Investors, founders report hot market for API startups

Startups that deliver their service via an API are having a moment. Or perhaps a year.

Speaking with founders and investors this year, it has become clear that the API model of delivering a product is more than an occasional hit-maker for companies like Twilio or Plaid. Instead, it appears that there is ample room for lots of API-powered startups to build and prosper.

TechCrunch took note of a cluster of funding rounds for API-powered startups earlier this year, only to see more of the same as startups like Alpaca (equities trading via an API) reported massive growth and Noyo (APIs that link players in the health insurance market) raised new capital.

There’s more to come. Twilio’s Jeff Lawson told TechCrunch recently that “the world is getting broken down into APIs” as “every part of the stack of business that a developer might need to build is eventually turning into APIs that developers can use.”

We should expect to see more startups, then, pursuing the business model as time passes.

To dig deep into the API-focused startup space, we’ve done something unusual today. Instead of merely ringing a bunch of VCs to get their take — though we did that as well — we took the time for this survey to also bring a number of entrepreneurs into the conversation.

With two sets of questions targeted at each group, here’s who we corresponded with:

And they had a lot to say.

Big themes

We’ll limit ourselves to two themes from investors and two from founders. But don’t worry, as we’ve embedded full responses down below.

Starting with investors, our chief takeaway was that the money folks are bullish on not only the current generation of API-powered startups, but also on their future. We asked about the possible union between API-powered startups and low-code/no-code technologies. Our hunch was that as more folks can code in some manner, and APIs get better, there comes a day when nontraditional developers can leverage application programming interfaces.

That day, if it comes, could provide a huge boost to the startups in the space, right? Root VC’s Edward seems to think so. He answered our question about the possibility of nontraditional developers interacting with APIs in the future with an enthusiastic yes, adding that he believes that “eventually almost everyone will be a programmer, but that our definition of programmer will expand to fit a much broader range of activities.” That could mean lots more folks out there ingesting, using and paying for access to APIs that startups will be there to offer.

Even more, Edwards added that the same forces work in reverse, that “API-driven businesses enable low-code implementations and give superpowers to junior developers or people who don’t consider themselves developers at all.”

Shasta’s Roth agreed, saying that “these are highly related segments: low code and APIs.”

Our second investor takeaway is that it’s too simplistic to merely say that API-focused startups are going to be akin to SaaS startups in many ways, albeit with lower gross margins. They are not worse businesses than SaaS startups. Instead, they are different. Roth noted, for example, that API-delivered startups should have strong gross retention (logo retention), but that they may not have strong upselling power (net retention). Adding to the nuance of the conversation around economics, the Accel duo said that while API-powered startups may have “endemically lower” gross margins than SaaS startups, they also often feature “lower spend on sales and marketing and stronger net retention, both via lower churn and faster, bigger expansion.”

So, the net retention point is probably not fully settled yet, but what is clear is that our previous view of API startup economics is probably a bit simplistic.

From our trio of founders, two quick things. First, the venture capital community is as active as you’d expect, especially when it comes to preemption. Second, their startups tend to have improving economic profiles over time. The question for them then becomes how far they can run the gross margin numbers up before they go public.

We’ll see. You’ll find full answers below, lightly edited for clarity:


Isaac Roth, Shasta Ventures

Are API-delivered startups a plank in your firm’s general investment thesis? If so, why? 

Yes, very much. But making APIs an investing pillar is like making SaaS an investing pillar — it’s too broad. Rather, we have integrated the understanding that there is a shift to composable capabilities and that it is no longer the domain of custom expensive integrators to hook these capabilities together. The secret sauce is what industries will this affect in which ways, what are the opportunities that arise as a result, which types of APIs will be adopted first and last, of course, where does value lie?

We also see increasing use of APIs by enterprises leading to startups creating solutions for enterprises to manage, monitor and secure APIs and home-grown applications created using those APIs. 

