Celebrating Execution


I am sitting at a board meeting, observing how a portfolio company of ours is out-executing its competitors by a wide margin.  Surprisingly, the primary narrative in tech startups is innovation, invention and capital leverage, and not on strong execution.  However, in my experience, strong execution has been biggest success factor.

In 2001, when Jim Collins published his influential book, Good to Great, he showcased how some companies were out-executing their competitors.  The difference between Gillette andWarner-Lambert was remarkable.  Similarly, it’s not difficult to see how much better Walmart has been out-executing Sears/K-Mart for decades.  A look at the stock chart provides a quick reminder.


By contrast, in the tech industry, the discussion digresses to innovation silver bullets, very quickly.  The tech press looks for headline-making bold claims.  VCs try to see how a company’s business model can create an unfair advantage.  Founders, consequently, come to us pitching some clever shortcut they have figured out.

No one questions what a big difference execution has made in the Walmart -Sears case, but when you argue the same for an e-commerce company, eyebrows go up: “Are you sure?  Was it not through a high-tech warehouse, or drone delivery?”, is the question you get.

We see over and over that the winners usually get there by fanatical execution: sweat, hard work and resilience.  I think we need to celebrate execution more in the tech community.

Merging into Monopolies


The big news this morning is that Uber China and Didi Chuxing are merging. The rivalry was one of the biggest recent tech battles, with both parties well armed with billions of dollars.  I don’t think there’s a doubt in anyone’s mind anymore about who will win the massive Chinese ride-hailing/ride-sharing market.

Come to think of it, I think the same question is now pretty much answered for the global market, as well.  It would be rational at this point for Uber to acquire Lyft for probably about 10% of its enterprise value, and get rid of the last credible threat to its US dominance.

The picture in Uber’s market reminds me of Peter Thiel’s much discussed 2014 article about monopolies.  I suspect we will see other examples of merging into monopolies in markets with strong network effects.

Enterprise Sales in Turkey


If you are a technology startup targeting Turkish enterprises, you’d better bring your magnifying glass with you.  Here’s an unpopular fact:

Turkish enterprises are tiny!

Let’s take a closer look.  On the 2015 Turkish Fortune 500 list, #500 is a company called Metag.  Annual sales: 284m TL (about $100m).  By contrast, the #500 company in the US is Burlington Stores with $5.1b in sales, 50X the size of Metag.

Another way to look at this would be by absolute size.  If you determine that your target market is companies with >$1b in sales, your target list would have only 40 companies in Turkey.

The US economy is about 20X the size of the Turkish economy ($16tn vs $800b), so you get the picture. It’s not just the relative market size.   There are a lot more enterprises to sell to in developed markets, and each one is probably a more sophisticated buyer of your services, with larger budgets.

Big Numbers


The press loves big numbers.  Why? Because, the readers love big numbers.  Look at the headline below:

FireShot Capture 10 - Meet the Man Who D_ - https___mail.google.com_mail_u_0_#inbox_1552f8f3a973098a

There are two very large numbers, $5.5 billion and $62.5 billion, referring to two highly ambitious and rapidly growing companies, Lyft and Uber.  I am a fan of both startups.  However, they are both quite tiny in the global enterprise scale.

When you say an $X billion company, one understands that it’s a top-line revenue number.  Apple is a $230 billion company and Walmart is a $480 billion company, in general business speak.

I would also understand referring to these businesses by market cap.  It’s the company value of these businesses based on very liquid market information.  By this measure, Apple is a $450 billion company and Walmart is a $220 billion company.

You’ll notice that these two sets of numbers are almost exactly asymmetrical, so it’s always critical to clarify what one means when the phrase “$X billion company” is brought up.  But I find both to be acceptable and practical.

Then, there is the startup version, with which I have a problem.  I mentioned that both Lyft and Uber are quite tiny.  To be specific, lats year Lyft had a rumored ~$100 million in revenues.  Uber, which is larger, had its revenues estimated at $1.5b billion.  Both numbers are very small compared to those of Apple and Walmart.  Bu this is not stopping the press to refer to them as $5.5 billion and $62.5 billion companies.

Why? VC round valuations.  These two companies, have been valued, in private rounds, details of which are not announced, at the larger numbers mentioned above.  So, all of a sudden, you find that you can get away with calling Uber a $62.5 billion company. And, jaws drop.  Your readers are awed. Links get clicked on.  Young entrepreneurs pull out their calculators and start multiplying their own metrics with the multiples they can extrapolate from Lyft and Uber numbers.  And, expectations rise, appetites get whetted…  And, bubbles form.

