Cloud communications platform Twilio (TWLO -1.44%) went public on June 23, raising $150 million and booking a first-day gain of over 90% from its IPO price of $15. The stock dipped slightly during the post-Brexit sell-off, but subsequently recovered and soared to the low $30s.

Twilio's platform enables developers to create apps that can interact with customers through messages, video, and voice. Major customers include Facebook's (META 0.27%) WhatsApp, Uber, and Airbnb. By connecting apps to mobile networks, Twilio's service lets WhatsApp users look up users through phone numbers, and enables Uber passengers to call or text their drivers. Thrivent Asset Management analyst David Rudow recently told the Wall Street Journal that Twilio "absolutely revolutionized" the way apps were built, since it would be tough and costly to build those services from scratch.

Image source: WhatsApp.

Twilio's growth indicates that excitement is justified. Revenue rose 78% to $88.8 million in 2014 and another 88% to $166.9 million in 2015. The company generates most of that revenue by charging customers per-minute rates for calls, per-message rates for text messages, and monthly rates for dedicated Twilio phone numbers. But that business model isn't profitable yet -- Twilio's net loss narrowed from $26.9 million to $26.8 million between 2013 and 2014, but widened considerably to $35.5 million in 2015.

High sales growth and low profitability are common in high-growth tech companies, but Twilio faces a much bigger risk that investors should be aware of -- its heavy dependence on Whatsapp.

How much does WhatsApp matter to Twilio?

In its S-1 filing, Twilio reveals that 17% of its 2015 revenue came from WhatsApp, up from 11% in 2013 and 13% in 2014. At first glance, that doesn't seem like a huge risk -- WhatsApp is the most popular messaging app in the world, and its active user base topped 1 billion users in April.

However, WhatsApp is also somewhat redundant in conjunction with Facebook's own Messenger app, which topped 900 million active users in April. While Messenger is gradually evolving into a monolithic app with mobile payments, in-app games, and chat bots, WhatsApp remains curiously separated from that ecosystem. 

Twilio's ability to match users based on their phone numbers is seemingly the backbone of WhatsApp, but the company warns that WhatsApp "has no obligation to provide any notice" if it decides to stop using the platform. WhatsApp's unclear plans for expansion also cast a shadow on Twilio -- it recently ditched its $1 annual fee (its only source of revenue), refuses to display ads within the app, and hasn't explained how it will monetize WhatsApp beyond vague plans to connect businesses to customers. But on the bright side, Twilio added API support for Facebook Messenger in April, indicating that the relationship between Facebook and Twilio remains strong. 

Understanding "no stack" companies

To understand why WhatsApp could eventually ditch Twilio, we should understand the difference between full-stack and no-stack companies. In the past, "full-stack" companies would create an app from scratch, creating their own communications features like the ones Twilio provides. That process was costly, time-consuming, and often resulted in compatibility issues.

This led to the rise of "no stack" start-ups, which develop a core technology and subscribe to other services. For example, Uber develops its core ride-hailing technology, but uses Twilio for text messages, PayPal's Braintree for billing, and Alphabet's Google Maps to find routes. Tapping into those "best in breed" services ensures that Uber provides a smooth experience for its customers, and lets it scale its operations in a cost-efficient way.

But Facebook isn't like Uber. It's a mainly full-stack player, with its own AI platform and data centers to host messages across its social network. Since Facebook has nearly a billion mobile daily active users worldwide, looking up users via Facebook profiles on WhatsApp could gradually reduce phone number-based searches. The recent addition of scanable profile codes to Messenger further eliminates the need for phone number exchanges. If WhatsApp introduces similar features, its dependence on Twilio's phone number service could dramatically decline.

Facebook Messenger codes. Image source: Facebook.

The bottom line

WhatsApp probably won't stop using Twilio anytime soon, since phone numbers and SMS messages will likely remain universal lookup and default communication tools for the foreseeable future. But over the long term, Twilio could shed revenue if Facebook uses its own social network as a search directory for WhatsApp and its own platform to deliver messages. If Facebook extends those services to other apps -- as it has already partially done with single sign-ons -- it could even evolve from a customer into a competitor.