Category: trends
Posts
map[]
Since writing The AI Agency: A Novel GTM for Machine Learning Startups, I’ve been meeting many companies who operate this way. These startups use machine learning to disrupt an industry traditionally dominated by agencies: law, accounting, recruiting, translation, debt collection, marketing…the list is long. I will publish a landscape soon on the area. If you’re operating an AI Agency, I’d love to hear from you.
In meeting many of these innovative businesses, I’ve observed they face four strategic questions.
map[]
In the car yesterday, I listened to an interview with Eugene Wei, who described the dynamics of consumer technology as if they were plain as day. It’s rare to hear someone lucidly describe complex topics like the reasons for Twitter’s success and struggles. Eugene wrote many of his ideas in this post though there are some ideas that I found only in the podcast.
Eugene was an early PM at Amazon, SVP of Product and Marketing at Hulu, Head of Product at Flipboard, and Head of Video at Oculus.
map[]
In 2015, I wrote about the trade-off facing vertical SaaS companies. Vertical SaaS companies focus their efforts on a particular group of customers. Procore targets construction with their software and Veeva targets pharmaceuticals with their CRM. This concentration limits the market size, but improves product market fit. Both of those businesses are now worth more than $3B. There is a new twist in SaaS with a parallel dynamic.
I’ve started to call them AI Agencies.
map[]
Reading the news in the past week made me wonder. Just where are we pricing SaaS companies today? The Nasdaq and the S&P have toyed with a bear market. Many stocks are down 10 to 50%. Absolute valuations are one consideration, but let’s understand it at a deeper level. Have multiples compressed?
The answer is yes, they have, but enterprise value to forward revenue multiples are still at some of the highest levels for SaaS companies in the past eight years.
map[]
Here are some predictions for 2019 and a review of my thoughts for 2018, many of which were wrong.
The M&A market slows meaningfully, especially at the multi-billion dollar level. The recent seasickness in the public markets forces most CEOs adopt a more conservative approach to acquisitions. Facing large swings in valuation, these leaders may struggle to advocate and articulate that large acquisitions are accretive and will be immediately rewarded by share appreciation after an acquisition.
map[]
Another week, another blockbuster software acquisition! This time SAP has agreed to purchase Qualtrics for $8B. Qualtrics is a Utah based provider of experience management software. Qualtrics writes and sells software to ask questions of employees and customers to help businesses improve customer experience, employee satisfaction products, and brand. Qualtrics had intended to go public this week. The company generated $342M in the last twelve months and had grown 27% annually.
map[]
I’ve written quite a bit about the public market software multiples. They’ve increased to near historic levels with forward revenue multiples approaching 9x. As the public markets have appreciated, something has happened that I didn’t expect. Some public companies are now fetching the mulitples of the most attractive private companies. I thought valuations between the markets would normalize because of a deflation in both public and private. Today, we have the beginnings of normalization by inflation.
map[]
Last week, AirBnB proposed granting their most valuable hosts shares in the company, much like employees. This requires a change in securities law. It’s a novel idea that has a history, promise and some risks.
In 1998, an online travel agency, TravelZoo, tried this. They granted free shares to 700,000 people who signed up to use the service. Administration of these shares proved a headache for the company. Many of those who received the shares were unaware of conversion deadlines.
map[]
About two years ago, Marketo was publicly traded and valued at roughly $1.1B. Vista Equity paid $1.8B to take the company private, a 64% premium. At the time it was taken private by Vista, Marketo generated $241M in trailing revenue, growing at 35% annually. Its net income margin was -31%.
Last week, Adobe announced they acquired Marketo for $4.75B. By the time of the sale, revenues had grown to $321M, growing at about 21% annually and profitable.
map[]
The SEC announced last week that it wants to find ways to let Main Street investors access stage private venture companies. This news item underscores an important trend that is reshaping the industry. Today in Startupland, startup access is the scarcest commodity. Everybody wants an allocation, an opportunity to invest in the very best companies. The SEC story highlights how much has changed in Startupland. In this post, I’ll touch on three.
map[]
Salesforce is worth $113 billion. 1% of $113 billion is $1.13 billion. ServiceNow is worth $34B and Workday is worth $33B. 3% of $33-34B is $1B. Atlassian is worth $20.5B. 5% of $20.5B is $1B. Why am I doing all this simple math you might ask?
We have reached a point in SaaS where a small fraction of an incumbent is a billion-dollar company. If you start a business tomorrow that is able to cleave 1% of revenue from Salesforce, you will have built a billion-dollar business.
map[]
Jacob’s Ladder is a toy of thin wooden blocks attached by ribbon. If you hold it in your hand and rotate it to touch the second block, it seems to set off a cascade of blocks falling from the top. The blocks haven’t changed positions, though they do rotate. It’s a moving optical illusion. When I watch this toy, I’m reminded of the current state of the fundraising market.
map[]
In the last six years, the median time between seed and Series A has more than tripled from about 200 days to about 750 days. Why? The seed market is in the midst of some secular changes. Seed rounds have declined 63% from their peak. Total dollars invested have fallen by 37%. But the median round size is up 3x in the same time period. In other words, investors are concentrating capital in fewer startups.
map[]
2018 is a blockbuster year for software M&A multiples. The prices companies fetch relative to their revenues surpass any of those in the past 7 years. Billion-dollar plus acquisitions in 2018 have commanded a median 17.7x trailing enterprise value to revenue multiple. Nothing in the past seven years is close. In fact, there is not a single acquisition in that range.
In 2018, three acquirers have paid greater than 14x trailing multiples, and two have paid greater than 20x trailing.
map[]
There are five forces driving the startup ecosystem today. They are working together to reinforce a high valuation environment. These forces are:
An infusion of capital into Startupland. There are many reasons for this. The money supply in the US has doubled in the last 10 years. A low interest rate environment means a low cost of capital, which means yield is hard to find for cash. VCs raise larger funds and more frequently.
map[]
Imagine you came across this ad.
