API First SaaS: Twilio and Agora

In this week’s article, we dive into a specific vertical of software-as-a-service (SaaS) companies: API-first SaaS companies. And in this context, we focus on two companies: Twilio and Agora.

API-First SaaS Companies

OK – for the sake of our non-technical readers, we have to introduce a handful of jargon:

What are SaaS companies? Companies that sell software via the internet. These companies typically generate recurring revenue from continued use of their products. SaaS companies come in various sizes and flavors. We did an in-depth look at SaaS business characteristics.

What is an API? When a person uses a piece of software, he/she does so via its user interface (UI), clicking buttons to initiate different actions. But when software uses other software, it does so via APIs. In other words, APIs are standardized functions that allow programs to communicate with other programs, enabling more complex functions to be performed.
What are API-first SaaS companies? These are special types of SaaS companies that provide their APIs to other companies (Figure 1).

This means that:

• API-first companies are selling mostly to developers (since developers are the consumers of APIs) – using a bottoms up approach. API-first companies leverage simple API implementation and low cost of adoption, so that it’s very easy for engineers to adopt their APIs.
• API-first companies typically monetize based on usage. The more frequent the APIs are used, the more revenue these companies generate. Even though at the beginning these APIs are low-cost (or free), the cost scales with usage. As a result, revenue of API-first SaaS companies can scale rapidly – the more their customers grow, the more they consume the APIs, the more revenue the API companies generate.
• Their APIs are typically used to provide infrastructure services that are required, but are typically not core to one’s business. This means that it is often the case that it’s easier for most companies to use external services provided by these API providers than to build their own solutions.

Here are a few examples: 

• Twilio provides communication services. When Uber wants to let you know that your car is arriving, it uses Twilio’s SMS service to send you a message.
• Stripe provides payments. When a business wants to accept credit card payment over the internet – it can do so easily using Stripe’s APIs.
• Okta provides identity management services. When an enterprise needs to manage its employees’ access to any application or device, they can do so easily with Okta’s APIs.

Communication APIs

For this article we focus on communications APIs and compare two companies that serve this vertical: Twilio and Agora.

Twilio

What does the company do?

In October 2014, Facebook bought WhatsApp for more than US$21.8 billion. At the time of the acquisition, WhatsApp only generated ~US$10 million in revenue, but had more than 450 million users worldwide. WhatsApp was able to grow rapidly because it allowed people to communicate internationally via text for free, a significant departure from the costly SMS that was prevalent at the time.

But we are here not to talk about WhatsApp. The company that powered WhatsApp is Twilio. The more than US$21 billion of value that WhatsApp was able to capture (with only 50 or so employees at the time) was only possible because Twilio provides simple-to-use APIs that enables WhatsApp to send and receive SMS, through the internet, at a global level. 

Twilio’s service removes the significant complexities (plugging into different telco carriers worldwide) that are required to provide a global telephone service. Thanks to Twilio, anyone can easily adopt a programmable messaging system.

Although the company started off with SMS and Voice, Twilio has expanded to other services (video, email, IoT) through product expansion and acquisitions (it acquired SendGrid, an email API provider, and Segment, an analytics platform).

Twillio’s service expansion continues to enable others (WhatsApp, Lyft, Uber, Airbnb, etc) to not only send programmatic messages (SMS, OTP, transactional messages), make phone calls, and send timely emails, but also have a holistic view of the customer engagement. Twilio now counts more than 190,000 businesses as its customers.

How has the company increased in value?

As it is always the case, the business that owns the customer relationship captures most of the value. Around the time of the WhatsApp acquisition, Twilio was valued at only US$1 billion in the private market. In 2016, two years after the WhatsApp acquisition, Twilio went public at around US$2.4 billion. Five years later, as of this writing, the company’s market cap has increased to US$66.6 billion.

The increase in Twilio’s market cap is driven by two key factors:

• Macroeconomic tailwinds – the explosion of digital businesses that require programmatic communication.
• Expansion of Twilio’s platform.

