While everyone was on the holiday break at the end of last year, Twilio wrote a blog post that very few people noticed. They announced that our firm, Union Square Ventures, had become an investor in Twilio.
Twilio is not a services for the masses. Yet.
It's a service that web developers can use to build telephony apps or build telephony into their app. This image on Twilio's home page says it all.
In the "Areas of Interest" post that I wrote at the start of the year, I wrote:
Developers are the new power users. If you cater to them, you can build a large user base with significant network effects.
We believe that one way to build a large network of web users is to build something that makes developers' lives easier. And Twilio does exactly that. It masks all the complexity of telephony into a finite number of API calls that web developers can use to build apps quickly and easily.
When we first met Twilio, the founder Jeff Lawson blew me away when he told me that the entire service was built on five API calls; <say>, <play>, <gather>, <record>, and <dial>. Clearly they've added a few more, like <conference> highlighted above.
Today, Twilio is announcing they've added SMS to the service, with <sms>. Now you can easily allow users to access your web app via SMS without having to set up shortcodes, dealing with aggregators, doing the configurations, etc. TechCrunch has the details.
My partner Albert led this investment for us. He's a web developer himself and has already saved himself countless hours with Twilio. Albert wrote a post about Twilio on the USV blog today. If you are a developer of web apps and are interested in adding telephony, you should check out Twilio.
<dial>Fred</dial><say>That’s a smart investment, congrats!</say>
This is a great pick. Telephony (both fixed and mobile) and the Internet are coming together into this uber-communication mashup. To pick up a stake in the bonds that will drive this connection is smart.
Wow, that’s great. Making it that easy will mean we will have it everywhere – can’t wait for that.Wonder what the revenue model looks like right now…
the revenue model is really straightforward because minutes on PSTN stillcost something. so twilio buys them and sells them via the service. samewith text messages.
Thanks – hadn’t seen that.I’ve had to learn a lot about the phone system recently, and wow is it messy, and not really clear where it’s going…how does that factor into your investment decision?
beautiful, nothing like simple business models…
I like it. Interested in seeing how this company grows.
I like this.What’s a phone anyway?Who needs the phone company?Telephony in its natural state should be part of many apps I use all the time. This seems a step in that direction.Congrats.
Just read the post on USV. Had no idea a service like this even existed. Really like how simple it sounds with just a handful of API calls.Also like how this service is tackling a very focused problem to start with. Your investments do a good job of starting on one part of the problem and building from there.
I have always wondered why web 2.0 companies cannot see the light of a developer program or community. Companies like Etsy have fallen behind the curve (with hopes of a comeback) with no real API or developer program. Today a market of 4 billion cell phones exists (we only have 1 billion computers) imagine how Twilio is positioned not just for what it can do today…but what the development community is going to want/need ……I’m going to keep a keen eye out for Twilio (and of course join the developer program!) YAY!!!!!!!!!!!!!!!!
We are pushing all of our companies, including etsy, to be more developer friendly
Etsy has been on a good trajectory….I think I am a bit hyper-critical of Etsy since they not only have created a marketplace….but have touched & enabled individuals around the world who are not computer savvy to see an opportunity which never existed…..Etsy allows me to “sell my passion to the world”. I have run and been part of lots of developer programs (Macromedia, Oracle, RespondTV) and it was that bit of magic that allowed the developers to go the extra mile when thinking about true change that they can influence. Fred & Beth, I am glad to hear that Etsy see’s the light and that Twilio has brought such a global opportunity to my finger tips (and your cell phone!).
At Etsy we are really focused on the API and our developer community for 2010. We expect to have some things related to that to anounce in the upcoming months.
my top 5 avc.com API:<think> <challenge> <communicate> <care> <lol>.Well, Ok:<9/11>
Fantastic. I missed this the first time the news came along, but the “<conference>My Room</conference>” notation grabbed me. Talk about an elevator pitch, twitter-style!
I’m glad you liked the ‘my room’ pitch. It works for me too$
I like the integration of voice and sms in one number that is very useful.There doesn’t seem to be an opportunity for revenue sharing at the moment, which is something which can be a big driver in the adoption of an sms service.
…or even the continued success of Twilio itself. You are speaking about revenue sharing with services built on top of Twilio, correct?
Can you elaborate on how that might work?
I’m not sure this is what @RichardForster meant, but I was going to suggest something like this: If I provide a service where my users send a text message to activate/use the service, they could pay a higher price for that message (through their phone bill), and Twilio and I could split the income
The ability to receive a revenue share from the sms received into a service (rather than paying to receive an sms) is a powerful incentive to create sms based apps.The US is different I know in that it is common to pay for every sms that is received however I know from experience that having to pay for inbound and outbound messaging can put off public agencies from implementing information services because they do not like the unknown cost potential. If a service can be cost neutral then that makes them happy!From a developer’s perspective again there are a whole host of services that could be provided via sms but if they have to pay every time an sms is received then many of those services will not appear. Advertising within an sms has been an unsuccessful model so far.It may not be possible using the current numbers that Twilio has to derive revenue from them. If not then they need to think about obtaining a short code (premium rate) and if possible (from a US regulatory perspective) share it with their subscribers using keywords so that their subscribers can derive revenue from a service or at least make it cost neutral to provide.
