Contact Us

Use the form on the right to contact us.

You can edit the text in this area, and change where the contact form on the right submits to, by entering edit mode using the modes on the bottom right. 

130 E 59th St., Floor 17
New York, NY

Blog

Becoming more human through mass automation

Steven Schlafman

Earlier this week. Amazon announced their latest innovation, Go. Think of Go as a futuristic grocery store. Using sensors, artificial intelligence and computer vision, Amazon is reinventing the shopping experience that we’ve all grown accustomed to for the last seventy years. That’s right. No more check out lines, registers or cashiers. If you want to buy an item, just grab it from the shelf, and then Amazon will automatically add the item to your virtual shopping cart. When you walk out of the store, Amazon will magically charge you for that item. Amazing, right? Yup. It’s also potentially scary when you think of the implications that this, and other forms automation, could have on our society.

Many industries are facing unprecedented changes largely driven by increasing wages and advancements robotics / artificial intelligence. This trend isn’t just limited to retail in the Amazon example but also transportation, food service, manufacturing, and administrative to throw out some examples. The number of jobs on the line is potentially massive. There are 3.4M cashiers nationwide according to the Bureau of Labor Statistics (BLS). There are 3.5M professional truck drivers in the U.S. according to the American Trucking Association. There are 4.7M food service workers in the U.S. (BLS). These are just a few examples. I don’t even need dig up all the numbers to conclude tens of millions of American jobs are at risk due to rising labor costs and automation.

All that said, I’m not here to paint a doomsday picture like many before me have. Hundreds if not thousands of articles have been written about our robot overloads and how we’ll eventually become slaves to them. I’m also not here to look at what we stand to lose. Instead, I’m here to look at what we all stand to gain in a world of mass automation. I believe if managed properly this massive shift could unlock enormous long term opportunities for our society and increase our overall quality of life. While there’s no doubt some pain will be felt in the short to mid-term, humanity has faced several major technological upheavals over the last thousand years and we’ve walked away every time with higher productivity, more time to focus on new activities and a higher quality of life. The mass automation era will be no different.

But first, how do we get there? Implicit within the concept of mass automation is the reality of significant structural unemployment. People will lose jobs, and those people will need to find new ways to support themselves and to support their families. This means several things, not all of which are bad. First, there’s a huge opportunity that exists around education and retraining. Retraining programs — if executed effectively — will yield not only a growth in talent available for existing American industries, but also an enormous increase in human capacity to tackle new or unsolved problems. As mass automation sets people free from menial work, socially, economically, technologically, and globally meaningful issues will become practically relevant in way they’ve never been before.

Of course, government and private retraining programs will hardly be enough to convert the millions of displaced laborers into newly productive workers in emerging industries, but they are a good start. Business, governments, and non-profits alike are already thinking about how to solve this issue. They’ll continue to do so. And I expect they’ll be successful. But for now, let’s move on. Assuming a large portion of the population no longer needs — or is able to — work in “traditional” industries, what will they do?

That brings us to the most interesting ramification of mass automation. How will we fill our time? Maybe some portion of the population will sit on the couch, drink beer, and watch reruns of Seinfeld ad infinitum. But I have more faith in us than that. I believe that we will begin, evermore rapidly, to solve the problems which have long perplexed humanity. More minds will be put to work against the problems of climate change, for example. Hopefully we’ll be able to invent and implement new responses to large societal issues like poverty, crime, sickness, pollution, the list goes on. But the true promise of increased human capacity goes beyond any one problem. By freeing our time and resources and redirecting them towards our largest problems, we’ll be able to focus on helping one another. Building and rebuilding communities. Engaging with each other emotional and spiritually. Being of service to our fellows. Ironically enough, I believe that mass automation will give us the capacity to be more human.

On top of all that, we get to reimagine the concept of work. What if we didn’t get up each day — 5 days a week — and sit in an office from 9 to 5. What if we engaged with the projects, the people, and the pursuits about which we’re most passionate? What if we did that always? And what if we were compensated not for our hours, but for our impact? What if everyone was guaranteed a universal basic income so that they could focus on these things? Making this shift will be difficult for many of us, but with strong, affordable retraining programs, millions of Americans will be granted opportunities that most of us can’t imagine today.