The explosion of digital businesses that require programmatic communication has helped Twilio expand its total addressable market (TAM) rapidly. Nowadays, customers expect to:

• Receive SMS notification if their rideshare car is arriving.
• Be able to send text and make video calls via their favorite communication app for free.
• Receive OTP notifications via SMS when logging into their banking app.
• Get an automatic call back when calling customer service agents.

This macro tailwind has helped Twilio expand its revenue stream into multiple sectors and service types. When Twillio went public, the company had customer concentration risk. It relied heavily on the revenue it generated from WhatsApp (about 15% of revenue at the time – or approximately US$35 million per year). Thanks to significant TAM expansion, currently, no customer contributes more than 10% of total revenue. See Figure 4 for Twilio’s quarterly revenue trend. The company made US$590 million in Q1 2021 alone and has been able to sustain 60% year-over-year revenue growth rate.

One benefit of an API-first company is that, when your customer is successful, your revenue goes up as well (variable revenue). This means that without acquiring new users, revenue per user expands. We can measure this by the dollar-based net expansion metric. Twilio has one of the best in the sector, consistently exceeding 130% (this means if a customer spends $1 today, that same customer will spend $1.30 the following year). The higher the dollar-based net expansion, the more efficient the business can be. See Figure 5.

Variable revenue that scales with usage is great – but unfortunately, costs are also variable. The company has to pay telco providers based on the volume of phone calls and text messages. This means that the margins for the business are lower than a typical SaaS company. Twilio’s gross margin from the last twelve month is 51%, while a more traditional SaaS company such as Asana has a gross margin of 88%.

The result of lower gross margins, combined with the fact that stitching together a global telecommunication service is operationally expensive and requires high R&D spend, is that Twilio is still not profitable. In fact, despite the fast revenue growth, the company continues to lose money.

This is why the company has been moving up the stack, from a programmable communications protocol, to a customer engagement platform (see Figure 7). It is trying to do so by building value-added software (contact center, marketing campaign tools, etc.) that have higher margins. Twilio is doing this through new product development, partnerships, and acquisitions.

Make no mistake, the complexity of what Twilio is building makes the business hard to replicate by others. As large as WhatsApp has become, it is estimated that Facebook is still paying Twilio north of US$100 million per year for its services. It’s cheaper for Facebook to use Twilio than to create its own internal solution. It’s not easy to build and maintain a global communication network.

Yes – Twilio is not profitable, but it is growing rapidly. Its revenue growth comes from the continued increase in total addressable market, and for investors, that is sufficient for now.

Agora

What does the company do?

Agora is another API-first SaaS company that specializes in the communications vertical. Unlike Twilio, which started with SMS and audio calling before expanding to other services, Agora started with real time audio and video broadcast APIs.

Agora was started by Tony Zhao. For Zhao, audio and video conferencing was in his DNA. He was a founding engineer at WebEx (a video conferencing tool that was purchased by Cisco). In 2008, he left Cisco and joined YY, one of the first live-streaming platforms in China. Zhao realized that the smartphone revolution meant that the need for realtime audio and video streaming was no longer limited to business conference rooms. With a smartphone in every pocket, everyone has easy access to camera and microphone. This means that real time audio and video streaming can expand to social, commerce, and other verticals as well.

In 2014, Zhao left YY and founded Agora, a platform that enables other businesses to create audio and video broadcasts easily. In an interview with GGV, Zhao elaborated on what he saw was a gap in the market (emphasis ours).

“Going briefly into the technical side of things, the main problem we were solving is a quality issue. The issue that comes with public internet is its instability, where the internet itself is the “best-effort network”. A best-effort network refers to a network service that attempts to deliver data packets to their destinations. But there’s no guarantee to transmit the packet on time to the receiver. In simple terms, this can result in loss of data or corrupted data.

To solve this, we created a software-defined real-time network. This ensures a certain quality of service, close to what you would get with a dedicated network. We also developed deep technology innovation around signal processing algorithms, so that in challenging environments like low-bandwidth networks and lower-powered devices, we can still make it work.”

Early on, the company did not have a business model. It developed a developer-friendly API service and gave 10,000 minutes free every month for engineers to create new use cases.

