All posts in “Startups”

Here’s why Spotify will go public via direct listing on April 3rd

Spotify explained why it’s ditching the traditional IPO for a direct listing on the NYSE on April 3rd today during its Investor Day presentation. With no lockup period and no intermediary bankers, Spotify thinks it can go public without all the typical shenanigans.

Spotify described the rationale for using a direct listing with five points:

  • List Without Selling Shares  – Spotify has plent of money with $1.3 billion in cash and securities, has no debt since it converted that into equity for investors, and has positive free cash flow
  • Liquidity – Investors and employees can sell on public market and sell at time of their choosing without investors shorting a lockup expiration, while new investors can join in
  • Equal Access – Bankers won’t get preferred access. Instead, the whole world will get access at the same time. “No underwriting syndicate, no limited float, no IPO allocations, no preferential treatment”.
  • Transparency – Spotify wants to show the facts about its business to everyone via today’s presentation, rather than giving more info to bankers in closed door meetings
  • Market-Driven Price Discovery – Rather than setting a specific price with bankers, Spotify will let the public decide what it’s worth. “We think the wisdom of crowds trumps expert intervention”.

Spotify won’t wait for the direct listing, and on March 26th will announce first quarter and 2018 guidance before markets open. It also announced today that there will be no lock-up period, so employees can start selling their shares immediately. This prevents a looming lock-up period expiration that can lead to a dump of shares on the market that sinks the price from spooking investors.

It’s unclear exactly what Spotify will be valued at on April 3rd, but during 2018 its shares have traded on the private markets for between $90 and $132.50, valuing the company at $23.4 billion at the top of the range. The music streaming service now has 159 million monthly active users (up 29 percent in 2017) and 71 million paying subscribers (up 46 percent in 2017.

During CEO Daniel Ek’s presentation, he explained that Spotify emerged as an alternative to piracy by convenience to make paying or ad-supported access easier than stealing. Now he sees the company as the sole leading music streaming service that’s a dedicated music company, subtly throwing shade at Apple, Google, and Amazon. “We’re not focused on selling hardware. We’re not focused on selling books. We’re focused on selling music and connecting artists with fans” said Ek.

Head of R&D Gustav Soderstrom outlined Spotify’s ubiquity strategy, opposed to trying to lock users into a “single platform ecosystem”. He says Spotify does “what’s best for the user and not for the company, and trying to solve the users’ problems by being everywhere.” That’s more shade for Apple, who’s HomePod only works with Apple Music despite customers obviously wishing they could play other streaming services through it.

By now being baked into a wide range of third-party hardware through the Spotify Connect program, Soderstrom says Spotify gets a more holistic understanding of its listeners. He declared that Spotify has 5X as much personalization data as its next closest competitor, and that allows it to know what to play you next. He cheekily calls this “self-driving music”.

By directing what people listen to, Spotify becomes the new top 40 radio — the hit-maker. That gives it leverage over the record labels so Spotify can get better licensing deals and favorable treatment. Now over 30 percent of Spotify listening is based on its own programming through featured playlists, artists, and more.

Spotify CEO Daniel Ek giving the Investor Day presentation

Wall Street loves a two-sided marketplace, so Spotify is positioning itself in the middle of artists and fans, with each side attracting the other. It’s both selling music streaming services to listeners, and selling the tools to reach and monetize those listeners to musicians. That’s both on its platform, and using its targeting and analytics info to deliver efficient ticket and merchandise promotions elsewhere. Ek discussed the flywheel that drives Spotify’s business, explaining that the more people discover music, the more they listen, and the more artists that become successful on the platform, and the more artists will embrace the platform and bring their fans.

Yet with music catalogues and prices mostly similar across the industry, Spotify will have to depend on its personalized recommendations and platform-agnositic strategy to beat its deep pocketed competitors. Music isn’t going away, so whoever can lock in listeners now at the dawn of streaming could keep coining off them for decades. That’s why Spotify not raising cash for marketing through a traditional IPO is a strange choice. But with its focus on playlists and suggestion data, Spotify could build melodic handcuffs for its listeners who wouldn’t dream of starting from scratch on a competitor.

You can follow along with the presentation here.

For more on Spotify’s not-an-IPO, check out our feature piece:

We want to hear about your robotics company

As you might have heard, last year’s TC Robotics event in Boston was such a hit we’ve decided to do it again — only on the West Coast, this time. On Friday, May 11, we’ll be holding TC Sessions: Robotics on the U.C. Berkeley campus. We’ve got a lot of big industry luminaries lined up that we can’t wait to tell you about, but in the meantime, we’d like to hear from you.

We’re going to have several opportunities for robotics companies to show off their goods in the lead up to and the event itself. We’re looking for the best and brightest in the robotics world — both startups and established companies alike. If you’ve got a technology you think will wow us, we want to hear from you.

