Twilio’s Dutch rival MessageBird plans an IPO
bloomberg.com
bloomberg.com
Twilio resells commodity product (SIP/SMS) at Saas margins by offering good developer experience and - by this point - major brand recognition. It’s not a defensible business long term, they just had a huge head start, and they still compete terribly for service outside North America.
This is why they are building on top of the underlying commodity and moving upmarket into things like call center automation products, which start to compete directly with many of their customers.
You can easily negotiate down your SIP/SMS Twilio prices the second you realize they have plenty of competitors.
If a smaller competitor gets their name out there by doing some basic guerrilla marketing against a behemoth public company with tons of money, it seems pretty benign. Everyone wins but the overpriced incumbent.
I have so much respect for Dropbox. I remember laughing when Dropbox gave away 2GB for free back in 2009(?) and I never really thought they had a valid business plan. I read Drew Houston talked to Steve Jobs and I still don't understand why he wouldn't sell. But then I'm poor. Anything that puts more than USD 20M in my pocket would be life changing for me.
> What Houston does is Dropbox, the digital storage service that has surged to 50 million users, with another joining every second. Jobs presciently saw this sapling as a strategic asset for Apple. Houston cut Jobs’ pitch short: He was determined to build a big company, he said, and wasn’t selling, no matter the status of the bidder (Houston considered Jobs his hero) or the prospects of a nine-digit price (he and Ferdowsi drove to the meeting in a Zipcar Prius).
https://web.archive.org/web/20200520184823/https://www.forbe...
And of course Houston is a billionare from Dropbox since it went public a little over two years ago in 2018.
—
I remember thinking Snapchat, namely, Spiegel and Murphy co-founders as being a bit crazy to not sell Snapchat to Facebook in late 2013 for $3B when Instagram was bought for $1B 1.5 years earlier.
However the two remaining founders had taken $10M a piece in the previous 2013 funding round. So like you said about $20M being insane as it would be for me. A lot easier to turn that down with all the possibly upsides and millions in your bank account.
Even with Snapchat not exploding the way it was thought it would later on: - a year after rebuffing FB, they raised almost $500M at a $10B+ valuation in late 2014. - 2.5 years later, in early-ish 2017, they went public at $30B+, raising $3.5B. - Tencent became the next big company to have a large stake in Snapchat. After being a previous small investor. Gathering 12-18% of the company by late 2017/early 2018 when Snap stock was doing pretty badly. - Today, the co-founders still have super voting rights and Snap stock is back to being above $30B while seemingly everyone copies them.
They unethically (IMO) kicked out the third co-founder for a while by then, like Twitter did before them, but less ruthlessly in the end since Snap’s 3rd guy, Brown, still made out very well.
Twitter’s Noah Glass on the other hand...it’s just too sad. No less seeing the worst Twitter guy of them all, Dorsey, getting all this praise for his recent big donations with almost no mentions of how he began his lies and consolidation of power [at Twitter]
Messagebird, Plivo, Nexmo (part of Vonage), Telnyx, Voxbone, Signalwire
You would think! I was one of the first interns at Twilio in 2011, when they were about 20 people. When I was deciding whether to go back full time, this was exactly my concern. The product was great, people loved it, but it was ultimately just a SaaS layer on top of open-source PBX software. Already they were starting to attract lower-cost competitors. I thought they didn't have much of a future.
And now, 9 years later, that tiny startup is a $30B company.
This was a great lesson for me on how facile this sort of business model analysis is. It turns out by radically simplifying a hard, important problem (programmatic communication) and providing a great product around it is worth a lot! It doesn't really matter if someone can hack together some of the pieces that make up Twilio in a weekend. Many, many businesses will pay a premium for a good, cohesive product that works well.
If you're targeting competitors' customers after they've spent the time, money, and effort to come to a conference, they're probably not a good prospect. They're deep and unlikely to switch easily or quickly. There are much better targets elsewhere.
* Early Twilio evangelist here, circa 2011-2013
Twilio and MessageBird have the same customer. Someone who is using Twilio could reasonably convert faster than someone who doesn't use any api-based telephony integration. You skip the step of convincing internal teams that you should do this in the first place.
At ZenPayroll we didn't just ignore people on ADP, Paychex, Intuit, etc. That would be nuts!
If you're trying to reach "people who integrate with telephony APIs" the Twilio conference is…the perfect place to find them!
re: spending time/money/effort to come to conferences:
1) They are networking events.