Scientists have discovered a revolutionary new treatment that makes you live longer. It enhances your memory and makes you more creative. It makes you look more attractive. It keeps you slim and lowers food cravings. It protects you from cancer and dementia. It wards off colds and the flu. It lowers your risk of heart attacks and stroke, not to mention diabetes. You’ll even feel happier, less depressed, and less anxious.
map[]
Technology innovations swing to a pendulum’s cadence. Sometimes innovations begin with infrastructure changes and reverberate up the stack. Other times, front-end engineers innovate at the application layer, which demand downstream changes in the infrastructure to scale.
The last major epoch of front end evolution has celebrated its ten year anniversary. We’re in a period of punctuated equilibrium. When will we see rapid speciation?
Web 2.0 and mobile applications built on iPhone and Android transformed the way users interacted with technology.
map[]
The venture capital markets are flush with capital. We’re approaching the heady days of the dot com era. In that epoch, despite the record volumes of venture dollars, startups went public quickly, in 4-5 years. Today, that timeframe is no longer realistic. In fact, the surfeit of private dollars delay IPOs.
From 2000-2005, the “typical” IPO-bound startup listed on an exchange 5 years after founding. After the Lehman collapse and the crisis of 2008, the IPO closed, and startup age spiked to 16 years median.
map[]
In the US, the median seed round has nearly quadrupled over the past seven years. In the mean time, seed investment has grown more than 7x and then fallen to a bit more than half of the high. As the market has grown and retrenched during that time period, I’ve been wondering about the geographic diversity of these seed dollars. Throughout these cycles, are startups in other states benefitting? Are they increasing their share of investment dollars?
map[]
At Saastr, Jason and I discussed the role of private equity in SaaS on stage as a potential acquisition path for SaaS startups. Private equity hasn’t been a common exit route for venture backed startups in the past. But that’s changing.
The chart above depicts the total disclosed value of US venture-backed SaaS startups which have been acquired by PE firms since 2010. The aggregate value has grown from zero to about $13B over that time period.
map[]
Much of the conversation today about crypto is about Bitcoin and currency. But that’s myopic. Soon, we will be talking about how crypto will change the software world. In fact, many founders have already started that pursuit. More than 30% of the initial coin offerings (ICO) in 2017 target developers and businesspeople with their products.
The numbers are still small. B2B crypto companies raised about $400M of ICO dollars in 2017.
map[]
Below are 7 predictions about the startup software ecosystem. How many of them do you agree with?
The tax holiday for repatriation creates one of the most active M&A environments of the past ten years. The repatriation holiday is part of the new tax plan. It permit companies to bring US dollars held abroad (from software sales in other countries) back to the US at a lower tax rate than before.
map[]
I’m grateful for all the entrepreneurs who have spent time teaching me how to build their companies. Over these past few days, I have been reading the book Antifragile written by Nassim Taleb. In the book, he writes about many provocative things, but the one that sticks out with me this holiday is about innovation, and it harkens back to the original title of this blog that I started to write nearly 7 years ago: ex post facto.
map[]
Initial Coin Offerings, a fundraising mechanism for companies using cryptocurrencies as a mechanism to buy their service, seem to be upending the world of venture capital. Filecoin raised $250M through an ICO. Tezos raised $232M. Bancor raised $153M. These are massive amounts of money. Recently, I’ve been wondering how prevalent ICOs are and whether they could potentially be a substitute for venture capital.
The chart above shows the number and size of ICOs since the beginning of this year.
map[]
Last week, I participated in two discussions about the changes in the SaaS world. I believe they are fundamental. The most important force shaping the industry today is competition. The level of competition in many core SaaS segments is intense.
Why? The SaaS era is about 20 years old. Salesforce was founded in 1999. Since then, many major categories of software have been saasified. Venture capitalists have financed many of those businesses.
map[]
I commuted to my first job on a bicycle. With my parent’s help, I bought a lemon yellow second-hand road bike that I pedaled 20 miles each way from 30th and N streets in Georgetown, Washington DC, over bustling Chain Bridge and the languid Potomac to an office park buried in Tyson’s Corner in Virginia. That was my workout each week. Then I moved to California and retired the bike.
map[]
A founder asked me if we had reached the point that SaaS is commodified. “Can you build a venture scale SaaS company anymore?” He made three key points to support the argument.
First, the technology barriers to starting a SaaS company continue to fall. Amazon, Google and Microsoft provide sophisticated, scalable, and easy to use infrastructure as a service. Next-generation machine learning tools are also available by API and improving all the time.
map[]
Harry Stebbins published a podcast with David Beisel this week in which they discussed the importance of voice. David says, “Voice is the most natural user interface possible.” I think the biggest challenge for voice is the skepticism and cynicism engendered by a decade or two of poor experiences. It’s no longer the technology.
One of my partners recently switched from an iPhone to an Essential phone and was stunned by the accuracy of the voice interface Google offers.
map[]
Invest Like the Best is one of my favorite podcasts. Hosted by Patrick O’Shaugnessy, Invest Like the Best profiles investors from many different disciplines. Recently, New York seed investor Jerry Neumann spoke on the show and talked about three key ideas on the technology innovation cycle.
First, he discussed reading Carlota Perez Technological Surge Cycle. I first read Perez’ book after Fred Wilson wrote about it. Jerry describes the theory very well in about 10 minutes.
map[]
Last week, I wrote about the decline of investments in San Francisco startups. On Hacker News, this post engendered a lengthy conversation on the challenges facing founders and start-up employees in San Francisco. In short, the cost of living in San Francisco and the Bay Area has reached untenable and unacceptable heights for many.
The average one-bedroom apartment in San Francisco costs $3500 per month. That is simply astronomical. It’s 30% more expensive than Manhattan.
map[]
In software, we’ve moved from a world where a customer buys a piece of software to run on their own infrastructure, to a world where a customer pays a vendor to run software on the vendor’s infrastructure. With machine learning, we may see another evolution of this. Machine learning startups create models based on data provided by customers. Should customers be compensated for their contribution?
Unlike the first wave of SaaS software, machine learning startups benefit from the data their customers share with them.
map[]
We’ve entered an era when computers can understand speech, computers can synthesize speech, computers can develop music, author encryption algorithms, create novel art, respond to customer support questions, and even generate new summaries and reports from data. Increasingly, humans will struggle to distinguish between computer-generated and human generated. Consequently, here’s an opportunity for startups to lead, not just technologically, but more broadly.