Are you seeing most API-delivered startups in the market for capital today find new places to apply APIs, or are you seeing the majority of startups pursuing the model working inside of market areas known to be API-friendly? What market segment is the ripest for API-delivered startup disruption?

The unbundling of financial services, which makes way for innovation and personalized experiences is a great opportunity. That one is easier to realize. An underappreciated opportunity is in HR and corporate finance where monolithic applications integrate many functionalities that could benefit from evolving separately and could be knit together by each enterprise in a manner that is oriented toward their unique needs.

Security is another industry where every solution seems to have its own stovepipe interface and yet most CIOs and CISOs want integrated panes of glass. There will be low-code solutions for aggregation security information and response.   Additionally, we predict a significant increase in the use of APIs within enterprises and CISOs to look for solutions to manage access and threats emerging from these APIs.

Finally, think about commerce — a segment that has already benefited from the API economy — it was the original poster child for APIs and is finally catching up to that promise. However, because the nature of commerce is being accelerated due to COVID there is a lot of room left here.

Is the economic profile of API-delivered startups, especially from a gross-margin perspective, still on track to land one level below that of SaaS startups?

Until APIs have proprietary value by aggregating data (see my article about this in Programmable Web) the switching cost is lower than SaaS because there isn’t as much stickiness from needing to retrain a workforce if you switch. This means customers have more pricing power. Similarly, APIs enable competition because they define a standard interaction, and this causes lower margins. But keep reading for how to overcome this.

Does strong retention rates amongst API-delivered startups countermand their more limited gross margin profile?

Related to the above, a well-performing API will retain customers but it may not have as strong net retention as SaaS unless the API business can aggregate more value either beneath the API (more functionality) or around the API (management, integration, workflow, compliance, risk management, etc.).

It appears that rounds for startups that deliver their product or service via an API are heating up. How competitive are their current venture deals? What has happened to valuations for these startups? Have their revenue multiples reached SaaS levels?

Investors who haven’t been developers themselves are getting starry-eyed about the large numbers. They see millions of API calls and tens of thousands of users and figure, “This thing is hot!” I’m cautious and excited. Excited because of the power that these APIs give to developers to create entirely new experiences. The engineer, the dreamer in me, loves these things and I enjoy trying them out! I really want access to the GPT-3 research API and every week I have an OMG moment finding some new API and imagining or prototyping what can be done. The reality is the world is not ready for much of the automation that is possible and by the time behaviors and expectations catch up there will be a lot of competition, there will be aggregation, there will be platform inertia. It’s as hard for startups to break through using an API as it is with any other kind of innovation.

One framing I have is that APIs are a lot like open source. People thought open source was a product category. Not discounting the values movement that IMO changed how we collaborate, for technology businesses it turned out to be a GTM strategy. APIs look a lot like that to me. They’re changing how we think about collaborating, especially among businesses that can supply unique parts of a value chain. But for technology providers, they look like a GTM strategy to me. 

Perhaps stretching the analogy, but we need to find out if Twilio isn’t the Red Hat of APIs where they managed to aggregate enough functionality to be a great business, created a wave of expectation and no other business could ever fulfill that expectation. For years many open-source businesses came and went and none were able to attain the recurring revenue scale of Red Hat. It was only when open source became a go-to-market strategy for platform SaaS (many of them API businesses) that we saw revenue scale. Most people describe this the other way: An open-source project gains success and then a business is wrapped around it by hosting it in the cloud. Which way you tell the story doesn’t change the business dynamic.

Consider this further mind twist on the API as a business model instead of go-to market. Imagine you take something like a CRM system or a financial management system and it has all the screens and workflows in a nice SaaS interface with mobile apps and all and you give that away entirely for free. Then you charge only for API access that allows the customer to integrate that captured data with other systems. Is this an API company? Or is it SaaS? 

Does the burgeoning market of no- and low-code tools point to a future where APIs can be ingested, linked and otherwise used by nontraditional developers? If so, how far away is that future?

