rational developers will determine that the risk reward equation isn’t good enough and focus their efforts elsewhere, as Topify is now doing.

In this case that is to the detriment of both individual Twitter customers who had been using Topify (like me) and to the ecosystem generally as my levels of engagement will now take a hit.

It is unclear to me at this point whether Twitter really sees itself as a platform company or a consumer focused internet service.  In light of developments like this it feels more and more like the latter, in which case better notification emails and email based interaction with the service would be high on my list of feature requests.

Parody on raising capital in today’s market

By | Advertising, Venture Capital | 3 Comments

For a little amusement this Thursday check this video from Luma Partners which has a parody investment pitch from a DSP company (DSP stands for ‘demand side platform’ and is a category of Adtech business that was super hot last year).  I love the irony that even though no investor wants to invest in a startup that pitches in cliches we do very much want to back the companies which have the substance the cliches imply.  Watch out for the ‘did I say unique’ about two thirds of the way through.

For some reason I can’t get WordPress to embed the video, so you will have to click this Youtube link.

If you have 2.5 mins and want a bit of education on adtech/DSPs coupled with some humour then it is well worth it.

Hat tip to Alex Rahaman for the pointer.

Last night Riverbed acquired our portfolio company Zeus Technology for $140m

By | Announcement, DFJ Esprit, Exits | 2 Comments


Last night we announced the sale of Zeus Technology to Riverbed for $140m. This exit is the culmination of a lot of hard work by a lot of people over a long period of time. My involvement with the company dates back to 2004 when I led a recapitalisation of the business and there was eight or nine years of trading history before that.

I think we can identify a couple of characteristics of Zeus that contributed heavily to the success of the business in the period I was involved.

Firstly (and most importantly) the product has always been fantastic. We weren’t always selling a lot of it, but with very few exceptions the customers we had were very happy, mostly because the software worked really well. As former Chairman and CEO Paul Brennan used to say ‘it does what it says on the tin’. It was also very easy to install, so we had no headaches with complicated implementation projects or big professional services teams.

Secondly, we kept the business lean until the market was ready for us. Zeus competes with a number of much larger businesses and our unique selling point is that we sell software vs the competition’s hardware. That difference only became important to large numbers of customers when cloud computing took off.

There are many ways to build a successful business. This is the way that worked for Zeus.

It is common to joke that success has many fathers, but in the case of Zeus the list of people who deserve credit is indeed long. Key contributors since 2004 in the role of chairman or CEO were Paul Brennan, Paul di Leo, Michele Fitzpatrick and Jim Darragh. At different points over the last seven years each of them played a critical role in moving the company forward. I also want to thank the key members of the exec team who pushed the business forward to exit over the last couple of years: David Day (CTO), Charles Hobley (CFO), Kosten Metrewelli (Head of Marketing), and Pal Kosten (Head of Sales).

I hope I get the chance to work with all of you again.

Finally, I should mention the founders Adam Twiss and Damian Reeve who wrote the initial Zeus code when they were undergrads at Cambridge and ran the company during its early years. I didn’t get to work with Adam at all and only for a short while with Damian but their contribution was the most important of all (kind of obvious, but needs saying).

UPDATE: Thanks to everyone who has sent a note of congratulations.  It means a lot, even if I don’t reply in person.

Enhanced by Zemanta

Social networks engender trust

By | Facebook, Social networks | 3 Comments

One of the reasons that I’m a fan of social media is that I believe that Facebook, Twitter, blogs etc. are a powerful force for good by virtue of the transparency they bring to society and the relationships between people they engender (here are two posts I’ve written previously on this topic).  I realise that not everyone holds this view and that some people are worried about the implications for privacy and traditional offline relationships and so I like to blog when I see relevant research in this area.