DT is a creative digital agency business based in Australia that developed a prototype that distinguishes between human speech and synthetic speech.
map[]
In January 2010, Andrew Parker wrote a post called the Spawn of Craigslist. Andrew identified companies that had built businesses by unbundling Craigslist. The vacation rentals link gave rise to AirBnB and HomeAway. Etsy dominated the arts and crafts for sale. This same unbundling is occurring to Excel.
Microsoft Office has more than 1 billion users globally. Assuming a 33% penetration of Excel, that’s a user population 300M users. Like Craigslist in the consumer world, Excel became the tool for nearly everyone to get stuff done at work.
map[]
As the number of SaaS applications has exploded, the SaaS ecosystem is responding to data fragmentation with middleware. This isn’t the middleware of the early 2000s, which was focused on helping developers build software. This is middleware that is focused on helping end-users unify data from the vast numbers of data repositories now existent.
Ten years ago, sales teams captured basic data on customers and prospects. Names, addresses, budget, time frame.
map[]
My algorithm is better than yours. My algorithm performs better on the precision/recall tradeoffs. It surfaces fewer false positives. It converges to an answer faster. Perhaps it requires a bit less data. Those statements might all be true. But none of these advantages confer a competitive sales advantage in the market. They aren’t technology innovations leading to a go-to-market advantage.
I first observed the use of large scale machine learning at Google.
map[]
At SaaStr earlier this year, I spoke about the huge potential of machine learning in SaaS. In that talk, I broke down some of the advances in ML that might be useful for software companies. In the discussion that ensued, I stressed the importance of not letting the technology obfuscate the value proposition of the software. Yes, ML is a huge step forward, but it’s not enough by itself. In fact, it likely isn’t the most challenging part of building a disruptive product.
map[]
From the millions of Amazon Alexas to the self-driving car, new products are coming to market infused with machine learning. The innovation offered by machine learning techniques are real, and they will changed the SaaS world. But how? How can startups use machine learning to their advantage?
There are four broad applications of machine learning:
Optimize - this morning, fastest way to travel from Sand Hill Road to South Park in San Francisco is highway 101.
map[]
The end of the year is fast approaching. Time for some quantitative analysis of the content that readers liked the best this year.
2016 was a year of change for SaaS, and most of the story was the public market. Valuations there fell from their highs in 2014. More than $70B of public SaaS market cap was taken private by both other publics and also by private equity firms. SaaS company formation seems to be slowing, but the companies that do raise, command larger Series As than ever.
map[]
For hundreds of startups in the as-a-service world, the scores of product launches at last week’s Amazon Web Services Reinvent Conference each were a warning shot across their bows. We are coming. We are coming right after you, with tens of billions of dollars on our balance sheet, hundreds of salespeople, and the broadest suite of software and infrastructure since Oracle. Anything that’s open source with traction, we will host.
map[]
A few weeks ago, I had my first customer support experience of the future. I was in a meeting when my Android’s caller ID told me American Express was calling. I stepped of the conference room and answered the call. A machine-generated woman’s voice identified itself as the American Express fraud department. “Do you have a bluetooth headset or headphones you can use with your phone?” she asked.
I replied that yes, I did.
map[]
The number of vendors selling to sales and marketing has exploded from 500 to more than 3000 over the last three years. Are we reaching the end of an expansionary cycle? The software pendulum tends to swing between software suites, offering a collection of different tools, and best-of-breed point solutions.
But, have we reached the point where the best-of-breed, fragmented ecosystem is a permanent fixture? Okta’s Businesses at Work 2016 report calculates most enterprises pay for somewhere between 10-15 corporate applications.
map[]
Why is this ad appearing on this site? This was a frequent questions both advertisers and publishers asked of Google. Behind the scenes, machine learning models match the best ad to the best website, given a set of constraints including budget and the dynamics of the ad auction. To untangle the decision chain across the many different targeting systems to answer the question is a knotty task indeed.
In March, the head of Google research Peter Norvig spoke at MIT.
map[]
It’s no secret I believe speech is the next input mechanism. We are in the Voice-to-Text Era. I wrote in late 2014 that speech is the fastest user interface, and the newest speech recognition experiments confirm it.
Andrew Ng, a luminary in the world of machine learning, and his teammates at Baidu, Stanford and University of Washington have developed Deep Speech 2, a neural network based speech recognition system. They tested the speech and accuracy of the system and compared it to people typing on their mobile phones.
map[]
In addition to increasing labor costs, startups in San Francisco are facing monotonically increasing real estate prices. JLL the real estate broker shared their data on the average asking rent in San Francisco from 2007 two 2016, year to date. In 2009, the average asking rent was $31.37. In 2016 that number has more than doubled to $73.05, for an average annual increase of just about 13%.
The most expensive neighborhood in San Francisco is Mission Bay/China basin at approximately $85 per square foot per year for class A real estate, followed by South market for an average of $80 per square foot.
map[]
The next big shift in SaaS is an evolution from software as a service as a displacer to a disruptor. Displacement technologies compete with incumbents on the same buying parameters. Disruptive companies change the way a buyer thinks about solving their need. Most SaaS products today are displacers.
SaaS products initially were viewed as a cheaper, often inferior product to their client/server peers. Five or ten years ago, that may have been true.
map[]
Which are the ripest areas for startups to disrupt using machine learning? At the core, machine learning/artificial intelligence relies on two key ingredients: advanced algorithms and data sets to train those algorithms. Novel algorithms are increasingly making their way into the public domain in the form of open-source libraries. So, the key differentiator for startups and ultimately long-term competitive advantage is access to proprietary data sets.
In the consumer world, there are natural and intrinsic monopolies at play.
map[]
How does a chat bot company transform a technology advantage into a distribution advantage? This is the key question facing startups in this part of the ecosystem.
Deep learning is advancing at an incredible rate. Google has released TensorFlow and Parsey McParseface. These advances will improve natural language understanding, enabling chat interfaces to amaze and delight users.