Yes, these are highly related segments: low code and APIs. The dream and vision here is that a nontrained developer can create automated workflows in a low-code platform by gluing together APIs. Automation of the long tail of business processes, including advanced functionality like ML. The reality is today we haven’t seen this. Some of the low-code solutions are really faux-low-code and some aren’t actually low code at all; they are just using a new nonstandard programming framework but require a large amount of development. There is a sweet spot to be realized where the low-code framework has enough built-in logic and integration that is suited to particular use cases that it is truly powerful while remaining worthwhile to use rather than just paying someone on Fiverr to knit together the required APIs in Python.

Related to the discussion of net retention and margins, a logical place for API businesses to grow is having a low-code type environment behind their API such that once an API integration is performed, the nature of what happens with the data being exchanged can be configured without redoing the API. For illustration by example imagine a simple API that sends money. Now imagine you want to program rules into that API such as any time you’re sending money to Germany, settle it in euros, unless it’s a customer who has requested to be settled in GBP. Those settings can be exposed in the API but it is likely to be easier to start configuring such business rules in the API vendor’s console — in particular, because one doesn’t need a programmer to set up these rules. 

Now imagine the money-sending API has a low-code interface allowing you to inject some logic into determining what currency to settle in, and includes nice pre-built integrations with algorithms that more sophisticated customers use. Now you’ve created more value for the customer by aggregating data on top of functionality and stickiness by capturing logic. That kind of thing sounds like a successful business with all the buzz: developer adoption, low code, APIs and SaaS margins. It could even have a “disruptive” business model by giving away the core API for free and only charging for the added logic capability. This is definitely the future and one I’m actively investing in.

Finally, how many net-new investments do you anticipate making into API-delivered startups in 2021?

We make 6-8 new investments per year. All of them will likely have an API angle. We’ve already made one unannounced this year that I’d characterize as API-forward, and I’m guessing we’d make about half next year in that same vein.

Daniel Levine and Vasant Natarajan, Accel

Are API-delivered startups a plank in your firm’s general investment thesis? If so, why? 

They’re a major focus for us. We’re huge believers in the software economy and tools and infrastructure derivatives that power it. APIs are key abstractions that help operationalize important parts of a software stack. Developers can build better and more reliable products when they leverage independently scalable, fully focused APIs. Where would Uber be without things like Twilio for telephony, Google Maps or Braintree for early mobile payments.

Are you seeing most API-delivered startups in the market for capital today find new places to apply APIs, or are you seeing the majority of startups pursuing the model working inside of market areas known to be API-friendly? What market segment is the ripest for API-delivered startup disruption?

The API-first startup space is still very nascent. The overwhelming majority of companies we see in the space are working in newer areas for APIs.

Imagine a diagram with a number of concentric circles on it. The very inside circle is the developer. In the second circle are traditional developer services and some DevOps-oriented like version control, CI/CD, application server hosting, databases, etc. There’s still more to do in that circle, but it’s less greenfield. In the third circle are specialty areas of engineering or business processes that abut the application developer’s core tools. That third circle is where products such as Stripe, Twilio, Scale, Mux, Checkr, Shippo, etc. sit. There is a ton to do in that third circle and the needs are fast-moving. It’s not obvious that one needs an API for background checks until there is a boom of companies hiring tens of thousands of people quickly and building their own software to do it. All of a sudden, those internal developers dramatically prefer hitting an API to emailing someone spreadsheets back and forth. Eventually, Checkr is born and it leverages those early adopters to build a better product at a better value than the incumbents. A lot of Stripe’s advantage early on was abstracting away painful business processes like signing up for a merchant account, or trying to model out one’s mix of credit card types.

And the circles are dynamic, like a black hole. Over time more and more things get sucked closer to the application developer. Those become APIs and all of sudden the areas adjacent to that problem enter the third circle.

Is the economic profile of API-delivered startups, especially from a gross margin perspective, still on track to land one level below that of SaaS startups? Does strong retention rates amongst API-delivered startups countermand their more limited gross-margin profile?

