28 comments

[ 2.4 ms ] story [ 67.7 ms ] thread
Weren't APIs the next big thing a decade or so back?
I think that was the web 2.0 "mashup" era while the current trend seems more focused on being "headless".
Yeah, there was definitely a trend a decade or so back when everyone was hyped for the fact they were releasing API access.

Generally speaking, anything that took off slowly let those APIs die so they could make money for themselves instead of letting third parties capture value.

Yes - it feels like the wave has passed, or rather is in full effect.

Twilio, Stripe, Lob are all relatively large companies now.

Yes, that was even part of the big marketing wave from .NET, SOAP, WSDL repository servers, BPEL and BizTalk.
> Graphics courtesy of Accel

> Daniel Levine is a partner at Accel.

> Accel’s first API investment was in Braintree

Advertorial

Every major SaaS already has an API available for their product as far as I can tell. What sort of nontrivial service is just API endpoints? e.g. if your product is just an API to get weather info or a headless CMS, isn't that just a feature of a more-complete product?

I'm genuinely curious, because it seems like API-based SaaS would be good products for engineers to launch.

Stripe is an example of it. They do have a dashboard, but at the end of the day, without implementing their AI or stripe.js, there is not much you can do with them.

It seems like a great business model for engineers to try out ideas.

Schema.io eCommerce API. Pretty cool actually if you look in to the details.
Something like unchained.shop. Hybris is uncool and too expensive to customize. I agree it's a good thing for engineers to launch.
Next up in the news from 2000: "Data warehouses on the rise due to increase in data required for machine learning". /sarcasm off

Seriously: APIs were always a thing and always will be as long as different programs have to talk to each other. I just recall the good old "SOA" days with loads of XML in SOAP and WSDL/UDDI [0], when everyone told the tale of solving the interoperability problems once and for all nicely captured by [1]. I agree, how business was handled was different (you had a contract, and wouldn't order with one click and your credit card), however this doesn't mean it is fundamentally new. We will see an increasing shift towards "pay as you go" services you can consume but i wonder where the limits of this trend are as you get highly dependent on thrid party providers to not shut down or raise prices over night, e.g. [2]

[0] https://en.wikipedia.org/wiki/Web_Services_Discovery

[1] https://xkcd.com/927/

[2] https://www.quora.com/What-is-Google-thinking-with-its-new-M...

Twilio/Sendgrid/Mulesoft hopes this becomes the case.

People also need to realize that you could make the argument that the massive behemoth that is AWS is just an API company. They give you simple APIs to perform actions (in their case that might be to spin up a compute node, etc).

In the Sendgrid example, they give you an API to send an email.

So it’s about companies who create services and the customer touch Point is only an API

At the end of the day, all of AWS is simply "I wish to mutate the state in your cloud such that your cloud will subsequently respond to request X with response f(X)." At the end of the day, that's an API. It's just that you're also warming a "cache" that happens to be Turing-complete :)
I can not imagine I use a SaaS that does not provides APIs.
This read like an ad.

I don't feel like APIs are the next big thing. They were the next big thing about 10 years ago and now there's just part of everything.

I think you need to understand the audience for this post.

Dan is trying to demonstrate that there is room for many more large VC style returns by investing in companies that's primary value is being an API.

His argument is that the market is bigger than people think it is today, and we've just seen the first ripple.

This post is to help him establish himself and Accel as someone to go talk to if you are building a API based company, because he's a believer you could be a unicorn.

OR if you are a LP, and want big returns on this 'new' wave come talk to Dan at Accel.

I do agree it's an Ad.

FWIW I like the graphics :) and it's a good argument.

(comment deleted)
>> Dan is trying to demonstrate that there is room for many more large VC style returns by investing in companies that's primary value is being an API.

I am sure you are trying to help them, but some more details would be helpful.

Any 12 year old with a phone/tablet/computer can create an "API". Is that what VCs are looking for? I imagine that there has to be more.

With a title like that, I was expecting to read about monetizing APIs using platforms such as RapidAPI or Apigee.
Right, except we were talking about moving everything to microservices a decade ago, and many things started moving in that direction. Enough so that the pendulum has swung back with articles talking about the downsides of moving to MSA.

Did Accel recently discover these companies all have APIs?

Please don't turn "API" into the next "blockchain", where I have to fend off VCs asking me how I incorporate APIs into my product or if I'm currently developing any APIs. That would be like me asking them about spreadsheets.
VCs need more places to stuff their money.

Many companies aren't companies, they're features.

Rebrand these vulnerable "feature" companies as "API" companies and go bull on them.

How did I know this would be written by a venture capitalist before even clicking on it?
Disclosure: I work on https://treenotation.org.

I made a my early career out of working on these API businesses in the late 2000's, early 2010's. The big change was the shift from XML/SOAP/WSDL to JSON, which just made things far easier for developers. Lots of companies did APIs before that, but they were ahead of their time as we needed JSON, which made things a lot simpler.

Fast forward to 2019. I'm pretty bullish that Tree Notation can unlock the next wave of innovation in the API space. Here is a demo video I did of a very early version exploring this idea back in 2014 (https://www.youtube.com/watch?v=DV4Nv23bBwQ&t=1s).

If anyone is interested and will be at this conference (I'll be there), shoot me an email and can do some more demos.