So let’s assume that the technology works. Let’s take for granted that despite the myriad ways a human can ask a question, the computer will understand the query.
map[]
Last week, Facebook launched the messenger chatbot platform, which like Slack, Telegram and others, presents a big market opportunity for startups to innovate. In addition, the companies and products that determine how to most efficiently distribute their product on these new platforms will benefit from user curiosity and less competition, both of which result in lower cost-of-customer acquisition.
These are the five questions I’m wondering about in this new world:
map[]
Quick. Casual. Human. Chat differs from other forms of communication. Because of these three attributes, chat seems to be reemerging as a potentially disruptive user interface for both consumers and business users.
The typical teenages emits more than 3000 text messages per month, not to mention messages on other networks. Most retailers provide customer support via chat. Hundreds of millions of people have tweeted. The combination of these three forces expose nearly everyone on the Internet to brief, useful messages every hour.
map[]
Which of the 16 major start of categories in information technology will reap disproportionate share of investment dollars in 2016? And which sectors are closely guarded secrets shielded by seed investors, that may have their breakout year this year?
Using Crunchbase data, I analyzed share of dollars commanded by each of these 16 categories over the last five years to understand the trends in both the seed and series A fundraising markets.
map[]
Just 25 months ago, Aileen Lee coined the term unicorn in her post Welcome to the Unicorn Club and last week the Economist declared these Unicorns “Gored.” Over the span of those two years, the Unicorn press cycle has swung from euphoric apotheosis to bleak nadir.
The rise and fall of unicorns in the press isn’t a unique phenomenon.
A company’s narrative moves like a clock: it starts at midnight, ticking off the hours.
map[]
Last week, a friend asked me what I thought the future of software would be. At some point, all the workflow processes that still conducted on paper and pencil will be digitized, all of the tasks completed in Excel will be optimized, and the majority of limitations of traditional software will be overcome. It’s not to say that software innovation in its current form will continue to exist for decades, but what is the fundamental shift that resets the ecosystem like the cloud did in the early 2000s?
map[]
Over the weekend, it seemed to me that the sentiment in the valley seemed to change abruptly. Dan Primack wrote Fear and Sadness in Silicon Valley, and Danielle Morrill punctuated her post Somewhere Over the Brainbow: The Unicorn Window is Closing images of unicorn cannibalism. While the attitudes may be changing, the effects haven’t yet revealed themselves in the data. So, is the environment truly changing?
The number of megarounds, investments of greater than $40M in US companies, continues to surpass last year’s figures.
map[]
I’ve been to many YC Demo Days and I always look forward to them. This year was no exception. There are so many wonderful ideas and companies founded by terrific entrepreneurs. In addition to the pitches themselves, the types of companies presenting forbear trends in the startup world more broadly.
To get a better sense of those trends, I’ve categorized more than 250 startups in 3 recent classes and plotted the evolution of the classes.
map[]
We all build products based on assumptions - assumptions about our users, who they are, how they think, what they expect. When the underlying assumptions underpinning product design no longer holds, new opportunities are created.
Jakob Nielsen, a [famed user experience researcher](https://en.wikipedia.org/wiki/Jakob_Nielsen_(usability_consultant)), writes about one such secular dislocation of UX assumptions in the Anti-Mac interface. It’s a post written in 1995 that remains remarkably relevant today. It contemplates all the assumptions underpinning the original Apple software design, and how the old fundamentals no longer apply for users who have grown up with technology.
map[]
Traditional software was initially sold by perpetual license. Then in the mid-00s with the advent of SaaS, the market shifted to per seat per year pricing. And simultaneously, freemium marketing strategies blossomed. Freemium companies provide software for free temporarily to entice users to try and use the product. Eventually, these users cross a threshold and convert to a paid subscriber. This threshold can be based on number of people using the product (Expensify), number of documents signed in a month (HelloSign), or additional product features needed by users (Yammer).
map[]
The median equity stake of US venture-backed post-Series A CEO has increased from 15% to 21%, a 40% increase in five years. This trend is also manifested in Series Bs, but as the chart above shows, post-Series C and D, total founder/CEO equity positions have remained constant.
Meanwhile the equity stakes of founding VP of Engineering and VP of Product have remained relatively constant throughout the same five year period across all stages of company.
map[]
The real promise of the Internet of Things isn’t simply linking millions of devices together, just like the real innovation of the web wasn’t networking a bunch of computers. Instead, the true and still unrealized potential of IoT is to transform business models; it’s enabling companies to sell products in entirely new and better ways that benefit both the company and the customer.
Around the turn of the millennium, startups began using the web browser to deliver their products to consumers and companies.
map[]
What will the world look like when cloud compute and storage are free?
Cloud computing prices are hurtling to zero. The chart above shows the logarithmic decline of the cost of a transistor cycle by 11 orders of magnitude (11 decimal places) over the past 40 years. AWS has decreased prices for EC2, elastic compute cloud, and S3, simple storage service, 42 times in eight years.
Storage is following a similar path.
map[]
More than 100B mobile applications have been downloaded since the launch of the Apple iOS and Google Play stores. As the number of users, downloads and apps have exploded, the dynamics of the app store have also changed. During the past 18 months, the competitive behavior within the Free Apps section of these app stores has evolved substantially in four meaningful ways:
First, the Android Play store has become substantially more welcoming to startups/new entrants than 18 months ago.
map[]
When starting a company, founders often decide between two structures: the S-Corp and the LLC. But, as James Surowiecki writes in “Companies with Benefits,” there is a new option called the B-Corporation, the Benefit Corporation.
The Benefit Corporation became an option in 2010 when Maryland’s legislature signed it into law. Subsequently, nearly twenty states have followed suit including California and Delaware. In the past four years, more than 1,000 companies have become B-Corporations including some very well known brands in technology and consumer goods: Warby Parker, Patagonia, Seventh Generation, Method Soaps, Etsy, New Belgium Brewery, Ben & Jerry’s, among others.
map[]
For consumer products, the power of mobile distribution is hard to overstate. Facebook serves more than 1B mobile users each month. Angry Birds reached 50M users in 35 days, a feat that took Instagram 18 months and Facebook more than 3 years.