This is why Twitter as the backbone for these types of services interests me. It’s still SMS-accessible, yet doesn’t cost the receiver, as they wouldn’t be handling it via SMS on their end. To make this a reasonably robust solution, though, Twitter accounts need to become much closer to ubiquitous. Or, Twitter needs to auto-create accounts the first time someone sends an SMS message to the service, similar to how Posterous creates a new account the first time it receives an email.
Interesting thoughts here, I’d like to keep this conversation going. We’re very focused on helping developers monetize their apps in 2010 and your thoughts on this are valuable to us – could you possibly drop me an email danielle (at) twilio (dot) com if you read this comment (or anyone else who would like to discuss monetizing apps)?Thanks!danielle @ Twilio
I’m a long-time friend and fan of twilio, and they really do rock. They’ve taken an industry which is traditionally an expensive pain-in-the-ass to work with, and made it dead simple. I had previously worked on a project using angel.com, and my god was twilio a breath of fresh air.Love the new features, and so happy you guys have invested in them.
man this looks so awesome. hope the product lives up to the promise. if so, this is another great “power to the people” play. work from home businesses could be a major stakeholder.i hope they will consider a simple product for stupid people who are not developers. perhaps the most common telephony apps can be boiled down into an easily customized app requiring no development knowledge. i think that would be a great addition to their product line all around.
I agree Kid there is a big opportunity for them if they were do that
KM – Are you thinking something like a form-based app builder or yahoo pipes? That’d be cool, and something a developer could build on top of the twilio platform. But you should take a look at their product, it really is quite simple, you can do a lot with just simple XML files, no programming required.They also have something called “twimlets”, which are simple canned functions that you can get started with just by building a URL string. Honestly, I don’t find that much easier than writing the XML, but I also write code all day…http://labs.twilio.com/twim…
wow, the twimlets thing is pretty simple. that’s really impressive. i’m going to have to find some time to play around with that more, seems simple yet powerful (best combo IMHO)
They are headed there. I’m doing my home and office phone system on twilio. Its alpha stage but not for much longer
very cool, so theoretically a marketplace could use Twilio to connect buyers and sellers by phone as well. Could be a really useful API. Looking forward to playing with it a bit.
dope! way easier than figuring out a users cell provider and then sending an email. and at 3¢ per message, a file is deleted, api bookmarked!
CAn’t agree more with your vision on what opportunity the developer masses are now 🙂 That’s exactly what Superfeedr is : making web developer’s lives much easier!
Smart investment! Twilio is great, I love how simple their API is. When I first heard about them, I whipped up a little (admittedly half-baked) app to try it. I had a functioning phone system in a couple hours (http://github.com/cpetersen…. I’m excited to try the SMS capabilities.In general I really like where the industry is going with these infrastructure as a service type plays. Things like twilio.com, heroku.com and couch.io are all really exciting.
Knowing a very small bit about telephony- this will be a huge shakeup, especially if you have to deploy massive amounts of phones. If it is a good deal, people will flock to it, and definitely make specialized applications for the corporate side…
I’m excited with the possibilities made possible and simple by Twilio. Another great pick by USV, i’d say that 1/3 of my top20 favorite web apps/apis/companys (twitter, mongodb, boxee, disqus, zemanta, foursquare and now twilio) are under USV umbrella.
Fred, love the Twilio concept. Given your involvement in user + developer platforms and developer-only platforms what do you see as the pros and cons in terms of network effect & execution in building a platform that caters only to developers like Twilio or platforms that cater to both users & developers like Twitter & Foursquare? For example, Tickreel has been designed for developers but we also built example apps to illustrate to developers the compelling use cases. Its hard to know where we will see the greater network effect. Given your experience, do user applications help drive developer adoption (as in Twitter) or does it make execution more challenging given you effectively have 2 customers from the beginning to manage?
One of the many benefits of having a developer platform is an upward spiral where user applications drive developer adoption and developer adoption drives more users. Twilio gets this far better than most.
Agreed there can be great synergies if you nail it for both users and developers as Twitter has shown. But my question is what are the costs and benefits of catering to both from the outset (where you get maximum potential network effect and illustrate a use case) versus catering to just either users or developers initially. Love to get some insights on this.
Goodbye, switchboards. We will not miss you, though.”We believe that one way to build a large network of web users is to build something that makes developers’ lives easier.” Developers==Users. Fred, it would be nice if you could dedicate a post for that?
I like Twilio as another push towards dumb pipes. It will be interesting to see if Google decides to get into this game by exposing a voice API.