2) They are paid for by companies so employees aren't by default that invested.
3) It's not a lot of effort. Employees like going to conferences. You meet people. It's paid for. You don't really have to do anything. It's a workation for most.
As for retention, much of Twilio is mostly a commodity. MessageBird came to the US and started a price war, which is really what you're competing on. Switching SMS APIs isn't the same as, say, switching off of SalesForce.
In payroll processing, there are a handful of major entrenched providers. Going (almost) directly at them is the only approach. Yes, you have to differentiate yourself but odds are there's a rip & replace coming.
In 2012, outside of SMS aggregators, sending and receiving text messages was still novel. Add in automated calling and there were some options with Asterisk (worked on that many times) but still novel. Going after that tiny market may have been cheap but probably not effective.
If MessageBird started a price war, that's a weak value prop by itself but could work so followups:
- How much have they driven down pricing across the space since they've come to the US?
- If that value prop is the main motivator, how much share have they taken?
- Are you going in on their IPO?
(I don't care either way, I don't own any Twilio shares anymore.)
Prices to the US are already super cheap. I think twilio was at or close to a penny when I first saw them. Prices elsewhere could use some lowering though, and it's a lot bigger deal if a 50 cent call becomes a 40 cent call than if a 1 cent call becomes a 0.8 cent call.
It kind of depends on what you're doing though. At my last job, I was a high volume customer, and we never went to the conventions, because everything worked / it was essentially a commodity for us; we had several vendors and directed traffic towards best results and price. I guess we would have met message bird earlier, instead of when they aquired one of our vendors.
You are informing a self-selected set of power users about a direct alternative - you can't get cheaper exposure than this.
How much money they would have to burn on advertising campaigns to get comparable exposure on such relevant audience ? Even if it leads to 0 conversions the awareness inside of that group is very valuable.
Oracle have been doing the trick of having all the taxis outside a competitor event decked out in the Oracle logo. They used to do it to Sybase in London in the 90s.
You know of the competitor and it stuck in your mind enough to remember because who doesn't love tacos.
Larry may be wasting all that money on taxis when all he needs is a taco truck.
Taiwan has democracy, doesn't threaten its neighbours, has a working legal system, respects human rights, and has an educated population with access to the free world.
Calling it part of China is highly offensive and bowing down to the communist propaganda is simply unacceptable. They have Huawei as a client so choose money over morals.
Let's make it more expensive for companies to give dictators what they want.
I don't know how many 200 OK I've received but the message was never delivered (and soooo many times during vital demos). I had several support cases with them and they simply ignored me. My product is built around SMS so I could have been easy money for them if it actually worked reliably.
I ditched them for Twilio, never had any issues since then. A couple of months later someone from sales at MessageBird reached out to me on LinkedIn and asked if I wanted to try their service. Spilled my guts to him and he wasn't surprised...
How certain are you that your messages are delivered? When I started digging, it was scary how often it failed successfully.
Their web UI is pretty buggy. Certain parts of their control panel sometimes causes my browser tab to freeze. Recently I found that pressing tab caused the page reload when filling out verification info for buying a new number.
Edit: For which ever country I choose from, they always have sms inbound Free. Feels like a marketing/sales trick ? I can't believe that MessageBird has managed to implement inbound SMS for all those countries, while Twilio hasn't.
Telephony in general is fairly cheap, so I'm guessing there's only one or two providers here providing VOIP termination, and without much competition they price their services as high as enterprise companies who need it will allow.
Reliability in the telecom space is a difficult problem, it's a large ecosystem of providers, aggregators, regulators, and lots and lots of hidden complexity. Then there's actually running the service, keeping the REST API up 24/7/365, making sure the TwiML processors are processing, keeping everything humming along smoothly for billions of requests a week.
When I was there, we would see "Twilio Killers" launch once or twice a year, normally competing around price. Their launches would almost always have the same 53 countries that they could deliver messages to, which we knew from operating in the space meant that they had just white-labeled Bandwidth's offering. They would be relying on a single aggregator and have the same reliability, resiliency, and reach as that one network.
There's a lot of reasons that people choose Twilio, network effects can't be denied, a lot of effort was expended on keeping it top of mind in developers heads. But that only scratches the surface.