In the world of mass automation we will have more time than ever before. I don’t believe that this time will be wasted. I believe it will be invested. In self-expression. In art. In education. In service to one another. I believe that our newfound freedom will lead not to the destruction of our society, but to its elevation. So when I hear about a fully automated supermarket, I think not about our robot overlords, but about our potential as humans, and about achieving that potential.

(Thanks you to my trusted RRE colleague Cooper Zelnick for editing this post)

Cutting Through the Machine Learning Hype

Jason Black

The tech ecosystem is well acquainted with buzzwords. From “Web 2.0” to “cloud computing” to “mobile first” to “on-demand,” it seems as though each passing year heralds the advent and popularization of new catchphrases to which fledgling companies attach themselves. But while the trends these phrases represent are real, and category-defining companies will inevitably give weight to newly coined buzzwords, so too will derivative startups seek to take advantage of concepts that remain ill-defined by experts and little-understood by everyone else.

In a June post, CB Insights encapsulated the frenzy (and absurdity) of the moment:

It’s clear that 9 of 10 investors have very little idea what AI is so if you’re a founder raising money, you should sprinkle some AI into your pitch deck. Use of ‘artificial intelligence,’ ‘AI,’ ‘chatbot,’ or ‘bot’ are winners right now and might get you a little valuation bump or get the process to move quicker.

If you want to drive home that you’re all about that AI, use terms like machine learning, neural networks, image recognition, deep learning, and NLP. Then sit back and watch the funding roll in.

Pitch decks and headlines today are lousy with references to “artificial intelligence” and “machine learning”. But what do those terms really mean? And how can you separate empty claims from real value creation when evaluating businesses and the technologies which underpin them? Having at least a passing knowledge of what you’re talking about is a good first step, so let’s start with the basics.

Definitions

Artificial Intelligence

The terms “artificial intelligence” and “machine learning” are frequently used interchangeably, but doing so introduces imprecision and ambiguity. Artificial intelligence, a term coined in 1956 at a Dartmouth College CS conference, refers to a line of research that seeks to recreate the characteristics possessed by human intelligence.

At the time, “General AI” was thought to be within reach. People believed that specific advancements (like teaching a computer to master checkers or chess) would allow us to learn how machines learn, and ultimately program computers to learn like we do. If we could use machines to mimic the rudimentary way that babies learn about the world, the reasoning went, soon we would have a fully functioning “grown up” artificial intelligence that could master new tasks at a similar or faster rate.

In hindsight, this was a bit too optimistic.

While the end goal of AI was — and still is — the creation of a sentient machine consciousness, we haven’t yet achieved generalized artificial Intelligence. Moreover, barring a major breakthrough in methodology, we don’t have a reasonable timeline for doing so. As a result, research (especially the types of research relevant to the VC and startup world) now focuses on a sub-field of AI known as machine learning aimed at solving individual tasks which can increase productivity and benefit businesses today.

Machine Learning

In contrast with AI’s stated goal of recreating human intelligence, machine learning tools seek to create predictive models around specific tasks. Simply put, machine learning is all about utility. Nothing too flashy, just supercharged statistics.

While there are plenty of good definitions for machine learning floating around, my favorite is Tom M. Mitchell’s 1997 definition:

A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.

Rather formal, but this definition is buzzword-free and gets straight to the elegance and simplicity of machine learning. Simply put, a machine is said to learn if its performance at a set of tasks improves as it’s given more data.

Need an example? How about one from your Statistics 101 course: simple linear regression. The goal (or Task) is to draw a “line of best fit” given some initial set of observed data. Through an iterative process that seeks to minimize the average distance from the regression line and the scatterplot of data (its Performance measure), linear regression improves its predictive “line of best fit” with each additional data point (Experience).

Red dots represent scatter plot of all data. The blue line minimizes average distance from the regression line (represented here by grey lines).

Red dots represent scatter plot of all data. The blue line minimizes average distance from the regression line (represented here by grey lines).

Boom. Machine learning.

Given that relatively low bar, nearly any tech company can claim to be “leveraging machine learning.” So where do we go from here? To further demystify the topic, it’s also useful to understand how machine learning algorithms are developed. With linear regression, the algorithm in question simply draws a line which gets as close to as many individual data points as possible. But how about a real world example?

While the math behind more sophisticated machine learning models quickly becomes incredibly complex, the underlying concepts are often very intuitive.