But the distribution advantages of mobile app stores hasn’t been observed as powerfully in SaaS or enterprise software for driving revenue. The majority of the top 20 slots of the Business Apps Category in the iTunes store aren’t occupied by startups.
map[]
Through last week, 80% of fast-growing technology companies who reported Q2 earnings had met or exceeded expectations, on average by 4%. As the earnings season has progressed, the tech sector is showing impressive signs of strength and predictability in the underlying companies’ business. Startups benefit from a booming public market for three reasons.
First, strong earnings buoy public technology share prices which drives M&A. The chart above shows the relationship between beating earnings and Enterprise Value to Trailing-Twelve Month Revenue Multiple (EV/TTM Rev).
map[]
I’ve never paid much attention to earnings reports in the past. But as I’ve been analyzing the tech industry more and more, and the stock market has begun to surpass record highs, I’ve been wondering whether the persistent bull market in tech is healthy. Or at the very least, whether the companies pushing the industry forward are able to understand the environment and their businesses well enough to meet the commitments they make to public market investors.
map[]
Through the first six months of 2014, VCs have raised about as much as all of 2013. If this pace of fund raising continues, 2014 would mark the biggest year for VCs since 2001, when the industry raised about $38B. This new money hasn’t yet hit the startup fundraising market in earnest, as the chart above shows. The second quarter of 2014 is the sixteenth largest by capital deployed sinced 1995, making it a top quartile quarter, but to break into the top five, that figure would need to triple.
map[]
In 2008, when I started working at Redpoint I knew very little about how the venture business worked, and before I started at the firm, I wanted to prepare by learning as much as I could about the industry. Unfortunately, not much was written about venture capital at the time.
In fact, I found only two books: a textbook on private equity and venture capital by HBS professor Joshua Lerner, and an out-of-print collection of 32 VC interviews called “Done Deals,” published in September 2000.
map[]
Vik Singh wrote a great post in VentureBeat last week titled “Why Salesforce Needed to Buy RelateIQ” in which he talks about a new era in SaaS, the Predictive Era, the era of intelligent software. We’ve just seen one of the first acquisitions in the category with RelateIQ*, but I believe we will see many, many more for a few reasons.
First and most importantly, prediction provides competitive differentiation in an increasingly competitive market.
map[]
It would seem hardware startups are booming. First, the amazing success of the GoPro business and IPO, which set a 23-year high-water mark for a consumer hardware company. Second, there seems to be a growing number of hardware startups bubbling in incubators like Lemnos Labs and Highway1. Third, Kickstarter and other crowdfunding sites have enabled hardware startups to mitigate one of the biggest risks in starting out: obtaining a reliable proxy for consumer demand.
map[]
Benjamin Morris, a writer for arguably the best computational journalism publication, fivethirtyeight, published “Lionel Messi is Impossible” which describes in words, statistics and charts why Lionel Messi is one of the greatest players in the world. Even if you’re not a soccer/football fan, the article is worth reading because it’s one of the finest examples of synthesizing data and a story to convey a point I’ve read in a very long time.
map[]
Google revealed many novel projects and products at yesterday’s Google IO Conference. At the moment, I’m most curious about the development of Android Wear, in part because of the beautiful Motorola 360 watch and in part because I suspect connected watches bring substantial change to the mobile device market.
It’s easy to write off watches as a niche product. But the addressable market of Android users is now large enough that even a single digit market penetration means tens of millions of users and a potentially interesting new platform for startups to build applications upon.
map[]
Jill LePore’s New Yorker polemic “The Disruption Machine” attempts to debunk the incredibly popular Innovator’s Dilemma, a theory written by HBS professor Clayton Christensen. I’ve been reading the debate around it with some interest. It’s becoming a really interesting conversation but I think the debate is focused on the wrong thing - whether or not these ideas are absolutely correct, even axiomatic. They aren’t always true. But that doesn’t mean these concepts are useless.
map[]
Since it was first written in 1982, the Simple Mail Transfer Protocol, the mechanism for sending emails, has remained largely remained unchanged. Today SMTP delivers 70 trillion emails to 5B inboxes each year. Overwhelmed by tens of thousands of emails, most of us can sympathize with Nick Bilton, who said:
There is no escape: Email is probably most invasive form of communication yet devised.
Unlike SMTP, email is changing very quickly.
map[]
BYOBI is an acronym I first heard on a telephone call with a VP of Technology at a large corporation. The word is almost unknown today, but I think that it will be one of the largest trends to impact data in the next five years.
BYOBI means Bring Your Own Business Intelligence. This VP of Technology was struggling to enable the sales people, marketers, engineers and others within his business to access the data they needed.
map[]
At an Internet of Things conference last week, I took part in a panel in which we discussed the future of connected devices. Will simple products win or will complex products dominate in the IoT?, we were asked. I think the question misses the point and raises another problem about the Internet of Things more broadly. It’s not about Things. It’s about Services. Software-as-a-Service and Platform-as-a-Service.
Simplicity and complexity are two sides of the same coin.
map[]
One of the most important trends in the Internet at the moment is unbundling. Entrepreneurs are picking apart Craigslist and eBay, vertical by vertical. At the same time, other entrepreneurs have replicated the core functions and features of Facebook and LinkedIn, creating hugely valuable companies.
But simply calling this trend unbundling doesn’t do the movement justice, particularly in the transactional web. The trend is more fundamental.
Craigslist and eBay are the canonical internet market places.
map[]
Since the first transistor, increasing speed has been at the core of much innovation in Silicon Valley . Over more than three decades, Moore’s Law has remained the engine of progress in chip technology. I’ve been wondering if a analogous productivity law will be written for data.
One level up the stack from the chip, software has benefitted tremendously from chip speed. I learned to code in Java, which I needed to compile before I could test my program.
map[]
From zero to $19B of business value in five years; WhatsApp’s sale to Facebook is an important moment in the history of the consumer web. The deal proves distribution, reach and large user bases aren’t the competitive moats they once were. Apple’s App Store and Google Play have leveled the playing field to such an extent that a startup can command 10% of the market cap from a $200B company.
map[]
Given all the momentum of the NoSQL movement, it would be easy to write off SQL-based technologies as forgotten, or simply standing still. But there’s a tremendous amount of innovation occurring in SQL databases. Amazon’s Redshift, an elastic data-warehousing solution launched in late 2012 is the most salient example.