Gross margin might be endemically lower than SaaS companies. But we often see lower spend on sales and marketing and stronger net retention, both via lower churn and faster, bigger expansion. We wouldn’t say that the economic profile is “one level below” just that it’s different.

The net retention rates in these businesses are somewhat latent. In most SaaS companies you have to go out and earn your renewal and upsell. API companies are priced to grow naturally with the apps using them. Retention is much more baked in. Moreover, they become mini-indices of the industries. Think about Checkr, as a common background-checking API to many of the world’s largest and fastest-growing gig-economy companies. If you wanted to own a basket of all the world’s on-demand labor marketplaces, you’d want to own Checkr!

It appears that rounds for startups that deliver their product or service via an API are heating up. How competitive are their current venture deals? What has happened to valuations for these startups? Have their revenue multiples reached SaaS levels?

All venture deals are very competitive, for good reason. So much of the global economy is rotating toward a software-led approach — markets are pricing in that potential. Specific to API-first companies, we’re seeing more venture firms in the mix than before. With companies like Twilio, Stripe, Segment and others building large businesses with hundreds of millions of revenue, the case studies are there to anchor early-stage investors.

Does the burgeoning market of no- and low-code tools point to a future where APIs can be ingested, linked and otherwise used by nontraditional developers? If so, how far away is that future?

We don’t think so, but that’s been happening for a while, well before the no- and low-code tools. There are many advantages that APIs have and one is that application developers can include it in their own application, effectively becoming a reseller. That can create some customer concentration and put pressure on margins, but it can be an incredible strategy for growth but also accessing less-technical markets. Some examples include:

  • Stripe powers payments for a number of e-commerce platforms, including Shopify. 
  • Checkr works with Workday, LinkedIn, Greenhouse, etc. to seamlessly provide background checks for their platforms.

Finally, how many net-new investments do you anticipate making into API-delivered startups in 2021?

We’re only constrained by the number of amazing teams and large market opportunities we intersect. But generally believe API-first opportunities are hidden in just about every major SaaS category, if teams are willing to seek them out! We’d love to talk to them. If we had to guess probably ~2-3 (which is a lot of any space in a year) across the team.

Lee Edwards, Root Ventures

Are API-delivered startups a plank in your firm’s general investment thesis? If so, why?

Yes, definitely. I talk about this here and in shorter quotable bites here.

There is an [approximately] three million global developer deficit relative to industry demand. Even during a recession, and even after layoffs, companies will pay for developer tools that amplify productivity. Honestly, they don’t have a choice, because developers will use whatever they want anyway.

Low code is infinitely more interesting than no code. The most popular programming language in the world is Excel. We are massively underestimating the number of people who can low-code in their area of expertise. The barriers to entry for coding are lowering every day.

API-driven businesses enable low-code implementations, and give superpowers to junior developers or people who don’t consider themselves developers at all.

I think API-driven businesses take advantage of these trends. Many larger companies take a buy-over-build approach. The top handful of e-commerce companies build their own payment gateways, for example. But for everyone else, there’s Stripe, which has unlocked an enormous amount of value in the tech ecosystem and managed to capture a percentage of it.

Are you seeing most API-delivered startups in the market for capital today find new places to apply APIs, or are you seeing the majority of startups pursuing the model working inside of market areas known to be API-friendly?

As a firm, we are probably seeing more API-driven companies in the space of serving a technical audience, but that may be a function of our firm’s brand and thesis.

What market segment is the ripest for API-delivered startup disruption?

There are huge swaths of overlooked and underserved industries. Flexport showed the Valley that billion-dollar businesses exist right under our noses that simply haven’t had access to the digitalization that other industries have had. Satya Nadella recently said that they’ve seen years of digitalization happen in months, because of COVID-19. So we keep an eye on a lot of the industries where we have direct experience as a firm and in our network — shipping, logistics, manufacturing, food, agriculture, transportation, mining, oil and gas, alternative energy — anything worth trillions of dollars that’s still running largely on paper and desktop computers. They’re hungry for innovation and the value is there.