This time it is another report from the Pew Research Center: Social Networking Sites and Our Lives, and a couple of the key findings were (as reported on FastCompany):

the Internet, in particular social networks, engender trust, and the more time you spend on them the more trusting you become

I think that is great.  One of the unfortunate features of late 20th century capitalism has been the decline in trust, which leaves people feeling more isolated and less fulfilled, so anything that reverses that trend has my backing.  I can almost hear the cynics arguing that trusting users of social networks are heading for a nasty surprise when someone steals their data and abuses that trust, but I just don’t see that happening, at least not on anything other than a very small scale.

Here is a little more detail on the correlation between internet use and trust:

As the report put it, "The typical Internet user is more than twice as likely as others to feel that people can be trusted," with regular Facebook users the most trusting of all. "A Facebook user who uses the site multiple times per day is 43% more likely than other Internet users and more than three times as likely as non-Internet users to feel that most people can be trusted." What’s more, while the average American has two "discussion confidants"–people they discuss important matters with–Facebookers who log in several times a day average 9% more close ties.

And finally a response to the fear that online relationships are being used to replace real world relationships:

The population on social networks has almost doubled over the past three years. Although some worry these online connections are being used to replace flesh and blood relationships, the Pew study found "little validity to concerns that people who use [social networks] experience smaller social networks, less closeness, or are exposed to less diversity." On the contrary, Americans "have more close social ties than they did two years ago," and "are less socially isolated."

Enhanced by Zemanta

Bringing transparency to venture capital

By | 50 Questions, Venture Capital | 6 Comments

There are two US VCs whose blogs I follow religiously, are perhaps my biggest source of inspiration in writing The Equity Kicker, and who I often quote here.  The first is Brad Feld and the second is Fred Wilson, and yesterday Fred posted a write up of Brad’s latest book Be Smarter than your Lawyer and Venture Capitalist, so I’m going to talk about both of them.

Brad’s new book isn’t available in the UK so I haven’t read it yet (I’ve pre-ordered my Kindle version, but I won’t get it until August 2nd), but from what I’ve read about it on Brad’s blog and from what Fred posted yesterday I’m sure it will be a high quality introduction to the venture capital process.

The contents of the book probably merit a post in their own right, and I will most likely write one after I’ve read it (as I did for Brad’s last book) but my point today is about what the book does rather than what is in it, and what the book does is shine a light on the workings of the venture capital industry.  Fred Wilson put it like this:

Venture Capital transactions have been a bit of a black art for a long time. It played to the advantage of the VCs because we do these things all day long all the time. But, like a marriage, it is not good for one side to take advantage of the other. It is not like selling a house. You have to live with the other side of the transaction after the deal closes. So over the past ten years, the VC industry has done much to increase transparency and trust with entrepreneurs. I’m proud to have played a part in that. Brad and Jason have done a huge amount of work in this area as well and their new book is going to be required reading for many. Well done.

Increasing transparency and trust between entrepreneurs and VCs is one of the main reasons I write this blog, and is raison d’etre for the 50 Questions series of posts and upcoming book.  As Fred says the situation is much improved over recent years, due in no small part to the rise of blogging and social media more generally, but I am regularly reminded (and occasionally surprised) at how far there still is to go.

I love feedback, and if there are any areas you would like me to write about or more general pointers for this blog, do let me know (if appropriate please check out the list of 50 Questions first though).

Enhanced by Zemanta

Revenue per active user – consumer internet benchmarks

By | Uncategorized | No Comments

The chart below is from Technology Review India.  Note that Groupon shouldn’t really be compared with the rest because its revenue per user includes the value of the products and services shipped.


Footnotes: 1.) The figure for Google users refers to the number of unique monthly search users, which doesn’t reflect all the people that see its ads and use its services. 2.) The figure for Groupon users refers to reported "cumulative customers" in 2010. 3.) The figure for active Zynga users refers to "monthly unique users" from October through December 2010. 4.) The figure for active Twitter users refers to a recent report from Business Insider that found that only 21 million Twitter users follow 32 or more accounts. Twitter considers an "active" user to be someone who is following 30 accounts, with a third of those accounts following back. 5.) Revenue figures for Facebook and Twitter are based on estimates from eMarketer, a research firm. 6.) Revenue figures for Zynga and Groupon come from their IPO filings with the Securities and Exchange Commission.