Developing a Machine Learning Model

Say you wanted to predict what new songs a particular Spotify user would enjoy. Follow your intuition.

You’d probably start with his or her existing library and expect that other users who have a large number of songs in common would be likely to enjoy the complement set of the songs in the other user’s library (a process called collaborative filtering). You might also analyze the acoustic elements in the user’s library to look for common traits such as an upbeat tempo or use of electric guitar (Spotify uses neural networks to do this, for example). Finally you might assign an appropriate weight to the tracks a user has listened to repeatedly, starred, or marked with a thumbs up/down.

Check out this visualization of the filters learned in the first convolutional layer of Spotify’s deep learning algorithm. The time axis is horizontal, the frequency axis is vertical (frequency increases from top to bottom).

Check out this visualization of the filters learned in the first convolutional layer of Spotify’s deep learning algorithm. The time axis is horizontal, the frequency axis is vertical (frequency increases from top to bottom).

All that’s left is to translate these intuitions into a mathematical representation that ingests the requisite data sources and outputs a ranked list of songs to present to the user. As the user listens, likes, and dislikes new music, these new data points (or Experience in our earlier terminology) can be fed back into the same models to update, and thus improve, that prediction list.

If you want to learn more about more complex machine learning algorithms, there are ample resources across the web that do a great job of explainingneural networks, deep learning, Bayesian networks, hidden Markov models, and many more modeling systems. But for our purposes, technical implementation is less relevant than understanding how startups create value by harnessing that technology. So let’s keep moving.

Where’s the value?

Now that we have covered what machine learning is, for what should savvy investors and skeptical readers be on the lookout? In my experience, the initial litmus test is to walk through the three fundamental building blocks of a machine learning model (task T, performance measure P, and experience E) and look for new or interesting approaches. It is these novelties which form the basis of differentiated products and successful startups.

Experience | Unique Data Sets

Without data, you can’t train a machine learning model. Full stop.

With a publicly available training set, you can train a machine learning model to do specified tasks, which is great, but then you are relying on tuning and tweaking the performance of your algorithm to outperform others. If everyone is building machine learning models with the same sets of training data, competitive advantages (at least at the outset) are all but non-existent.

By contrast, a unique and proprietary data set confers an unfair advantage. Only Facebook has access to its Social Graph. Only Uber has access to the pickup/dropoff points of every rider in its network. These are data sets that only one company can use to train their machine learning models. The value of that is obvious. It’s basic scarcity of a private resource. And it can create an enormous moat.

Take *Digital Genius, as an example. The Company offers customer service automation tools and counts numerous Fortune 500 companies as clients. These relationships offer Digital Genius exclusive access to millions of historical customer service chat logs, which represent millions of appropriate responses to a wide swath of customer queries. Using this data, Digital Genius trains its Natural Language Processing (NLP) algorithms beforebeginning to interact with new, live customers.

In order to attain the same level of performance, a competitor would have to amass a similar number of chat logs from scratch. Practically speaking, this would require performing millions of live customer interactions, many of which would likely be frustrating and useless for the customers themselves. While the algorithm would eventually learn and improve, the model’s day one performance would be lackluster at best, and the company itself would be unlikely to gain traction in the market. Thus, having the proprietary data sets from their largest clients gives Digital Genius a real, differentiated value proposition in the chat automation space.

Of course, another way to go about gaining access to a unique data set is to capture one that has never existed. The coming wave of IoT and the proliferation of sensors promise to unlock troves of new data sets that have never before been analyzed. Companies which get proprietary access to new data sets, or those which create proprietary data sets themselves, can thus outperform the competition.

*OTTO Motors (a division of Clearpath Robotics), has captured one of the most robust data sets of indoor industrial environments on the planet from their network of autonomous materials transport robots (pictured below). Every time an OTTO robot makes its way around the factory floor, information about its environment — moving forklifts, walking workers, path obstructions — can be sent back to a centralized database. If the company then develops a more robust model to navigate around forklifts, for example, the OTTO Motors team can backtest and debug their improvements against real-world, historical environment data without needing to actually test their robots or even use physical environments.

An OTTO 1500 robot autonomously navigates around a warehouse.

An OTTO 1500 robot autonomously navigates around a warehouse.