Specifically, we’re looking for technology that will make for great videos and stage demos. We’re also searching for startups who are interested in participating in a pitch competition. Bonus points for new technologies we haven’t seen before — and for companies based in and around the Bay Area. Think you’ve got what it takes? Fill out the form below. We’ll reach out to those companies that meet the criteria.

More information on the upcoming TC Sessions: Robotics event can be found here.

The red-hot AI hardware space gets even hotter with $56M for a startup called SambaNova Systems

Another massive financing round for an AI chip company is coming in today, this time for SambaNova Systems — a startup founded by a pair of Stanford professors and a longtime chip company executive — to build out the next generation of hardware to supercharge AI-centric operations.

SambaNova joins an already quite large class of startups looking to attack the problem of making AI operations much more efficient and faster by rethinking the actual substrate where the computations happen. While the GPU has become increasingly popular among developers for its ability to handle the kinds of lightweight mathematics in very speedy fashion necessary for AI operations. Startups like SambaNova look to create a new platform from scratch, all the way down to the hardware, that is optimized exactly for those operations. The hope is that by doing that, it will be able to outclass a GPU in terms of speed, power usage, and even potentially the actual size of the chip. SambaNova today said it has raised a massive $56 million series A financing round led by GV, with participation from Redline Capital and Atlantic Bridge Ventures.

SambaNova is the product of technology from Kunle Olukotun and Chris Ré, two professors at Stanford, and led by former SVP of development Rodrigo Liang, who was also a VP at Sun for almost 8 years. When looking at the landscape, the team at SambaNova looked to work their way backwards, first identifying what operations need to happen more efficiently and then figuring out what kind of hardware needs to be in place in order to make that happen. That boils down to a lot of calculations stemming from a field of mathematics called linear algebra done very, very quickly, but it’s something that existing CPUs aren’t exactly tuned to do. And a common criticism from most of the founders in this space is that Nvidia GPUs, while much more powerful than CPUs when it comes to these operations, are still ripe for disruption.

“You’ve got these huge [computational] demands, but you have the slowing down of Moore’s law,” Olukotun said. “The question is, how do you meet these demands while Moore’s law slows. Fundamentally you have to develop computing that’s more efficient. If you look at the current approaches to improve these applications based on multiple big cores or many small, or even FPGA or GPU, we fundamentally don’t think you can get to the efficiencies you need. You need an approach that’s different in the algorithms you use and the underlying hardware that’s also required. You need a combination of the two in order to achieve the performance and flexibility levels you need in order to move forward.”

While a $56 million funding round for a series A might sound massive, it’s becoming a pretty standard number for startups looking to attack this space, which has an opportunity to beat massive chipmakers and create a new generation of hardware that will be omnipresent among any device that is built around artificial intelligence — whether that’s a chip sitting on an autonomous vehicle doing rapid image processing to potentially even a server within a healthcare organization training models for complex medical problems. Graphcore, another chip startup, got $50 million in funding from Sequoia Capital, while Cerebras Systems also received significant funding from Benchmark Capital.

Olukotun and Liang wouldn’t go into the specifics of the architecture, but they are looking to redo the operational hardware to optimize for the AI-centric frameworks that have become increasingly popular in fields like image and speech recognition. At its core, that involves a lot of rethinking of how interaction with memory occurs and what happens with heat dissipation for the hardware, among other complex problems. Apple, Google with its TPU, and reportedly Amazon have taken an intense interest in this space to design their own hardware that’s optimized for products like Siri or Alexa, which makes sense because dropping that latency to as close to zero as possible with as much accuracy in the end improves the user experience. A great user experience leads to more lock-in for those platforms, and while the larger players may end up making their own hardware, GV’s Dave Munichiello — who is joining the company’s board — says this is basically a validation that everyone else is going to need the technology soon enough.

“Large companies see a need for specialized hardware and infrastructure,” he said. “AI and large-scale data analytics are so essential to providing services the largest companies provide that they’re willing to invest in their own infrastructure, and that tells us more investment is coming. What Amazon and Google and Microsoft and Apple are doing today will be what the rest of the Fortune 100 are investing in in 5 years. I think it just creates a really interesting market and an opportunity to sell a unique product. It just means the market is really large, if you believe in your company’s technical differentiation, you welcome competition.”

There is certainly going to be a lot of competition in this area, and not just from those startups. While SambaNova wants to create a true platform, there are a lot of different interpretations of where it should go — such as whether it should be two separate pieces of hardware that handle either inference or machine training. Intel, too, is betting on an array of products, as well as a technology called Field Programmable Gate Arrays (or FPGA), which would allow for a more modular approach in building hardware specified for AI and are designed to be flexible and change over time. Both Munichiello’s and Olukotun’s arguments are that these require developers who have a special expertise of FPGA, which a sort of niche-within-a-niche that most organizations will probably not have readily available.