Enhanced by Zemanta

Even bankers believe web company prices are inflated

By | Uncategorized | One Comment

On Tuesday the Wall Street Journal reported the results from a survey of bankers which found that three quarters of respondents believe that the ‘valuations of many private internet businesses aren’t justified’.  The survey went on to ask what was driving the prices up, with the following results:

  • 34% believe it is down to demand for pre-IPO shares outstripping supply
  • 25% believe it is down to the growth of the internet itself
  • 22% put it down to businesses’ profitability
  • 19% thought good performance from recent internet IPOs is the cause

The first and last of these both refer to market dynamics and the second ‘growth of the internet’ is too vague and broad to be anything other than a catch all ‘we think the market is hot’ which means that only 22% of the respondents believe the valuations are being driven by fundamentals.

However, despite this lack of faith in the foundations of recent rises in valuations 62% of bankers also feel that there is the ‘possibility of a second internet stock bubble similar to the one seen in the late 1990s’.

From where I’m sitting that feels less likely today than it did six months ago.  A defining characteristic of bubbles is uncertainty over how long they have to run, but the macro picture of stretched government and consumer balance sheets coupled with lacklustre stock market performance and widespread feeling that current private company valuations are toppy leave me thinking that this part won’t run as long as the first internet bubble. 

My conviction on this point has been strengthened recently by persistent stories of bankers advising companies to raise money now because it will get more difficult again in the near future.

Disclosure: I genuinely believe everything I’ve written above, but as a VC it is in my interest to talk down the valuations of private companies, particularly at the moment as DFJ Esprit is very clearly a net buyer (we have recently had a run of good exits and closed a new fund).

Enhanced by Zemanta

50 Questions: What should I put in my business plan?

By | 50 Questions, Venture Capital | 3 Comments

Twenty-eighth in a series of weekly posts by myself and Nicholas Lovell of Gamesbrief which answer the fifty questions you should ask before raising venture capital.  We expect the series to run for a year after which we will collate the posts into a book.  You can find the rationale behind the series here, and the list of questions here.  We welcome your comments on any and every aspect of what we are doing.


The business plan has two purposes, to introduce your company to prospective investors and as a document investors who are in their due diligence process can look to for additional information about the company.

As I wrote in my last 50 Questions post, I wouldn’t advise that your business plan is the first document that you send to a VC, that should be an executive summary and covering email.  However, some people will ask to see a business plan before a meeting and for others the business plan will be the first contact they have with your business (particularly the partners of your sponsor) and so the business plan should work as an introductory document.  That means putting an executive summary at the beginning the contents of which should be as I described here (same post as linked above).  Note that the executive summary should stand on its own, i.e. the reader should be able to get an good high level feel for the whole business without needing to read any other part of the business plan.

Turning to the body of the business plan.  The first thing is to make it well structured.  It should have a good logical flow such that it reads well end to end and it should also be well structured so the reader can quickly dip in and find say the biography of the marketing director or the financial projections.

Content wise, there should be sections on the product, market, team, competition and financial projections.  Additionally, depending on the sector and characteristics of the individual company you might want sections on vision, go to market strategy, barriers to entry, unit economics, exit strategy, history of the company, use of capital, financing history, and anything else that makes your company stand out (in a good way).  Don’t go to overboard on the additional sections though and aim for a document of around fifteen pages.  If prospective investors are interested enough in your business to want more information than that they will ask.

For guidance on the contents of the individual sections please see:

CEO and board must be united

By | Startup general interest, Venture Capital | No Comments

Brad Feld put a great post up yesterday titled Note to CEOs: Decisions come from you, not the board which makes the point that it is a mistake for CEOs to blame the board for a decision that has been made, either because they don’t agree with it, or because they want to distance themselves from it.  Either way the CEO is abdicating some responsibility for the decision, and that undermines his or her authority and de-motivates anyone who sees what is going on.  Nobody tries as hard to do things they know their boss doesn’t believe in.