This same data-race is even more competitive on the road. The reason why the Google Self-Driving Car, Tesla Autopilot, and Uber Self-Driving teams all tout (or forecast) the number of autonomous miles driven is because each additional mile captures valuable data about changing environments that engineers can then use to test against as they improve their autonomous navigation algorithms. But relative to the global total number of miles driven per year (an estimated 3.15 trillion miles in 2015 in the US alone), only a de minimus number of those are being captured by the three projects mentioned above, leaving greenfield opportunity for startups like Cruise AutomationnuTonomy, and Zoox.

The final, and most experimental approach to leveraging unique data sets is to programmatically generate data which is then used to train machine learning algorithms. This technique is best suited for creating data sets that are difficult or impossible to collect.

Here’s an example. In order to create a machine learning algorithm to predict the direction a person is looking in a real world environment, you first have to train on sample data that has gaze direction correctly labeled. Given the literal billions of images that we have of people looking, with their eyes open, in different directions in every conceivable environment, you’d think this would be a trivial task. The data set—it would seem—already exists.

The problem is that the data isn’t labeled, and manually labeling, let alone determining, a person’s exact gaze direction based on a photograph is way too hard for a human to do to any degree of accuracy or in a reasonable length of time. Despite possessing a vast repository of images, we can’t even create good enough approximations of gaze direction for a machine to train on. We don’t have a complete, labeled set of data.

Programmatically generated eyes used to train machine learning algorithms to determine gaze direction.

Programmatically generated eyes used to train machine learning algorithms to determine gaze direction.

In order to tackle this problem, a set of researchers at the University of Cambridge programmatically generated renderings of an artificial eye and coupled each image with its corresponding gaze direction. By generating over 10,000 images in a variety of different lighting conditions, the researchers generated enough labeled data to train a machine learning algorithm (in this case, a neural network) to predict gaze direction in photos of people the machine had not previously encountered. By programmatically generating a labeled data set, we sidestepped the problems inherent to our existing repository of real-world data.

While means of finding, collection, or generating data on which to train machine learning models are varied, evaluating the sources of data a company has access to (especially those which competitors can’t access) is a great starting point when evaluating a startup or its technology. But there’s more to machine learning than just Experience.

Task | Differentiated Approaches

Just as access to a unique data set is inherently valuable, developing a new approach to a machine learning Task (T) or starting work on a new or neglected Task provide alternative paths to creating value.

DeepMind, a company Google acquired for over $500M in 2014, developed a model generation approach that enabled them to pull ahead of the pack in a branch of machine learning known as deep learning (hence the name). While their acquisition went largely unnoticed by the mainstream press, it was difficult to miss the headlines as their machine learning algorithm dubbed “AlphaGo” squared off against the world champion of Go in early 2016.

The rules of the game of Go are relatively simple, yet the number of possible board positions in the game outnumber the atoms in the universe. Traditional machine learning techniques by themselves simply could not produce an effective strategy given the number of possible outcomes. However, DeepMind’s differentiated approach to these existing techniques enabled the team not only to best the current world champion of the game, Lee Sedol, but do so in such a way that spectators described the machine’s performance as “genius” and “beautiful.”

However, the sophistication of performance on one Task does not translate well to other domains. Use the same code from the AlphaGo project to respond to customer service enquiries or navigate around a factory floor and the performance would likely be abysmal. Practically, the approximate 1:1 ratio between Task and machine learning model means that for the short- and medium-term there are innumerable Tasks for which no machine learning model has yet been trained.

For this reason, identifying neglected Tasks can be quite lucrative and easier than one might expect. One might assume, for example, that since a significant amount of time, effort, and money has been spent on improving photo analysis, that video analysis has enjoyed the same performance gains. Not so. While some of the models from static image analysis have carried over, the complexity associated with moving images and audio has discouraged development, especially as plenty of low hanging fruit in the photo identification space still remains.

*Dextro’s Stream API annotating live Periscope videos in real time.

*Dextro’s Stream API annotating live Periscope videos in real time.

This created a great opportunity for *Dextro and Clarifai to quickly pull out ahead in applying machine learning to understand the content in videos. These advancements in video analysis now enables video distributors to create searchable videos based on not just the manually submitted metadata from the users who upload, but also the content contained within the video like the transcript of the video, the category of video, and even individual objects or concepts that appear throughout the video.

Performance | Step Function Improvement

The final major value driver for startups seeking to harness machine learning technology is meaningfully outperforming the competition at a known Task.