Because, the press loves big numbers.



The Disconnect in Online Advertising

Online ads suck, but it’s the main driver of today’s internet.

If you did not already know that, a great place to look for confirmation is Mary Meeker’s 213 slide Internet Trends deck, in which, slides 41 thru 110 are dedicated to the examination of how different forms of online ads are growing and that it’s accelerating.  It’s depressing really.  And the money slide is this one:

FireShot Capture 6 - 2016 Internet Trends Report_ - http___www.slideshare.net_kleinerpe

In summary, it will get worse before it gets better.

This is confusing to me.  Because, on the other side of the internet coin is that high quality, frictionless offerings are winning on the internet.  Look at the most successful apps.  Booking.com is eating the lunch of other hotel OTAs, primarily because it works so well.  Uber is a lesson in on-demand UX.  Airbnb’s user experience has helped its massive global network effect, in taking the big lead it has in global marketshare.

Ads, on the other hand, are pulling in the opposite direction.  1010!, a great game, is pissing me off with 5-second video after video, trying to get me to accidentally click on an app install.  And, I am not alone. Back to Meeker’s deck:


81% = Mute Video Ads
62% = Annoyed with / Put Off by Brand Forcing Pre-Roll Viewing
93% = Consider Using Ad Blocking Software

Here’s the slide:

FireShot Capture 7 - 2016 Internet Trends — Kleiner Perkins C_ - http___www.kpcb.com_internet-trends

This disconnect between how fast online advertising is growing while remaining a nuisance for internet users has made it tough for us to build strong investment theses for ad driven business models.  It’s also probably why we get turned on by technology that looks to improve the efficacy in the sector.

Microchunking of Apps


I find myself thinking these days about how the app universe will evolve.  There are quite a few ideas and analyses on some trends that hint at what’s ahead of us:

The common thread through these is the reduction of applications to their smallest viable (and functional) form, while remaining accessible through multiple platforms, including devices and social media, and maintaining their ability for transactions.  In essence, we are going through “microchunking” of applications, to borrow Fred Wilson’s term referring to the content revolution we saw with the Web 2.0 wave last decade.

I see two areas to watch as this trend develops.  One is the need for flexible SW development frameworks.  I suspect these will be open source. But there will be support and services opportunities around them that should allow for creation of sizeable businesses.

The second is a similar opportunity on the operations side. Think of it as a set of standards providing context and interoperability for microchunked apps.  As applied to the need around human transport, an example may be standards for my mobile device to signal that :

  1. I am at point A.
  2. I need to get to point B.
  3. I am willing to pay X for this service.

This signal can then get acquired and processed by an open marketplace, and matched with the supply side signals from providers, whether they are taxis, Uber drivers, etc.

I am keenly waiting for Open Bazaar to see how it fares in the wild.

Intentions and Context

Girl With Mobile Smart Phone

Ben Evans penned his best post of the year last week, which has kept me thinking on it since.  One problem I have with it is its title, though.  I have trouble distinguishing between mobile and the web at this point.  What we call phones refer to a fraction of the utility I receive from mine, and any business that creates its value on the “connected world” (my preferred noun for what includes mobile, web, the internet, etc.) needs to think of these platforms in a unified manner.

The topic is a vast one.  It touches on

  • programming frameworks
  • the future of native apps vs. evolved browsers
  • whether we will continue to rely on legacy input/output interfaces such as mice, keyboards, touch screens, LCD displays, etc.
  • how we navigate and ultimately get to where we need in the massive world of data and information.

I am not smart enough to wrap my mind around what we have coming. I do, however, notice that the final point on the list above is exciting a lot of entrepreneurs.  It’s the quest for the new search paradigm. Google’s Page Rank was a great benchmark, in its speed and accuracy, for our expectations on how our information needs are met.

For the next generation of solutions, Fred Wilson is proposing that we rely on contextual runtimes.  I like his examples:

If I’m building a lunchtime meal delivery service for tech startups, that’s a Slack bot.

If I’m building a ridesharing service, that’s going to run in Google Maps and Apple Maps.

If I’m building a “how do I look” fashion advisor service, that’s going to run in Siri or Google Now.

If I’m building an “NBA dashboard app”, that is mostly going to run on the mobile notifications rails.

I find myself thinking about where context will reside, whether on the cloud or the mobile OS, and how much will be implicit versus explicit.  Essentially, we are looking for better ways to sniff for intentions, the way Google persuaded us to type it into a search box.