I’ve seen this play out in a large number of boards and problems have always followed.

The right way is for the board to present a consensus, including the CEO.  Behind closed doors there may be heated debate and even disagreement, but presenting a united front is a pre-requisite for good leadership.

Taking this line of thinking to its logical conclusion you end up with the following position (from Brad’s post):

I’ve come to believe that the only real operating decision that a board makes is to fire the CEO. Sure, the board – and individual board members – are often involved in many operational decisions, but the ultimate decision is (and should be) the CEO’s. If the CEO is not in a position to be the ultimate decision maker, he shouldn’t be the CEO. And if board members don’t trust the CEO to make the decision, they should take one of two actions available to them – leave the board or replace the CEO.

Backing up a little, the overall role of the board is to make sure the company has the right strategy and is appropriately resourced from a cash and people perspective.  So there are bunch of critical non-operating decisions that the board is involved with, but as Brad says with the one obvious exception all operating decisions are the CEO’s.

I’ve been a little brief here, and if you are into this topic, or want the link between the board presenting a united front and fact that the CEO must be the ultimate decision maker then reading Brad’s post (and the comments) will help.

Enhanced by Zemanta

The consumerisation of enterprise software

By | Enterprise2.0, Startup general interest | No Comments

The consumerisation of IT is a much talked about theme of the last ten years or so.  The definition per Wikipedia is:

Consumerization … describes the trend for new information technology to emerge first in the consumer market and then spread into business organizations, resulting in the convergence of the IT and consumer electronics industries, and a shift in IT innovation from large businesses to the home. For example, many people now find that their home based IT equipment and services are both more capable and less expensive than what is provided in their workplace.

A newer but perhaps unsurprising trend is for enterprise software companies to start adopting the disciplines of consumer internet companies, or the ‘consumerisation of enterprise sofware’.  I had an inkling that things were heading in this direction back in 2007 when I wrote about edge-in adoption of social software at the enterprise, and now I think they are finally getting there.

Here is a list of the relevant ‘disciplines of consumer internet companies’:

  • Focus on creating amazing user experiences
  • Key metrics are customer lifetime value, customers acquisition costs and customer usage
  • Daily/weekly/monthly iteration of the product based on observations of customer use
  • Launch with minimal feature set (minimum viable product)
  • Low cost development leveraging open source software and the cloud
  • Minimal expense from the customer to get started
  • Minimal time commitment from the customer to get started
  • Great customer support

Companies employing these disciplines are winning because they offer a better service for the end customer.  Too many IT projects still fail and by following the above disciplines new wave enterprise software companies reduce both the chances and the cost of failure.  Two big ticks.

This doesn’t necessarily mean that selling has got any simpler though, at least to large enterprises who still have to run complex processes to ensure internal consistency, value for money, and compliance.  Whilst it is undoubtedly becoming easier to make small sales via the web, enterprise sales folk are still needed to land big deals. has expensive enterprise sales people and so will just about every other large software company (it is pretty much impossible to be a large software company without selling large enterprise deals).  Ben Horrowitz did a good job of explaining why on Techcrunch:

large companies employ complex processes to ensure that major purchases make sense. These processes generally span many different organizations and stakeholders. It is not unusual for a purchasing decision to include people from many different IT departments (e.g. development, security, operations) and business functions (e.g. Finance, IT, Legal). The decision often involves technical decision makers, economic decision makers, and risk management decision makers.

Often these processes are so complex that almost nobody inside the company knows how they work. Excellent enterprise sales reps will guide a company through their own purchasing processes. Without an enterprise sales rep, many companies literally do not know how to buy new technology products. A top notch enterprise sales person not only knows her customer’s process better than the customer, but will be skilled at characterizing the value of her product to each decision maker independently. This will involve product demonstrations, proof of concepts, complete return on investment analysis and even competitive positioning. The sales rep will work with the various constituents to help characterize the value proposition to their management teams.

It is difficult to do all that on the web.

Enhanced by Zemanta