One great example is Prosper which makes loans to individuals and SMBs. Their Task is the same as any other lender on the market — to accurately evaluate the risk of lending money to a particular individual or business. Given that Prosper and their peers in both the alternative and the traditional lending world live or die by their ability to predict creditworthiness, Performance (P) is absolutely critical to the success of their business. So how do relatively young alternative lenders outperform even the largest financial institutions out there?

Instead of taking in tens of data points about a particular borrower Prosper draw an order of magnitude more data. In addition to using a larger and differentiated data set, the new wave of alternative lenders like Prosper have been rigorously scouring research papers and doing their own internal development in order to incorporate bleeding edge machine learning algorithms to their data sets. Together, the Performance characteristics of the resulting machine learning models represent a unique and differentiated ability to issue profitable loans to a whole group of consumers and businesses who have historically been turned away by legacy institutions.

Being able to judge the performance of a startup’s machine learning models against that of the competition is another great way to cull the most innovative companies and separate out the mere peddlers of hype and buzz.

Back to Business

To be clear, there’s much more to machine learning than hyped-up pitch decks and empty promises. The trick is culling the wheat from the chaff. Armed with clear definitions and a working knowledge of the simple concepts underlying the buzzwords and headlines, go forth and pick through presentations with confidence!

But remember this caveat.

Yes, machine learning — when harnessed appropriately — is both real and powerful. But the ultimate success or failure of any business hinges much more on the market opportunity, productization, and the team’s ability to sell than it does on specific implementations of machine learning algorithms. Just as compelling tech is a necessary but insufficient condition to create a successful tech company, great tech in the absence of a viable business is unlikely to become anything more than a science project.


Big thanks to Cooper Zelnick for being sounding board and an editor on this one. Shoutout to Ryan Atallah and Sven Kreiss for proofing for technical errors as well.

* Denotes an RRE portfolio company.

Talent Playbook

Maria Palma

Our founding teams always tell us that Talent is one of their biggest challenges.  While it's gotten easier and less costly to start companies in the last few decades, getting the right team is still a major challenge and yet is a core differentiator for your company.

Given that Hiring and Talent can often seem unapproachable and complicated, we put this Talent Playbook together as a starting point to cover some of the fundamentals. We interviewed Heads of Talent in our network to highlight a few best practices. This also includes an excel template to get you started on hiring when you aren't at a point to invest in more sophisticated systems yet.

Why we Invested In Nanit, a Computer Vision Baby Monitor

Will Porteous

We are extremely excited to announce our seed investment in Nanit, the next generation baby monitor leveraging computer vision. When I was first introduced to the co-founders of Nanit, Andrew Berman and Assaf Glazer, I was instantly impressed with their vision for the product and deep knowledge of computer vision.

Nanit fits a pattern we see time and time again in New York, where early founding teams are at the intersection of multiple disciplines. Assaf, a PhD in machine learning and computer vision, was able to form a strong team quickly, including former executives from Diapers.com, Philips Medical, and a former venture capitalist from Northwest Venture Partners. We have seen this pattern in prior entrepreneurs, such as Ben Kaufman, founder of Quirky (despite the company’s challenges) and Bre Pettis, founder of Makerbot. Both founders surrounded themselves with an interdisciplinary team to tackle unique challenges.

The Nanit team leverages technology to answer one of the most fundamental questions that all parents ask, “Is my child safe and sleeping well through the night?”. This question leaves many parents sleep deprived and exhausted.

On average, moms and dads lose 20.3 hours worth of sleep per week during the baby’s first year alone. -Daily Mail


I certainly remember being a new parent myself and wish there would have been such products.  All we had for baby monitors was fuzzy audio and occasional beeping, which left you with no idea what was actually going on. We certainly didn’t have baby monitors leveraging machine learning.

In the evolution of baby monitors, first, you could hear your child. Next, you could see your child. Now, with Nanit, you can hear, see, track and understand your child’s sleep patterns. This means parents can see not only whether their baby is sleeping, but how well their baby is developing.

All of the data collected is easily visible through the Nanit mobile app where parents can see trends over time and gain actionable insights. One of Nanit’s beta testers explained how she can check the app in the morning to see how her baby slept. If her child had a rough night of sleep, she would add in an extra nap during the day.