Nvidia has been a massive benefactor in the explosion of AI systems, but it clearly exposed a ton of interest in investing in a new breed of silicon. There’s certainly an argument for developer lock-in on Nvidia’s platforms like Cuda. But there are a lot of new frameworks, like TensorFlow, that are creating a layer of abstraction that are increasingly popular with developers. That, too represents an opportunity for both SambaNova and other startups, who can just work to plug into those popular frameworks, Olukotun said. Cerebras Systems CEO Andrew Feldman actually also addressed some of this on stage at the Goldman Sachs Technology and Internet Conference last month.

“Nvidia has spent a long time building an ecosystem around their GPUs, and for the most part, with the combination of TensorFlow, Google has killed most of its value,” Feldman said at the conference. “What TensorFlow does is, it says to researchers and AI professionals, you don’t have to get into the guts of the hardware. You can write at the upper layers and you can write in Python, you can use scripts, you don’t have to worry about what’s happening underneath. Then you can compile it very simply and directly to a CPU, TPU, GPU, to many different hardwares, including ours. If in order to do work you have to be the type of engineer that can do hand-tuned assembly or can live deep in the guts of hardware there will be no adoption… We’ll just take in their TensorFlow, we don’t have to worry about anything else.”

(As an aside, I was once told that Cuda and those other lower-level platforms are really used by AI wonks like Yann LeCun building weird AI stuff in the corners of the Internet.)

There are, also, two big question marks for SambaNova: first, it’s very new, having started in just November while many of these efforts for both startups and larger companies have been years in the making. Munichiello’s answer to this is that the development for those technologies did, indeed, begin a while ago — and that’s not a terrible thing as SambaNova just gets started in the current generation of AI needs. And the second, among some in the valley, is that most of the industry just might not need hardware that’s does these operations in a blazing fast manner. The latter, you might argue, could just be alleviated by the fact that so many of these companies are getting so much funding, with some already reaching close to billion-dollar valuations.

But, in the end, you can now add SambaNova to the list of AI startups that have raised enormous rounds of funding — one that stretches out to include a myriad of companies around the world like Graphcore and Cerebras Systems, as well as a lot of reported activity out of China with companies like Cambricon Technology and Horizon Robotics. This effort does, indeed, require significant investment not only because it’s hardware at its base, but it has to actually convince customers to deploy that hardware and start tapping the platforms it creates, which supporting existing frameworks hopefully alleviates.

“The challenge you see is that the industry, over the last ten years, has underinvested in semiconductor design,” Liang said. “If you look at the innovations at the startup level all the way through big companies, we really haven’t pushed the envelope on semiconductor design. It was very expensive and the returns were not quite as good. Here we are, suddenly you have a need for semiconductor design, and to do low-power design requires a different skillset. If you look at this transition to intelligent software, it’s one of the biggest transitions we’ve seen in this industry in a long time. You’re not accelerating old software, you want to create that platform that’s flexible enough [to optimize these operations] — and you want to think about all the pieces. It’s not just about machine learning.”

Airtable raises $52M to give non-coders tools to build complex software

A massive company probably has plenty of engineers on staff and the resources to build a complex backbone of interconnected information that can contain tons of data and make acting on it easy — but for smaller companies, and for those that aren’t technical, those tools aren’t very accessible.

That’s what convinced Howie Liu to create Airtable, a startup that looks to turn what seems like just a normal spreadsheet into a robust database tool, hiding the complexity of what’s happening in the background while those without any programming experience create intricate systems to get their work done. Today, they’re trying to take that one step further with a new tool called Blocks, a set of mix-and-match operations like SMS and integrating maps that users can just drop into their systems. Think of it as a way to give a small business owner with a non-technical background to meticulously track all the performance activity across, say, a network of food trucks by just entering a bunch of dollar values and dropping in one of these tools.

“We really want to take this power you have in software creation and ‘consumerize’ that into a form anyone can use,” Liu said. “At the same time, from a business standpoint, we saw this bigger opportunity underneath the low-code app platforms in general. Those platforms solve the needs of heavyweight expensive use cases where you have a budget and have a lot of time. I would position Airtable making a leap toward a graphical user interface, versus a lot of products that are admin driven.”

Liu said the company has raised an additional $52 million in financing in a round led by CRV and Caffeinated Capital, with participation from Freestyle Ventures and Slow Ventures. All this is going toward a way to build a system that is trying to abstract out even the process of programming itself, though there’s always going to be some limited scope as to how custom of a system you can actually make with what amounts to a set of logic operation legos. That being said, the goal here is to boil down all of the most common sets of operations with the long tail left to the average programmers (and larger enterprises often have these kinds of highly-customized needs).

