Why Twilio Acquired Segment | Hacker News Reader

Why Twilio Acquired Segment

rudderstack.com

81 pointscpard26 comments

I feel like Twilio is making a move into becoming the go-to Enterprise/B2C API platform, being already the biggest company in SMS and Voice message APIs, acquiring Sendgrid to gain advantage in Email APIs and then Segment to be able to manage customer data. Think they might go acquire some kind of customer/user analytics API next.

Realized this when I use Twilio SMS + Video, Sendgrid and Segment together before their acquisitions and now I'm just sort of using Twilio, Twilio Sendgrid and now Twilio Segment.

It isnt a bad spot to be in. My only gripe against Twilio is its definitely priced for enterprise.

I wouldnt build a startup on Twilio its costs eventually add up much higher than competitors. "Just because someone spends a lot on marketing doesn’t mean you should be paying for their marketing budget, especially as the core product commoditizes."[0] On the other hand they definitely have a rich API and I suppose they have tremendous uptime to boot. I never had a single issue setting up Twilio or Sendgrid any issues I had were with a web framework or other libraries, or network firewall but never Twilios APIs specifically.

[0]: http://alanquayle.com/2018/02/sms-api-pricing/

Who do you use instead?

I have tried bandwidth.com before. They are cheaper and claim they own the VOIP infra so can afford to be.

https://www.bandwidth.com/twilio-alternative/

For a fun project though so can't comment on uptime etc. However, if they are powering the VOIP infra behind the other providers, should be good I guess?

what did you do about the minimum monthly requirement?

Got a free trial key. Didn't know they had a min requirement

It seems like the most successful SaaS companies eventually expand beyond their initial product category to become a platform with a wide range of services and a whole ecosystem around them. Some go really wide, like Salesforce, while others seem to try to keep some coherence around the core product (at least for now), e.g. Zendesk and Intercom. It seems Twillio wants to go quite wide too.

I wonder if Stripe and Twilio merging could ever make sense - that would make for quite the powerhouse.

Beyond the developer friendly aspect, I would think running a payment infra is very different business than running SMS/email/integrations? Former has lot of challenges around compliance, fraud etc too.

If Twilio wants to own the entire developer ecosystem, maybe they will do others like APM first (Datadogs etc)?

Not to say a merger makes sense, but telco-adjacent work like voice and SMS also has massive challenges around stamping out fraud and lots of regulatory frameworks to manage.

"Twilio was a surprise acquirer. Many expected Segment to be picked off by one of the usual suspects like Adobe or Salesforce."

In terms of culture, Twilio would be a better fit, given their focus on developer evangelism in user acquisition which is similar to Segment's strategy. In fact, if you look at Sendgrid, which Twilio has acquired 2 years ago, it also adopts a similar strategy.

Never used Sendgrid (have always used mailchimp) but was it also considered very developer friendly?

Yes, we use it for transactional email and it's a breeze to integrate, just works, and has great support.

> just works

Unless you try to send multiple requests to it's API at the same time. We've had to build in retry with random backoff into our mail sending helper because SendGrid's API doesn't like multiple requests at once.

> has great support.

Anecdata, but my experience with Sendgrid's support during a production outage was that to get a response from a human being we ultimately had to reach out to somebody on LinkedIn. Everything else via official channels just received automated replies promising responses in x business days. Also FWIW while our production e-mails weren't being sent, the Sendgrid API was happily returning HTTP 200 success responses.

I agree that it definitely was a breeze to integrate with though.

We converted to a paying account after having an inactive account for a year, some automated system turned off our account AFTER we started paying, the API accepted all of the emails and didn't send or save anything. We had to go in and rebuild all of the lost emails on our backend (eng cost > total Sendgrid spend per year). Support was unresponsive until we escalated via a sales phone number. Typical large company automating dangerous behaviors and then not having a good support system to deal with the fallout.

Wow that sucks.

We had one emergency where someone figured out how to mass mail through our app (flawed "invite via email" feature), we were able to immediately reach support through the live chat feature and stop some 200k spam emails from being processed (they were in their queue by the time we noticed). Saved our butts.

FWIW, my experience is they used to have great support. Sendgrids support seems to have gotten considerably worse since they married Twilio. I loved Sendgrid several years ago but they seem to be having more issues now and drastically slower to respond to tickets. Some of their older APIs have suddenly stopped working (no code changes on our end, and supposedly no changes on theirs...) and it's taking weeks for support to reply. Years ago you'd be able to chat with someone online almost immediately, now you submit a ticket and hope they respond.

Cant be worse than Mailchimp and mandrill. Mailchimp acquired Mandrill and it's essentially abandonware - no updates for years - and when it goes down its often an hr before they even realize, during which they have just silently discarded messasges* and they never follow up with affected customers. (* sometimes they will show as errored in activity log but no retry, no notification). I hope Twilio doesnt do a Mailchimp

IIRC they were the ones who created mandrill but after some time they decided to stop investing in it (because it wasn’t aligned with their goals, or something along those lines). Rather than completely shutting it down, they left it online to support people who were already using it. You can debate the merits of that approach, but as I recall they communicated clearly about it at the time.

While developer evangelism is important, the business strategies must converge otherwise there would be no value in the acquisition.

In the current environment of everyone against 'big tech' it would be hard for Adobe or Salesforce to justify the acquisition of Segment from a competition standpoint, for they would consolidate the CRM market even further.

Acquiring companies that are adjacent to your core business makes total sense in adding value to your platform, and not acquiring core functionalities allows you to claim that competition is getting stronger.

My prediction is something similar with Twilio + Stripe. If Twilio wasn't already public with a higher mkt cap than Stripe's most recent valuation, I'd guess Stripe acquires Twilio, but as it is not sure. Anyway here's my thinking:

1. Complimentary businesses: Stripe is a developer interface for collecting payments from customers, Twilio is a developer interface for communicating with (and now understanding / "segmenting") your customers.

2. Similar culture: very high emphasis on great DX and developer evangelism. Shoutout to Stripe's docs for being the best in the biz, and Twilio's are not too shabby either.

3. Unique opportunities: Stripe collects payment for their customers, Twilio collects payment from their customers. If everything was under one roof, you could have a situation where Stripe becomes a de-facto bank account. Revenue comes in to your Stripe account and many of your consumables (paying for Twilio's services) could come directly out of that account, removing the need to egress funds to your bank at all, thus reducing financial transaction overhead (which at scale could be a sizeable gain in efficiency).

You can also argue that Twilio+The Trade Desk makes total sense:

Enterprises could have a single point of truth (Segment) and target existing and potential customers via sms (Twilio), email (SendGrid) and digital ads (The Trade Desk).

As every business, becomes a software business - the underlying features of building a software will be outsourced to specialists.

I am just calling Unbundling of Building

Wrote some thoughts here:

https://twitter.com/ankitpansari_/status/1314915795433000960...

There are two important concepts here - 'Abstraction' and 'Undifferentiated Heavy Lifting'

A whole group of software development functions - hosting, payments, storage, analytics, communication are being abstracted by modern services such as AWS, Twilio, Stripe, and Segment.

A software developer can now focus on the core value of the product.

Werner Vogels, the CTO of AWS, coined 'Undifferentiated Heavy Lifting'.

He suggests that businesses should use resources on-demand and pay for what they use, leveraging each other competency.

Shared some slides on my Tweet here - https://twitter.com/ankitpansari_/status/1314915795433000960

Reply on news.ycombinator.com