If you are interested in learning more or purchasing a Nanit for someone you know, they launched today and are being sold at a discounted price for a limited time here. We love companies that use hardware and software to build rich user experiences and are proud to support Nanit alongside some great partners, including Mark Suster from Upfront Ventures, Flextronics, and Jacobs Technion-Cornell Institute.

Blockchain and Media, a One-Two Punch

Alice Lloyd George

Why we invested in Mediachain

In 1967, three years after winning the world heavyweight championship, Muhammad Ali was stripped of his title after refusing to join the U.S. army. Ali was convicted of draft evasion and sentenced to five years in jail. At the height of the controversy, this photograph of Ali echoing martyr Saint Sebastian was taken for Esquire. It was one of the magazine’s most iconic covers, helping to fuel Ali’s status as a symbol of counterculture during the Vietnam war. The case of “the People’s Champion” worked its way up to the Supreme Court, and the conviction was eventually overturned.

My maternal grandfather, Carl Fischer, was a photographer in the 1960s. He worked with art director George Lois and shot that iconic cover as well as others for Esquire. Lois, famous for his art direction, was equally famous for stealing credit from the many creatives he worked with, including Julian Koenig, known for the Volkswagen “Think Small” ad. In 2008 the Museum of Modern Art had an exhibition of Lois’ work in which he was credited for both the photographs and the magazine covers. Carl’s name was not mentioned (the museum later corrected the attribution.)

Needless to say, when I met Mediachain co-founder Jesse Walden last year, his vision for giving power back to creators resonated. The road is littered with attempts to tackle the issue of attribution through digital rights management (DRM.) Traditional approaches have relied on erecting walls coupled with litigation. Now, permissionless technologies like the blockchain have inspired entrepreneurs to rethink infrastructures and come up with ways to take down those walls and share openly, while also preserving critical information.

One of the key learnings of the blockchain is that it is possible to exchange data or tokens (like bitcoin) without having to trust a central entity or middleman. Mediachain is an exciting application of this insight. So how does it work? Mediachain is an open-source protocol that allows parties to cryptographically sign statements about units of media, which is to say, any information about a song, image, or GIF, can be published to a blockchain, creating a tamper-proof record that can be referenced at any time. That information is written to a decentralized datastore, and perceptual resolvers (ie. visual Shazams) can be plugged in to query what is essentially a universal media library for a match. (More on how it works here.) It’s even possible to identify original images that have been edited, cropped, or repurposed. In a case like the Muhammad Ali shot, we would easily be able to see metadata for both Carl Fischer as creator of the image, and George Lois as art director who used the image in a magazine cover.

The Mediachain team is working with established organizations like Getty Images, the Digital Public Library of America (DPLA), MOMA and Europeana. A registry of existing creative commons images is just the first step. The protocol will also enable anyone to automatically claim authorship at the time of image creation, e.g. as soon as you snap an Instagram. This is significant, as we are living in an era of explosive online content growth. Benedict Evans at a16z recently estimated that in 2015 alone, 2–3 trillion photos were shared across Facebook, WhatsApp, Instagram and Snapchat. But while the volume and velocity of user-generated content has taken off, the intelligence we have about that content has not kept up. Our ability to discover and understand media (via computer vision) and attribute it (via metadata) is going to unlock all kinds of untapped value. Imagine if, while reading an article, you could easily tip the creator of a GIF that’s embedded. Or if you could get credit when aggregators, like the Fat Jew or FuckJerry, steal your joke and take it viral.

There are a few other innovative projects extending the blockchain metaphor to other areas. These include OpenBazaar, a decentralized e-commerce platform that has already been downloaded 100,000 times since mid-May, as well as Blockstack, which offers decentralized services for identity and authentication, and 21.co, which is making a future of machine-to-machine payments possible [disclosure — RRE is an investor]. Critics of such projects call them “academic.” But many technologies that are revolutionary today started out looking like research projects, notably ARPANET, circa 1968. In the case of Mediachain, a decentralized and persistent metadata layer makes a number of interesting things immediately possible:

  • Real connectivity can be built between platforms. Today platforms are fragmented — their data is siloed and not interoperable.
     
  • Creators and their audiences can now connect directly. This reestablishes a flow of value that was previously broken. Destination platforms have been built that do a great job at this, such as Patreon, started by my friend Sam Yam, as well as the Kickstarters and Indiegogos of the world. Mediachain can enable the distributed equivalent, a way to establish the connection between creators and their audiences while those audiences are naturally roaming the web.
     