Redshift’s ability to process huge volumes of data is breathtaking. When running Redshift on solid state drives (SSDs), one team at FlyData queried 1 terabyte of data in less than 10 seconds.
map[]
As recently as six months ago, it was easy to disregard the Internet of Things (IoT) as just a theoretical market that Cisco measured in the trillions, but whose potential never seemed to materialize. That’s all changing.
The past year ushered in a new era for the Internet of Things for three reasons. First, venture capitalists invested nearly $1B of capital in the IoT in 2013, more than 3% of all VC investments by dollars.
map[]
Yesterday, I walked the sprawling floor of the Consumer Electronics Show. These are the six key themes/trends I noted.
The Coming of Era of Drones - I saw barrel-rolling quad-copters and parkouring wheeled robots dancing in synchrony to hiphop music. It was awe-inspiring and terrifying at the same time. If you watch the video, I think you’ll agree the mechanisms to control drones are rapidly becoming more sophisticated. Setting aside regulatory questions, there seem to be fewer and fewer barriers to large scale drone deployments in the very near future.
map[]
Bitcoin has captivated the imaginations of many with its quasi-anonymous, hyper cost-efficient payment network. The potential for Bitcoin to change foreign exchange is hard to overstate. In the same vein, the technologies that enable the internet of things (IoT) like Bluetooth Low Energy and Apple’s Beacons and Electric Imp’s infrastructure will transform the way we interact with the physical world to something akin to the mall in Minority Report.
The startups that bring Bitcoin and IoT to the mass-market won’t declare themselves Bitcoin companies or connected devices companies.
map[]
For retailers of any size from startup to Fortune 50, the tablet application has become more important than the mobile application. IBM’s annual Black Friday ecommerce report, which tracks 800 internet retailers proves the point.
Tablet users buy at a conversion rate which is 300% greater than mobile phones and despite generating only 60% of smart phone traffic, drive 65% more transactions at 14% larger basket sizes. Ultimately, the average tablet user is worth 3.
map[]
The typical mobile phone’s home screen is occupied by more than 30 applications. A digital tragedy of the commons, each additional mobile application a user downloads decreases the odds of an average application re-engaging a user. After all, the time spent on mobile isn’t increasing fast enough to cover the marginal application.
Developers have used push notifications effectively to re-engage users, but push overload drives more and more users to opt out of push.
map[]
When I first started programming at a startup, I were proud even snobbish about the language we used: Java. I derided JavaScript. I couldn’t have been more wrong.
JavaScript has become the dominant language of the web. Every browser, PC and mobile, understands Javascript. It is a lingua franca among computers. Some hardware even speaks JavaScript.
Because of its universality, JavaScript’s growth has compounded. The chart above shows the number of code repositories by language on GitHub, as analyzed by RedMonk.
map[]
Twitter is the first form of communication that changes depending on who is using it. Like the telephone, the fax, email, Twitter enables all types of communications: friendly banter, customer support requests, news syndication and public service announcements to name a few.
But Twitter imposes a three new wrinkles that set it apart as a new form of communication. First, the recipients of a tweet are knowable but never really known.
map[]
I’ve always wanted to read a newspaper like The Daily Prophet from Harry Potter, a daily with animated images and with articles that respond to the reader.
Last week, the Guardian, the world’s third-most widely read paper, published NSA Files: Decoded which explores the many different points of view on the NSA’s information collection efforts. As you scroll down, senators and attorneys and experts begin briefing you. Interactive data charts invite you to explore trends and embedded documents beg you to dig deeper.
map[]
In 2008, iGoogle represented 20% of traffic to Google. Seven years later, the mobile phone is the home screen of choice for a billion people. iGoogle is dead. Mobile killed it.
The typical mobile phone user checks their phone 110 times per day. I suspect iGoogle’s most avid users visited about 10 times daily. By that math, mobile home screens generate an order of magnitude more engagement. If Google or any company wants to provide the home screen to those billion people, that home screen must be a mobile app.
map[]
Quietly mentioned in yesterday’s press conference about Google’s Android update is a new feature that will change the way people use their mobile phones, search deep-linking. With KitKat, Google is applying its world-class crawling and search technology to the content and data within mobile applications.
Quoting from the Verge, >A search for a restaurant will offer a link directly to that restaurant page in the OpenTable app if you have it installed, allowing you to set up a reservation.
map[]
This week, Netflix announced its US userbase grew to 31M subscribers surpassing HBO for the first time. The magnitude of Netflix’s milestone is hard to overstate. In a bid to compete with Netflix, HBO has partnered with Comcast, which serves 21M subscribers, to trial an Internet-only subscription plus HBO, the first time HBO is available to US consumers without a full cable subscription. It’s a clash of behemoths.
Separately, the NY Times revealed they have amassed more than 700,000 digital subscribers who provide upwards of 20% of circulation revenues and grow about 40% year over year.
map[]
Intent to purchase is the engine of the consumer web. Creating and capturing intent motivates almost every dollar invested into ecommerce and advertising. Intent is also the fuel for the Internet’s most successful business model, Google’s AdWords + Search.
As the internet has evolved, so have the ways of creating and capturing intent. From display to search to retargeting to collections, each new technique has leveraged data in a novel form to discover consumers’ wishes.
map[]
Paying for an Uber is a breeze. I step out of the car and go on my way. Behind the scenes, Uber has identified me, recorded my fare, added a gratuity, and sent me a digital receipt that I can easily forward to Expensify. That’s the way I want all my payments to be. Paying for Uber with a mobile phone is less work than using a credit card, which is the biggest competitor for share of wallet of in-person payments.
map[]
Over the course of the past few months, I’ve built six projects that use APIs from Twitter, Google, LinkedIn, Mailchimp and others. These APIs enable me to extract data that I use to analyze this blog and build tools to be more effective. Once I extract the data from their platforms, no API platform understands why or how I use the data. While web companies measure many things on websites and mobile apps, very few apply the same rigor to APIs.
map[]
During the past week, I’ve been tapping out letters on a 1921 Underwood portable typewriter that my wife gave me as a present. Sitting in front of it and watching the letter hammers pound ink onto paper reminded me that computer programming is still a very new field. Not 50 years ago, prehistoric programmers punched FORTRAN code on punch cards in this way. In the past 10 to 15 years has computer programming entered mainstream culture and the number of developers has soared, popularized by the dot com era and the increasing prevalence of computers and mobile phones.
map[]
The NoSQL movement launched officially in 1999 but rose to prominence much later perhaps closer to 2008 when Hadoop and other key value pair technologies became en vogue.