We focused a ton on being reliable, being easy-to-use, and on backwards compatibility. There's a lot of unmaintained WordPress plugins that were written 9 years ago, abandoned 8 years ago, and still work fine because API compatibility was a top priority. There's something to be said about being able to just build an integration and then mostly ignore it and it just hums along sending SMS out reliably. A developer spending a day debugging a less reliable service can obliterate the savings in price, sometimes, depending on volume and use case.
Ease-of-use and network effects shouldn't be discounted. Most mainstream languages have a reasonably good library, Twilio officially maintained 6 of them while I was there. Our Developer Education team spent a ton of time and efforts creating tutorials, quickstarts, blog posts, and presentations. The Developer Evangelism team could be found at nearly every Hackathon, Conference, and Meet Up, spreading the word about Twilio, with well-polished live coding demos that would spark a lot of "Wow moments."
Even knowing exactly how every part of it worked, there's something magical about a presenter having everyone in the room text a number, then having a 3 line ruby program hit the API, pull down all the messages, and display them in the terminal. Then with two or three more lines of code, text everyone that texted in a message. That kind of demo takes maybe 10 minutes to do, but people would practice and polish it until it was seamless. You could look around the rooms at some of the events and see the wheels turning in people's heads as they started thinking about how they could integrate this functionality into their product.
First-mover, network-effects, reliable, resilient, easy-to-use, and in the beginning there were a number of highly promoted price drops to keep the momentum up.
MessageBird, as you said, came to the US already with a _well_ established business. Obviously they weren't competing for the same investors/checks as most of the rest of the batch, but it was still hilarious to hear them pitch next to everyone else.
Direct listings have been getting a little more popular; in a direct listing a company just puts their stock up for sale on an exchange without going through the IPO process, and it just starts trading at whatever the market thinks is fair. But it's still fairly rare to see a direct listing.
Been using it for a while for mostly US sms, works fine:
I have begged Twilio engineers in person at multiple Signal and over email/support channels as well as the CEO and here in this very forum: For the Love Of God will you please add an 'email' twiml verb ?
If you want to CC: a SMS to email currently, within Twilio, you need to build functions, sign up for a Sendgrid account, authenticate in sendgrid with real names and physical addresses ...
It's a no-brainer use-case and it is nothing but pain to implement.
Actually it's not:
First, email and sms are different transport mechanisms.
Second, email requires a To (envelope) address.
Third, some mgmt. like bounce and DKIM/SPF is needed.
You could write a subroutine to hide some of that, but I'm not sure there's a clear general solution.
When I say that it is a no-brainer, I mean that it is a no-brainer feature ... that is, something that everyone would have a very ready use-case for.
Second, I think we're all well aware of the vast gulf between email service and the mobile phone network. That's the point of a service like Twilio: they have both systems (email and SMS) terminating inside of their own cloud environment. This means that "forwarding" an SMS to an email address, which would be rocket science for me, is easy for them.
In fact, it's so easy that we can readily describe exactly how to implement it: just give us an email verb in twiml.
There are a lot of ways I would use this, but the basic use-case is as follows ... my primary phone number lives at Twilio and I receive SMS with this simple twiml:
<Response>
<Message to='+14155551212'>{{From}}: {{Body}}</Message>
</Response>
<Response>
<Message to='+14155551212'>{{From}}: {{Body}}</Message>
<Message to='me@mydomain.com'>Subject: Incoming SMS {{Body}}</Message>
</Response>
The immediate issues that comes to mind is spam and unwanted messages, etc., but that's easy to solve - just have a simple challenge/response proof that you own whatever email address you are sending to and only allow people to send to email addresses they own. I am, after all, just emailing myself.
cc'ing SMS, in programmable messaging, is currently very very difficult and time consuming. You either need to host code at some third party site[1] or you need to sign up for a sendgrid account (again, third party, additional account, etc.) and write functions to pass onto the sendgrid account ... but you can't actually fire off emails from sendgrid unless you authenticate with actual real names and physical addresses ...
I just want to cc: myself from within twiml. Christ.
[1] https://stackoverflow.com/questions/44768719/can-i-forward-s...
Have a look at the SMS Server. This product is able to forward SMS to e-mail and visa versa. Works with almost any SMS and e-mail provider including Twilio.
Here's a link to the website: https://www.auronsoftware.com/products/sms-server/
Disclaimer: Auron Software is my company and I'm the main developer.