  • Attribution also helps publishers. Spotify recently paid a $30 million settlement to resolve a lawsuit for royalties that they withheld because they didn’t know whom to pay.
     
  • With true data hygiene comes valuable analytics. This can help publishers and institutions that want to better track how their content gets circulated. Just over a year ago, our portfolio company Buzzfeed played a key role in how The Dress went viral, with 38 million views of their main post and visits from every country in the world within 12 hours — see the GIF below, which gives a rough idea of how the image blew up internationally in just a few days.

When it comes to analytics, not all organizations are as sophisticated as Buzzfeed. Mediachain’s solution will enable us to track how content flows through the web, and help publishers better understand how, where, and when their content gets shared.

This just scratches the surface for how we might capture cultural value as it increasingly goes digital. The question is what kinds of new applications people might come up with on top of the open data shared in Mediachain. Developers, creators and rights-owners interested in contributing can get involved on GitHub or their public Slack.

Today we are announcing RRE’s participation in Mediachain Labs’ seed financing. I’m thrilled we get to work with companies like Mediachain and am excited to see the team build out their ambitious vision.

How We Think About New Media Companies

Will Porteous

Given all the recent assertions in the press about a “Digital Media Bloodbath,” this seemed like a good time to affirm our commitment to media investing and to explain our media investment thesis. Media has been a focus for our firm since our first fund more than twenty years ago. While each year we consider a lot of media related investment opportunities, we make relatively few investments in the area. This is largely because we rarely see new companies innovating broadly enough to build major long term, independent businesses. Innovation in media often begins with content, but building a successful new media business depends just as much on innovation in distribution and in the advertiser’s experience. Drawing on some examples from our portfolio (including Business Insider, theSkimm, BuzzFeed, and Giphy) and from companies we admire, let me try to explain why.

Engaging Content

Most promising new media companies start with original content. If they are successful, they will begin to capture a meaningful share of an audience’s time. Our successful media founders have typically been strongly creative people with experience in the legacy media industry. They know how to fashion something engaging, often out of familiar material. Henry Blodget, Founder and CEO of Business Insider, recognized that telling a story in a timely way through charts (often arranged sequentially as slideshows) allowed an audience to engage directly with the data and reach their own conclusions quickly. This high-velocity storytelling helped propel Business Insider to the #1 spot in daily online business news. Similarly, the founders of theSkimm recognized a need while working at NBC when they were constantly recapping the news for their female friends. Building on that experience, they have achieved breakthrough engagement and reach with millennial women (and some men), by approaching the morning news with a fresh voice and a format that’s easy to consume on mobile. As a result, their subscriber growth and open rates have reached extraordinary levels. But great content is just one critical element in creating a truly innovative media company. Founders need to focus just as much on creating innovative distribution for their original content and on creating a fundamentally better experience for advertisers.

Efficient Distribution

We live in a post-portal world in which content travels freely through different messaging channels (email, text, Snapchat, Slack, a multitude of messaging services) and across major social media platforms (Facebook, Twitter, Google, YouTube). Successful distribution today depends on promotion, broad appeal, and your audience’s willingness to engage, react and share. Understanding these platforms and the forces that drive engagement and proliferation are critical if you want your content to achieve meaningful distribution.

We’ve been investors and on the board of BuzzFeed since May of 2010. The power of sharing as a mode of distribution is one of BuzzFeed founder Jonah Peretti’s great insights. He recognized early on that as platforms become a bigger and bigger part of the digital ecosystem, sharing of engaging content would increase, and that content would exist in many different formats. BuzzFeed has made early, and sometimes surprising, investments that have helped build a global, cross-platform network powered by technology and data science. That network has allowed BuzzFeed’s team to cover a breaking news event like the November 2015 attacks in Paris on par with other global media organizations like CNN or the BBC, reaching millions of people around the world through their site, apps, and distributed video. It’s also brought about creative programming and experimentation with, say, exploding fruit or mesmerizing recipes that resonate with millions of people. The company understands how audiences engage with content and why it is shared. This has given Buzzfeed a powerful cultural currency.