This does sound really simple and effective. And the combination of first-to-market and developer-friendly is a definite winner.I’ve been wondering why something like the SMS piece doesn’t already exist via Twitter (apart from the obvious answer of the Fail Whale). To me, services built on top of twitter still haven’t come close to realizing their potential. A zillion clients, search tools, and spam bots are nice and all, but why not use it as an entry point for users to reach web services? A service like this could certainly handle examples similar to those in the Twilio video: @cornerbar what are your hours? @lunchcounter what are today’s specials? @towhomitmayconcern what can I ask you? This seems to be just a small piece of what Twilio can do, but it would seem to be a piece that businesses could implement over twitter for just the price of some software on their web server, without recurring costs: just an xml file with the response data and a php file to receive and process the twitter “request” via the xml file. Surely a lot of work could go into processing semantic requests, but a really simple version could simply return “I didn’t understand. Try hours or specials or directions…” when a request can’t be parsed. The obvious downside is that users would need to (not only know what a twitter account is, but also) have a twitter account to use it.All that said, the potential to provide services immediately that take advantage of the Twilio interface is fantastic, and it’s all in terms that any client can understand.
They probably could work together and resolve some of the problems mentioned here…
Yep. Definitely not mutually exclusive, and the back end would be doing the same thing for both.
At first glance, looks like a ‘global HTML element name space’ problem: Or, suppose there are 1 million Web pages with Trilio ‘tags’ or HTML ‘elements’. Suppose HTML 6 has a new tag <sms> for something quite different than what the Trilio tag does. Now the Trilio software has to be updated, every end user trying to make use of the Trilio software has to download a new version, and 1 million Web pages have to be updated before HTML 6 makes much progress. Real soon, now, y’hear? So, this is a ‘network effect’ that becomes Monday morning Excedrin headache 847,223,485,982. Or, so it seems at first glance.So, as usual for such namespace issues, need a ‘registration hierarchy’ as was attempted in CMIS/P. The simple way, usually effective enough practice, is just for a company like Trilio to use tags, say, <trilio-sms>, etc.
When I worked for Apple as a systems engineer building telephony apps was a ton of fun (but not always so easy). My first exposure to Twilio was at a weekend startup competition and Twilio was the runaway star of the event purely because of their developer friendly stance, their approachability and the high value to low effort ratio. GREAT investment guys.Am curious how they landed on your radar… were they introduced? A cold “call” (sorry, no pun intended)? or some other approach?
congrats, Twilio is an absolutely amazing idea. the marketplace has needed something to allow web creativity to freely extend into telephony. Personally, I have brainstormed on the platform for hours at end.It will be really interesting to see if the company can survive the temptations of a middleware company and not develop apps that compete with its targeted customers. Grasshopper and Ringcentral are just the beginning.per minute pricing is a bit steep but nonetheless, a very cool company.the pbx is dead. long live the pbx.
It’s interesting how the old maxims still apply: in a gold rush, all the money is made by the people selling the mining picks and gold pans, not by the prospectors themselves.
Fred…I think you nailed another one…sorry…I’m excited. If I understand this correctly…this is huge.
I have to say…get a VP of Marketing STAT!! Cause someones timing was off. 🙂
SMS is a command line interface – whether you use twilio or you provision through an aggregator. SMS is an absolutely killer platform for making impulse connections with potential merchants using a a web proxy.
What about MMS? Can/will it allow send/receive of pics/videos? [future roadmap]
If anyone is looking for some business/application ideas for using Twilio for… I have had some SMS & VXML ideas in the past:I think using SMS to then send URLs to the user, so they can click and view things in their smartphone, is going to really come into play. Or calling a # in a job classified ad to hear a full description about the job & company [audio is more personal than text]. Ditto on calling a number on a product at a store that I’m looking at to walk me through the features.http://blog.stevepoland.com…http://blog.stevepoland.com…http://blog.stevepoland.com…http://blog.stevepoland.com…http://blog.stevepoland.com…http://blog.stevepoland.com… [a dating or chatroulette idea for the phone]http://blog.stevepoland.com…http://blog.stevepoland.com…http://blog.stevepoland.com…
Great news! One of my startup clients has been looking for outbound SMS! I left a msg and definitely need to talk to them!!!
At TOPP, we’re starting on a few projects that could integrate voice and SMS nicely, but we hadn’t looked too closely at possibly providers yet. This weekend, I gave the Twilio API a run through and found it super easy to use. Thanks for the heads up, Fred. The result of my weekend project is http://phone-idol.com. No one’s taken me up on my offer to sing online just yet, but the experience definitely proved to me how easy Twilio has made it to integrate sms & voice w/ the web. I’ll definitely be back.
thanks so much for sharing that nick. i love hearing stories like this
Twilio is providing a core piece of infrastructure that is mission critical to the businesses that use it. But charging for our service, we assure our customers that we will still be around tomorrow. Telecom also has hard costs on our end, so we would have had our hands tied and had to give up a lot more of the company raising funding if we had gone with a freemium model.”Pay to play” probably doesn’t work for every API, but for ours it is the only way to go.Danielle @ Twilio