All this is coming at a time when businesses are increasingly chasing the long tail of small- to medium-sized businesses, the ones that aren’t really on the grid but represent a massive market opportunity. Those businesses also probably don’t have the kinds of resources to hire engineers while companies like Google or Facebook are camping out on college campuses looking to snap up students graduating with technical majors. That’s part of the reason why Excel had become so popular trying to abstract out a lot of complex operations necessary to run a business, but at the same time, Liu said that kind of philosophy should be able to be taken a step further.

“If you look at cloud, you have Amazon’s [cloud infrastructure] EC2, which abstracted the hardware level and you can build on existing machine intelligence,” Liu said. “Then, you get the OS level and up. Containers, Heroku, and other tools have extracted away the operation level complexity. But you have to write the app and modal logic. Our goal is to go a big leap forward on top of that and abstract out the app code layer. You should be able to directly use our interface, and blocks, all these plug and play lego pieces that give you more dynamic functionality — whether a map view or an integration with Twilio.”

And, really, all these platforms like Twilio have tried to make themselves pretty friendly to coding beginners as-is. Twilio has a lot of really good documentation for first-time developers to learn to use their platforms. But Airtable hopes to serve as a way to interconnect all these things in a complex web, creating a relational database behind the scenes that users can operate on in a more simplistic matter that’s still accurate, fast, and reliable.

“Obviously MySQL is great if you want to use code or custom SQL queries to interface with the data,” Liu said. “But, ultimately, you’d never as a business end user consider using literally a terminal-based SQL prompt as the primary interface to and from your data. Certainly you wouldn’t put that on your designs. Clearly you would want some interface on top of the SQL level database. We basically expose the full value of a relational database like Postgres to the end user, but we also give them something equally but more important: the interface on the top that makes the data immediately visible.”

There’s been a lot of activity trying to rethink these sort of fundamental formats that the average user is used to, but are ripe for more flexibility. Coda, a startup trying to rethink the notion behind a word document, raised $60 million, and all this points towards moves to try to create a more robust toolkit for non-technical users. That also means that it’s going to be an increasingly hot space, and especially look like an opportunity for companies that are already looking to host these kinds of services online like Amazon or Microsoft and have the buy-in from those businesses.

Liu, too, said that the goal of the company was to go after all potential business cases right away by creating a what-you-see-is-what-you-get one size fits all platform — which is usually called a horizontal approach. That’s often a very risky move, and it’s probably the biggest question mark for the company as there’s an opportunity for some other startups or companies to come in and grab niches of that whole pie in specific areas (like, say, a custom GUI programming interface for healthcare). But Liu said the opportunity for Airtable was to go horizontal from day one.

“There’s this assumption that software has to involve literally writing code,” Liu said. “It’s sort of a difficult thing to extricate ourselves from because we have built so much with writing code. But when you think about what goes into a useful application, especially in the business-to-business internal tools in a company use case which forms the bulk of software that’s consumed in terms of lines of code written, most of them are primarily a relational database model, and the relational database aspect of it is not an arbitrary format.

Sierra Leone just ran the first blockchain-based election

The citizens of Sierra Leone went to the polls on March 7 but this time something was different: the country recorded votes at 70% of the polling to the blockchain using a technology that is the first of its kind in actual practice.

The tech, created by Leonardo Gammar of Agora, anonymously stored votes in an immutable ledger, thereby offering instant access to the election results.

“Anonymized votes/ballots are being recorded on Agora’s blockchain, which will be publicly available for any interested party to review, count and validate,” said Gammar. “This is the first time a government election is using blockchain technology.”

“Sierra Leone wishes to create an environment of trust with the voters in a contentious election, especially looking at how the election will be publicly viewed post-election. By using blockchain as a means to immutably record ballots and results, the country hopes to create legitimacy around the election and reduce fall-out from opposition parties,” he said.

Why is this interesting? While this is little more than a proof of concept – it is not a complete voting record but instead captured a seemingly acceptable plurality of votes – it’s fascinating to see the technology be implemented in Sierra Leone, a country of about 7.4 million people. The goal ultimately is to reduce voting costs by cutting out paper ballots as well as reducing corruption in the voting process.

Gammar, for his part, sees the value of a decentralizes system.

“We’re the only company in the world that has built a fully-functional blockchain voting platform. Other electronic voting machines are ‘block boxes’ that have been increasingly shown to be vulnerable to security attacks. For that reason, many US states and foreign nations have been moving back to paper,” he said. “If you believe that most countries will use some form of digital voting 50 years from now, then blockchain is the only technology that has been created which can provide an end-to-end verifiable and fully-transparent voting solution for this future.”

One election in one country isn’t a movement – yet. However, Gammar and his team plan on expanding their product to other African countries and, eventually, to the rest of the world.

As for the election it is still unclear who won and there will be a run-off election on March 27. The winner will succeed President Ernest Bai Koroma who has run the country for a full decade.