Understanding search as a means of distribution is just as important. To this end, we are big admirers of companies like Bustle. Founder and CEO Bryan Goldberg’s deep understanding of search engine optimization and the search habits of his core demographic — teen and young millennial women — has enabled the company’s content to achieve incredible durability and reach long past the date of publication. The associated ad revenue opportunity has similar endurance, particularly given Bustle’s rich, almost “glossy,” ad format. While treating search as a distribution mechanism, the company has achieved an almost recurring revenue stream on the pieces it has previously published, propelling it to terrific revenue growth.

And then there are companies that recognize that controlling distribution of great content in a new format can be just as powerful a business enabler. We’ve always loved GIFs because of the way they tell emotionally engaging stories in a short, easily shareable format. When they first started, the founders of Giphy created a lot of engaging GIFs. But they soon recognized that all valuable historic video content would eventually be atomically reduced to a series of GIFs (at least the good parts). Fortunately for Giphy, organizing all this new content requires powerful new search capabilities that fall outside of traditional web search. They have focused relentlessly on creating those capabilities and indexing this vast new content landscape.

BuzzFeed, Bustle, and Giphy have all innovated considerably around distribution and this has helped propel their growth. We are deeply interested in new forms of distribution in themselves, particularly ways to reach strong affinity groups and self-organizing communities.

Effective Advertising

For more than 20 years we’ve been fascinated by the slow pace of innovation in ad formats and experiences. For media businesses that depend on advertising, this seems counterintuitive since better ad experiences have the potential to unlock major new revenue streams. Innovating around the advertiser’s experience has actually created a vast amount of value and we wish we saw more companies focused on this.

Through AdWords, Google gave advertisers a way to reach specific audiences that had declared their interests and intentions, dramatically increasing the efficiency of cost-per-click advertising and breathing new life into display. Unfortunately, this still left brand advertisers searching for a better way to tell their story. Facebook promised the beginning of a better relationship for brands, with sponsored stories reflecting our accumulated likes and the collective wisdom of one’s social graph. And yet, even today, we still see an awful lot of poorly targeted display advertising on Facebook. And the platform takes an incredibly artificial approach to our relationship with a brand: Personally, while I may shop at Target (and even like their brand values), I’m never going to take the time to visit and “like” their page.

Across the edges of our screens, the old display advertising model has been dying for some time now. Display ads are awful for telling a brand’s story. The ad exchanges have wrung the last bit of efficiency out of this miserable format. While they have their utility when an audience’s intent can be clearly discerned, all too often they appear where it can’t.

We believe that in every media format, the important thing is to give the brand advertiser a means to engage the audience emotionally, just as the content does. At BuzzFeed, Jonah Peretti recognized this and drove the creation of the sponsored content model. This gave brands a fresh medium to tell a story through shareable content, while holding them accountable for engaging the audience. Today, there are more ways than ever for publishers to be true partners to a brand, helping hone their voice and find their targeted audiences. BuzzFeed’s test-and-learn approach lets brands explore new formats like quizzes, shorter-than-short form content, and no-sound-required video. These explorations have opened up entirely new revenue streams for BuzzFeed in the form of cross-platform branded content packages that are integrated with the company’s owned and operated site and apps. It’s one of the reasons the company has continued to grow at a terrific rate.

Today, we welcome the ad innovations from SnapChat, with Lenses that overlay a brand lifestyle on our own and publisher driven Discover channels that include short, high-quality video ads. Now it feels like we are getting something new that works for brands, though a lot of what’s happening in those channels feels a lot like TV. We look forward to seeing new innovations in ad experiences that captivate audiences and advertisers alike.

So What’s Next?

We will continue to look for new media companies innovating across content, distribution and advertising. For us, those are the essential elements of building major long term, independent media businesses. We know that platforms and formats will continue to change and this will create opportunities for new entrants. With the shift to mobile, this seems to be happening on a greater scale than ever before. And, as it becomes cheaper and cheaper to produce and stream high quality video, new online media businesses are leading with everything from GIFs, 8-second Snapchat videos, 45-second How-To segments, to 3-minute serials.

Judging by the deals that have caught our attention recently, we expect to continue to see innovation in news — particularly the decentralized sourcing and curation of “stories.” We also think there are still huge opportunities for media companies to share data with advertisers about the audience they are reaching without compromising privacy. In fact, we think this is essential for engaging advertisers in any new platform or ad experience.

It has been a privilege for us to work with so many great media entrepreneurs and we look forward to partnering with the next generation in the years to come.