This gave birth to the pricing scheme that is still adopted today:

If you look at the pricing plan (Figure 8), you can see that Agora gives a generous number of free minutes. This is done because the company wants to optimize for a different use case than Twilio (who also provides video and audio call services). Agora is focused on one-to-many real-time audio and video broadcast services – this use case maximizes its revenue growth: 

• If an app provides a 1:1 voice call, it’ll take 10,000 minutes before the app gets charged by Agora.
• In contrast, if an app provides a broadcast (one-to-many), then it’ll take 1 minute of one broadcast to 10,000 people for the free minute to be used up.
• After the free tier is used up, the company charges US$0.99 per 1,000 minutes for vVoice and US$3.99 per 1,000 minutes for 720P video. The fees can go up quite quickly!

How has the company increased its value?

Here’s the share price history of Agora (Figure 9). For the most part, the share price has been flat over the past year, with one exception.

In February 2021, a reddit user posted that Clubhouse, the extremely hot audio broadcast app, was built in a week, using Agora’s APIs. Retail investors jumped in, leading to a massive share price increase. But it seems that the increase in valuation was short lived. Agora’s share price has declined more than 61% from its peak in February 2021, as Clubhouse itself appears to be losing momentum.

Clubhouse may not do to Agora what WhatsApp has done to Twilio.

Nevertheless, there are several similarities between Twilio’s and Agora’s businesses.

First, both enjoy fast revenue growth. If the businesses that use their APIs grow rapidly, so will the revenue of Agora and Twilio. Agora’s quarterly revenue trend can be seen in Figure 10 below.

Here are some key takeaways:

• Revenue spiked in Q1 2020 due to the lockdown in China. The lockdown led to increased consumption of ecommerce and social media. Currently, the bulk of Agora’s revenue (more than 80%) comes from China (where live streaming broadcast is extremely popular for social media and ecommerce purposes). Agora powers YY, Xiaomi, Daoyu and a plethora of other apps in China.
• If you take into account the one-time revenue spike in China (light blue bar), then year-over-year growth in revenue is only 13%.
• But if you exclude the revenue spike from China (light blue bar), then the year-over-year revenue increase is 80%.

Second, similar to Twilio, Agora enjoys a strong net dollar expansion rate (Figure 10).

Third, as is with Twilio, Agora has variable revenue, but also variable costs. Agora has to expend variable costs in providing its services. As a result, its gross margin is ~62% from the past twelve months. Respectable, but not as high as your typical SaaS company.

What is the next growth opportunity for Agora?

One thing that Twilio has done but Agora is still searching for is a larger TAM. For Agora to achieve its next stage of growth, it has to achieve expansion of use cases, similar to what Twilio has done.

The current growth story of continued expansion in China (via the ecommerce and live streaming verticals) has already been baked into its current valuation. For Agora to grow beyond its current US$6 billion market cap, it needs to expand its use case into other verticals.

Here are some things to consider:

• One vertical that Agora is aggressively pursuing is the remote education market (where the main use case is one-to-many video broadcast). The company recently acquired Easemob and Netless to add chat and whiteboard APIs to its product suite. 
• Agora’s education vertical pursuit needs to expand beyond China. The growth of the online education sector in China might be slowing down as the government is cracking down on that industry.
• Just because Clubhouse is starting to lose momentum, does not mean the real time broadcast social media space won’t take off. The large US tech giants have recently announced their own version of Clubhouse: Spotify announced Greenroom, Facebook announced Live Audio Rooms, and Twitter announced Spaces. It is unclear if these services use Agora’s API, however.
• In addition to education and broadcast services, Agora is also pursuing remote healthcare, gaming, VR, and other enterprise services.

The Final Word

It’s still an open question on whether Agora can achieve similar growth as Twilio has enjoyed. A large part of Agora’s value increase will be determined by the company’s ability in pursuing a larger TAM. Nevertheless, the two companies (surprisingly) enjoy similar valuations. Twilio is currently valued at 23x EV/Sales while Agora is valued at 20x.