Is the economic profile of API-delivered startups, especially from a gross-margin perspective, still on track to land one level below that of SaaS startups?

That may be true. Many API-driven businesses are very infrastructure heavy. In the early stages, we tend to not think a ton about the cost of software infrastructure and focus more on building a product that people love and want to pay for. Later stages, and certainly as a company approaches a public filing, this begins to matter a lot more. I think we’ve seen that many AI/ML-driven businesses bring home less than traditional software margins <someone tweeted or blogged about this, I forgot who>, but I think API-driven businesses are often somewhere in between. On the other hand, sometimes the thing on the other side of the API is very expensive. Planet Labs owns hardware in actual outer space and purchases rocket payload space, for example.

Does strong retention rates amongst API-delivered startups countermand their more limited gross-margin profile?

Absolutely. I think one reckoning we are seeing in some SaaS and subscription-model companies is in churn and net retention. Investors are focusing more and more on those metrics earlier and earlier. API businesses often serve something that is critical to the business and expensive to replace. Yes, lock-in is strong, but maybe more importantly, a company focused on a specific offering is probably going to do it better than your internal team of 2-5 engineers. We see this over and over when Daily.co pitches its API for video chat to companies building tools with real-time video communication components.

It appears that rounds for startups that deliver their product or service via an API are heating up. How competitive are their current venture deals?

Extremely. It’s the most competitive segment we invest in.

What has happened to valuations for these startups? Have their revenue multiples reached SaaS levels?

In some cases they have exceeded them. Some factors:

  • Frothy IPO and high-value acquisitions (FAANG+MSFT) in 2018-2020.
  • Focus on churn and retention — API-driven businesses do well here for reasons stated above.
  • Early-stage venture valuations are not very tightly coupled to revenue multiples and often reflect “the highest price someone is willing to pay” — competition between venture firms drives this, as does the entry of deep-pocketed multistage firms who know the best way to get ownership in a competitive vertical is to invest earlier and earlier. They’re less price sensitive, because they plan to buy up with their winners over multiple rounds. The downside is, they’re more aggressive with cutting off the ones they perceive as not winners — and this is a huge negative signal that can kill a Series A financing.

Does the burgeoning market of no- and low-code tools point to a future where APIs can be ingested, linked and otherwise used by nontraditional developers?

Yes!

If so, how far away is that future?

It’s close. The hardest part of solving a problem is understanding the domain. We already see domain experts hacking and editing code, writing scripts and macros and formulas in Excel. Programming is just the language of problem solving — anyone who knows how to solve a problem methodically can learn it, and the tools and resources for doing it get easier and easier by the day. I do believe that eventually almost everyone will be a programmer, but that our definition of programmer will expand to fit a much broader range of activities.

Finally, how many net-new investments do you anticipate making into API-delivered startups in 2021?

2-3

Shannon Goggin, Noyo

How quickly is the market you sell into becoming more amenable to ingesting your service, or related services, via an API via a hosted software service (SaaS)?

Noyo works with health insurance companies and benefits software to trade sensitive customer data. APIs are a hot topic across the industry right now, and many insurance companies are actively working to map out their own API strategies. Most recognize that better data connectivity is needed and are used to hearing about APIs as an abstract concept, but there is a big need for education. While the conversation about hosted versus cloud-based solutions does come up, it has not been a blocker; in fact, APIs enhance security and compliance in many ways (e.g., granular data access controls).

Are investors trying to preempt your next round? 

It’s a huge opportunity to apply great technology to an important space; there is a lot of engagement.

Is the amount of lift, or help that you provide customers to get them up and running with your service, going down over time?

The great thing about this technology is it scales very well. We are deliberate about engaging closely with our clients throughout the onboarding experience — having a deep and nuanced understanding of how they are using the technology lets us build the most flexible and powerful solutions.

How focused are you on boosting the vertical depth of your service, compared to expanding it horizontally? How has your planning on this point changed in the last year?