Today, it’s hard to argue with the success of the movement. Large banks, insurance companies, biotech companies and dotcoms rely on NoSQL to power their services and inform their most important decisions.
I first saw MapReduce at Google. I’ll never forget the conversation I shared with a search engineer who described MapReduce.
map[]
The venture capital fund of the future will perform the same tasks as the venture funds of today: help portfolio companies, evaluate new investment opportunities and build networks of other investors, potential hires, and founders.
But to succeed over the next 25 years, venture capital firms must increase the scale and sophistication of each of these duties by order of magnitude through technology. In the five years I’ve worked in the venture business, I have witnessed tremendous change in the industry driven by intensifying competition for returns.
map[]
Mobile apps are like packaged software, a friend who is a head of product at a successful mobile first company told me over breakfast. On the web, you can launch a product that’s 80% functional to hit a promised launch date. Around 1 am that night, when most users are asleep, you can surreptitiously push an update with bug fixes and new features, reboot the servers, and no one is the wiser.
map[]
There’s a force in mobile app distribution that isn’t talked about much despite its magnitude: mobile social networks. Mobile social nets are becoming the predominant mobile app paid discovery/distribution platforms.
How big is this force? Line, the Japanese mobile messenger with more than 200M users is on a $400M run rate, 60% of which is derived from app distribution. Kakao Talk will pay out about $600M to developers this year - which is about 10% of iTunes global developer payout in 2012.
map[]
There’s a concept in computer programming called operator overloading that neatly describes the current Series A investment market. Overloading means using the same code for two very different purposes depending on inputs. For example, a programmer could redefine the + function so that instead of adding two numbers, the function would multiply the numbers if one of the numbers was negative. Overloading can create lots of confusion in programming because the same function name can return very different results.
map[]
Product and engineering teams are inseparable and at the core of most startups. One degree removed, sales and customer support teams ply the voice of the customer to influence product and eng. Of the five teams in a startup, marketing teams tend to have the least influence because traditional product marketers must influence others to enact change.
Marketing works with product and engineering to update branding and communication within the product.
map[]
Control4’s IPO last week and Nest’s massive success mark the beginning of mainstream, useful connected devices. In the decade to come, the most successful will be united by one common characteristic. They will anticipate our needs and make us lazier. Living is hard work. So many things to remember, like buying milk on the way home tonight. Plenty of software has been written to help us manage our daily torrent of to dos.
map[]
Crowdfunding platforms solve three key problems for startups, two of which are obvious. Kickstarter and Indiegogo, among others validate demand and provide short term financing by marketing product ideas and accepting pre-payment for future delivery. Consumers vote with their dollars to catalyze product development.
The third, less obvious benefit, is the platforms cajole startups to follow a product management process similar to Amazon’s process, described below by Ian McAllister in his Quora post What Is Amazon’s Product Management Process?
map[]
In 2005, I started as a customer support rep in the AdSense team at Google. Over the next few months, I was trained in the art of handling hundreds of emails per day.
My training focused on using a customer service tool called Trakken that Google had acquired and heavily modified to enable AdWords and AdSense teams to manage the torrent of inbound support emails from customers. In AdSense, a team of fewer than 100 people supported a few hundred thousand customers almost exclusively through email with customer satisfaction scores consistently above 90%.
map[]
More and more companies realize their proprietary data contains insights that drive tremendous competitive advantage. Enabling an organization to make data driven decisions is a long term process. Below is the current big data adoption process and where we are within it:
Companies generate proprietary data whose volumes can’t be handled by existing tools. The main adopters of these technologies are financial services, healthcare, genomics and web companies. Companies build or buying the tools and expertise to store and process that data.
map[]
We’re already seeing the impact to hardware vendors of peak smart phone. Over the past six months, Apple, Samsung and HTC share prices have fallen 20 to 25%. These share prices have tumbled because smart phone penetration is hurtling towards 70% in the US. True saturation has occurred in certain segments: 87% of 30 to 49 year olds with income greater than $75k own a smart phone. Does the app economy suffer from the same challenge of decreasing growth as hardware?
map[]
I’m asked with some frequency which startup sectors are booming. Mobile messaging and big data are knee-jerk reactions at this point. But these days I often respond “financial services.”
In the last two years, financial services startups have been innovating impressively quickly and challenging some of the fundamental ways in which capital and credit are distributed. I count seven major categories of innovation to date:
Math Based Currencies/Payment Networks - Bitcoin might be the most well-known and best publicized math-based currency, but there are a handful of others in the market, like OpenCoin and LiteCoin.
map[]
Recently, hardware companies have been popping up in all kinds of places. Nest is building the next thermostat. Sonos sells seamless and beautiful soun systems. Thalmic Labs and Leap Motion are innovating in alternate forms of computing control. Electric Imp is building the platform-as-a-service to connect devices to the web via Wifi and so on. Hardware investments are blossoming because software is reinvigorating the market. Hardware companies face deeper logistical challenges, greater likelihood of margin erosion, and more complex go-to-market strategies than their software cousins.
map[]
We all type quite a bit. I’ve never measured how many words each day I type but I imagine it’s probably a few thousand each week between my laptop and my mobile phone and across emails and blog posts. And no one can deny the toll this takes on our wrists. In the past week, I’ve been suffering from some carpal tunnel pain particularly as I’ve been coding more. Coding often requires all kinds of odd key combinations because of the varying syntaxes of CSS, Ruby, R, MySQL and the English language.
map[]
Reading through the tech press since the Facebook IPO, you might get the impression venture capitalists are still reeling from that apocalyptic offering, believe no further successes can be had in the consumer web, and so are fleeing the consumer web in droves to pursue enterprise investments.
