OK, the competition isn’t fierce. Just four technology companies have gone public in the US this year. But the fifth one, Twilio, which made its IPO intentions known last week, is bound to be the most interesting.
Because the closer you look at Twilio, the more it represents a new kind of tech company for a new era of tech.
All of the information in this piece is derived from Twilio’s IPO prospectus and other public sources. The startup, privately valued at $1 billion last year, intends to raise up to $100 million by going public on the New York Stock Exchange in June under the ticker symbol TWLO.
Twilio, launched in 2008, has never turned a profit. But its losses have narrowed, and revenue grew to $166.9 million in 2015. Revenue last quarter was up 78% from a year ago, while the net loss shrank by 25%.
In many respects, Twilio is just another startup in the broad category known as software-as-a-service, or SaaS. But that underplays the startup and what it represents. To understand Twilio’s place on the internet, let’s take a closer look at the soon-to-be-public company.
Twilio calls itself a “cloud communications platform,” but remove a layer of jargon, and what Twilio does is send text messages on behalf of other services. When you receive an SMS to confirm your phone number, remind you of an appointment, or generate a password, there’s a good chance it was sent by Twilio, which has 28,648 paying clients.
SMS may have peaked in most of the world, usurped by messaging apps, but it’s still the backbone of so many mobile internet services. These apps generally rely on phone numbers, instead of email addresses, to identify users, and only the ability to text unites them all.
Take WhatsApp, the messaging app said to have killed SMS by offering the same functionality for free. It is among Twilio’s top customers, representing 17% of revenue last year, or $28.4 million. That’s generated entirely from WhatsApp sending text messages to verify the identity of its more than 1 billion active users when they log in.
Twilio has quietly provided the SMS backend for WhatsApp since 2013, around the time the messaging app began a period of explosive growth and a year before it was acquired by Facebook for $19 billion.
Of course, relying on one company for so much of its revenue is a risk for Twilio, especially since WhatsApp hasn’t signed any commitment to keep using the service. Three of Twilio’s top 10 customers currently aren’t under long-term contracts. For that reason, Twilio distinguishes between “variable revenue” from clients like WhatsApp and “base revenue” from clients that have signed contracts of a year or longer. The latter accounted for a healthy 84% of total revenue last quarter.
Twilio argues persuasively that its performance is best measured by base revenue from ongoing clients that have signed long-term commitments. It notes, for instance, that the group of clients it signed up in 2012 have doubled their spend every year since. Last quarter, revenue from long-term contracts was up 170% from the same customers a year prior, a metric it calls “dollar-based expansion rate.”
Think about that dynamic the next time a business-to-business service asks your company to sign a one- or two-year contract at a substantial discount. It’s valuing the certainty of extended business over the flexibility of letting customers come and go as they please. It may also be polishing its books in advance of raising money from investors.
To see how Twilio is striking that balance, you can look at its gross profit margin, which is a measure of the company’s efficiency. Twilio has never made money: It lost $38.9 million last year. But gross profit only accounts for expenses related to generating more revenue—for instance, it has to pay more to telecommunications networks in order to send more text messages. If the gross margin were to decrease, it could be a sign of unsustainable growth.
That doesn’t seem to be happening. Twilio’s gross margin was 55% last quarter, up slightly from a year ago. It says the dip in late 2014 was due to a price cut, but the company clearly hasn’t been discounting as a primary method of achieving growth.
These metrics—base revenue, dollar-based expansion rate, gross profit margin—are all pretty standard for software-as-a-service. They are also used to evaluate public companies like Zendesk, which provides software for customer service, and Atlassian, which provides software for group chat and project management. They are all hosted services that charge more as they are used more. The strategy is to become indispensable to the operations of growing businesses.
But Twilio is no ordinary SaaS startup. It differs from most of the others because Twilio isn’t really used directly by people. There is not much of an interface for interacting with Twilio, the way there is for Zendesk or Atlassian. Twilio is instead used by apps and services. It is plumbing.Twilio is more like plumbing than software.
Twilio’s products are its APIs, primarily the ones for programmatically sending text messages and phone calls. Say you’re a developer, and the boss says, “Build a service that texts the weather to people every morning.” You could do it all yourself, handling all the complexities of international phone numbers, regulations, and so on. Or you could use Twilio for that, and focus on the weather forecast, which is hard enough.
Uber has been using Twilio in much the same way since 2011. Twilio sends texts to Uber customers to confirm their identity or send updates about their ride. It also connects drivers and riders by voice or text without revealing either person’s phone number.
That’s notable because Uber is otherwise considered a “full-stack startup,” meaning it owns all critical aspects of its business. For instance, Uber’s lack of proprietary mapping software was considered a liability and led it to make investments there. It is spending heavily on research and development in robotics and self-driving cars to better control those areas, as well. But the infrastructure for messaging customers is not something Uber has felt compelled to bring in-house.
“We built the Uber experience without Twilio initially, and the problem was, people were not getting the high quality experience we were promising,” Uber CEO Travis Kalanick says in a glowing testimonial on Twilio’s website. “I sleep easier, and my engineers sleep easier because we’re not dealing with situations where it’s taking 15 to 20 minutes for a text to be delivered.”
That all makes Twilio sound less like SaaS and more like a cloud hosting service. The best comparison is perhaps to Amazon Web Services (AWS). Nowadays, most startups use AWS or its competitors for hosting instead of worrying about administering servers themselves. It’s cheaper, more reliable, and easier to scale than going it alone.
So if Twilio is akin to AWS, then here’s the really fascinating part: Twilio hosts “substantially all” of its own service on AWS.
Twilio customers, in other words, are outsourcing messaging to Twilio, which in turn outsources to Amazon. That’s either an inefficient quirk, or it points to a new kind of company in the startup stack, a layer in between infrastructure and software. Call it AWS as a service—AWSaaS?Call it AWS as a service—or, awkwardly, AWSaaS.
There are other examples of this. Stripe facilitates payments for other companies, handling all the mess of credit card transactions, but it hosts the service with Amazon. Wit.ai, which was recently acquired by Facebook, performs natural language processing using AWS servers.
You could argue this is a precarious position to be in because Amazon could always decide to make messaging a feature of AWS. Amazon already offers a push-notification service that competes with one of Twilio’s newest products.
But you could just as easily argue that Twilio’s value is simplicity and ease of use. For developers dealing with complexities of messaging lots of users at once, that may be worth the extra “tax” on top of what Amazon would charge. (At its place in the stack, Twilio will always be more expensive than AWS.) Certainly lots of large tech companies have decided that’s the case.
Twilio is a layer on top of Amazon Web Services in a way like The Wirecutter‘s product recommendations are a layer on top of Amazon.com. The value proposition is that it “just works.” And while Amazon offers a range of amazing services for developers, it’s not known for the user-friendliness of its APIs. Getting started with Twilio or Stripe is easier than getting started with AWS.
AWSaaS may not catch on as a name for this trend, but if Twilio continues to grow as a public company, it will be the best example of this new layer and new kind of startup.