We are focused on the health insurance vertical, but rapidly expanding the breadth of our product in the space. We started with a specific use case to launch and have quickly expanded into adjacent use cases. We’ll keep doing this until we can power every transaction within the health insurance space.

Are your gross margins improving over time?

Yes. This is a business with meaningful economies of scale.

Anshu Sharma, Skyflow

How quickly is the market you sell into becoming more amenable to ingesting your service, or related services, via an API via a hosted software service (SaaS)?

I am surprised at how open companies are to a SaaS API for something as critical as cybersecurity. While I have spent over a decade in SaaS including some very large deals during my time at Salesforce, the scope of the projects by large companies including banks and healthcare companies is simply beyond what was a possibility just a few years ago. We have truly moved from “why SaaS” to a “why not SaaS” era.

Are investors trying to preempt your next round? 

We get emails every week from investors. The interest in both APIs and privacy markets has never been higher.

Is the amount of lift, or help that you provide customers to get them up and running with your service, going down over time?

Our largest deployment to date took less than two hours to get up and running. That’s the power of a simple API that solves a complex problem.

How focused are you on boosting the vertical depth of your service, compared to expanding it horizontally? How has your planning on this point changed in the last year?

We have been focused on fintech and health tech companies as our leading verticals. While every company in the world that has customer data (B2C) needs a privacy data vault, the need for financial services and healthcare is the highest — one of them is literally getting fined or breached every day.

Are your gross margins improving over time?

Like Jeff Bezos, we don’t obsess over margins. We intend to focus on free cash flow to build Skyflow into a public company. (Our margins are well north of most SaaS companies.)

Is your net retention improving over time?

We are early to answer this question but customers are baking us into their core services and we expect positive net churn as usage grows faster than churn.

Do investors have the same growth expectations for your business as they do their SaaS portfolio companies?

I don’t ask investors what they expect. We focus on building our own business with its own unique characteristics. I believe fitting into someone else’s model would have killed businesses like AWS, Snowflake, Salesforce, etc. We will find our own model of growth and success.

Nina Kuruvilla, Daily.co

How quickly is the market you sell into becoming more amenable to ingesting your service, or related services, via an API via a hosted software service (SaaS)?

Very quickly. Live video requires both a lot of client-side code and significant infrastructure, so it’s relatively easy for us to convince developers and product managers that they should “buy” (via an API) rather than “build” (from scratch). That’s gotten even easier in 2020 as more products and more teams are thinking about video as a core feature.

Are investors trying to preempt your next round?

Yes.

Is the amount of lift, or help that you provide customers to get them up and running with your service, going down over time?

No. But we think this is because of the increase in the number and kinds of developers who are integrating live video into their websites and applications. 2020 has been a year of explosive growth for video, so many of our new customers are less experienced with video than our average customer was in 2019. People are also operating on much more aggressive timelines for rolling out video.

How focused are you on boosting the vertical depth of your service, compared to expanding it horizontally?

Vertical depth is a key element of our strategy. We describe ourselves as a “full-stack” solution for live video. We have core video call features, plus a prebuilt UI that lets you get up and running with just a few lines of code, and an extensive list of add-on features like recording, live streaming out to YouTube Live and Facebook Live, and detailed call performance data. Additional “vertical” features on our roadmap include transcription, more extensive text chat features and use-case specific plugins that make it even easier for our customers to roll out specific types of video experiences (for example, webinars, social games, healthcare provider queues, and e-commerce platform integrations).

How has your planning on this point changed in the last year?

Our planning hasn’t changed. We have always believed that to win in our market requires execution on both horizontal (use cases and sectors) and vertical (a full complement of features that are both “low level” and “high level”). That’s a challenge, but it’s also an interesting set of problems to work on that span both business and technical architecture!

Are your gross margins improving over time?

Yes.

Is your net retention improving over time?

Yes.

Do investors have the same growth expectations for your business as they do their SaaS portfolio companies?

Yes.