That’s because in the past year or so most major tech publications have swung from focusing on consumer products to enterprise companies. GigaOm made this transition first, now TechCrunch and PandoDaily are following suit.
map[]
On the day of Tableau’s IPO, a company known for innovating in data visualization, I thought I would share the most impressive HCI concept I’ve seen in a long time.
In my view, Bret Victor is on the forefront of human computer interaction design. In the first two or three minutes of this video at Stanford, he demonstrates his home-built software that combines data analysis with visualization. It’s magnificent and really hard to describe because it’s so novel.
map[]
Craigslist’s generic market place gave rise to fifty or more single-purpose alternatives some of which are quite valuable companies: AirBnB, Care.com, 99Designs, oDesk, CustomMade, Gazelle.
Each of these startups operates verticalized market places that solve customer problems better than Craigslist’s generic platform. In so doing, these startups have grown their observed markets.
The same unbundling trend will happen to social networks.
Generic social networks will continue to succeed. But their most attractive users in the best segments will be systematically poached by verticalized competition.
map[]
When the neologism was popularized in 2004 by Tim O'Reilly, the words Web 2.0 captured a desire for the web to become interactive. It was a description of a movement towards social media and engaging users on the web. But more than an idea, it carried a design aesthetic which focused on the user, user experience and engagement. After all, users wouldn’t participate on a hostile site.
We’re no longer in the Web 2.
map[]
I believe machine learning will drive the next big wave of innovation in consumer web services. The very same technologies that power Google’s search and Netflix video recommendation engine will become far more common and useful, perhaps even predominant in the consumer web.
Every great consumer product has a little bit of magic. Apple employs static software and hardware design to anticipate user needs - to create that magic. Compared to user experience design and forethought, intelligence, when it works, is far more compelling and impressive.
map[]
Last night I spoke at the Enterprise Tech VC Panel. We discussed five trends in the seed market and the outlook for 2013. These are the five most important trends for 2013, in my view. MicroVC Funds Have Doubled Their Assets Call it micro-VC or mega-seed fund, there’s a new investor class which raises funds between $50 and $100M to invest in seed-stage companies. Felicis manages a $70M fund, Jeff Clavier at Softech invests from a $55M fund and Steve Anderson of Baseline has raised at two funds totaling $100M.
map[]
For sales people, social proof is one of the most powerful forms of influence, as Robert Ciadini proved in his seminal book on the topic. It’s no secret that the best leads are referrals. Second best is the friend who is a customer in common: “Oh, Peter chose Salesforce for his CRM. Maybe I should consider them too.” Social proof confers the trust of a relationship to the salesperson improving close rates.
map[]
On a Saturday morning in August of 2006, Sergey Brin and a team of Googlers flew to Los Angeles to meet Tom Anderson and his MySpace team. By the next afternoon, the two founders shook hands on a three year $900M contract. About twelve months later, I found myself as the product manager for the team in Marrakesh, Google’s executive board room, reporting the state of affairs to Eric, Larry and Sergey.
map[]
With a litany of articles in recent months highlighting the number of companies with valuations greater than $1B, I’ve started to wonder about the valuation trends for the highest profile venture backed companies. Venture capitalists are increasing market prices in Series A and Series B rounds aggressively in effort to reap disproportionate returns. And the variances in the prices of different startups in these early rounds is enormous, indicating a relatively inelastic market.
map[]
Brokers serve two key roles within ecosystems. First, they introduce buyers and sellers. Second, they lend their expertise to help buyers and sellers make the right decisions in the market. The internet neutralizes the first value proposition of brokers by leveling the information asymmetry between buyers and sellers at a far greater scale with much better data than any broker ever could. And for products and services with relatively small price points, commodity goods or where the cost of failure is low, market places often nullify the value (and cost) of an advisor.
map[]
Hardware competition is cut-throat. Walking through the halls of CES, every Android phone is identical. LG has copied each Samsung model. Because of the competition, tablet prices are plummeting - I saw a $49 7inch tablet (and some were giving Nexus 7s away for free with subscription. Not constrained to large competitors, startups like FitBit and Jawbone launched competing (and nearly identical) products within weeks of each other: the FitBit Flex and the Jawbone Up Even newer segments like gesture control of TVs or GoPro self-recording cameras are addressed by six competitors, many of them white-label.
map[]
Call it what you like. The Series A Crunch or Silicon Valley’s Financial Cliff, there’s a lot of talk about the challenge seed stage companies facing insurmountable odds raising Series A investment - PandoDaily’s analysis pegs the odds at 20% based on anecdotal data.
The three horsemen of the seedpocalypse In the past 3 years, the three major trends influencing the seed market are:
The decreasing cost of starting a company is balanced by growing labor costs.
map[]
Fred Wilson’s perspectives on trends in consumer web investment created a big brouhaha over the weekend. Commenting on a WSJ article, Wilson offered his confirmatory observations that follow-on investments in the consumer web have become more challenging as momentum investors have shifted toward enterprise. Over the past 18 months, valuations of later stage consumer internet companies have ballooned into the hundreds of millions propelled by enormous user growth. For many of these startups, revenue hasn’t been able to keep pace with rising serving costs.
map[]
When I walk into a bank today, I might deposit my cash in exchange for some interest rate. The interest rate is my cut of the profits the bank made on my deposit through their trading and lending activities. The more assets under management, the more money a bank can make on its own account. Like banks, web companies are in the business of maximizing gigabytes of user data under management.
map[]
The throwaway line in pitches these days is “we’ll sell our data.” Most of the time, this notion is wrong. Data is the most valuable outcome of building a successful product. It’s the insight, the secret, the keys to the kingdom. Don’t sell the keys to the kingdom.
Data provides economies of scale and insights used to develop huge barriers to entry and it should be kept within an organization.
map[]
The UK is just starting to enforce EU’s cookie consent law. Visit the FT for the first time in a while and you’ll see: In effect, this law requires consumers to define a Personal API. Consumers carry a trove of data: browsing data, ad click data, friend network data and social sharing data. Web businesses have realized the value of that data ($12B in 2011). This EU law encourages consumers to become aware of the value of this data and manage it.