Auren Hoffman

Hello fellow data nerds. My guest today is Jeff Lawson. Jeff is the co-founder and CEO of Twilio. If you've ever received a two factor authentication code or an order confirmation code via text message, then you've interacted with Twilio. Prior to Twilio, Jeff co-founded, uh, Versity. Is that how I pronounce it?

Jeff (00:52.258)
That is, that is how you pronounce it.

Auren Hoffman (01:01.111)
Okay. Uh, which he sold for $30 million back when he was just 23 years old. He was also the co-founding CTO of StubHub and one of the first product managers at AWS. And he founded Twilio in 2008 in the height of the financial crisis. Twilio went public in 2016, and now has a market cap of over $10 billion. Jeff, welcome to World of Dats.

Jeff (01:14.862)
Thank you, Auren, great to be here.

Auren Hoffman (01:16.731)
I'm excited now. The idea of empowering developers is kind of central to Twilio strategy. And it was kind of like one of the core founding thesis, I think, when you started. And now it's a little bit more common, but I think back in 2008, it was quite uncommon. How do you think of Twilio as a company because it's so developer focused?

Jeff (01:41.598)
You know, to us, it always seemed like a few things. It was like the confluence of several factors. Number one, it was, you know, at some point, like, you don't have to worry as much about markets and all that. Like you just worry about a customer. You're like, okay, is there a customer who has a hard problem that if I solve it, they can pay me. And what we found with communications was it sat really well at the intersection of, of those notions. Right. Because.

Auren Hoffman (01:56.486)
Yeah.

Jeff (02:10.382)
Companies spend a lot of money on communications. It's a well-understood domain. They've got budgets. The value of it is well-understood. But it was inaccessible to developers, historically, because if you're a developer, you are often asked to add capabilities into the products you're building or the experiences you're building. And if you'd ask the average developer, hey, I want you to build a webpage. Sure, no problem. I want you to send an email. Okay, I know how to do that. Okay.

I want you to send a text message. It would be like the record screech, like what? You know, like I don't know how to do that.

Auren Hoffman (02:42.39)
We're at all these like telco gateways and stuff.

Jeff (02:45.366)
Yeah, like making a phone, you know, ring somewhere in the world, like, that is a completely foreign, foreign thing. And I remember I had that very experience myself, as a developer. So you know, for example, at StubHub, we were building StubHub. And I remember thinking like, Okay, well, if I've got a seller, and a buyer who have to like meet up at the venue to exchange the ticket, like, how am I going to coordinate that interaction? And we thought, well, wouldn't it be it'd be easy if we just allowed them to essentially like, you know,

call each other or to send a text message. And it was like, that's a really neat idea. How do I do it? I have no frigging idea how to make that happen. And when I realized that I had a similar experience at every single company I had started prior, I was like, huh, there's this big unmet need going on in the market where the experiences that we all have every day in these digital products need communications, yet the people who are tasked with building those experiences can't do it. And so some amount of the time,

Auren Hoffman (03:31.835)
Ha ha ha.

Jeff (03:43.774)
you know, maybe they figure it out, but oftentimes it's just, no, it's just not getting built. And I had that experience, right? So every time I had these needs, what I would do is I would say, okay, well, I don't know how to do this. And I pick up the phone and I call the companies who seemed like they didn't know how this stuff works. I call carriers, I call the, you know, equipment people like the Cisco's of the world and say like, okay, we have this idea, we're trying to make this happen. How do we do it? And remember, every time you would if you could get a salesperson on the phone, which oftentimes you couldn't, but if you got a salesperson on the phone,

they would say, oh, okay, well, we think we can help you with that idea. We're gonna, you know, first of all, we're going to, you know, you need to roll out the copper wires to from the carrier to your data center. And then it's like, on them, we got to rack up a bunch of hardware, and then you're gonna buy some software, but it doesn't do what you want it to do. So then you have to bring in our professional services who are going to come and we think we can bang this whole thing into submission and get the idea that you need. That'll be about four years and the seven million dollars like click sign here to get started.

Auren Hoffman (04:17.863)
needs assessment.

Auren Hoffman (04:37.915)
Hahaha

Jeff (04:41.782)
And every time I was like, what do you, you know, first of all, millions of dollars for, like, I don't know, that kind of money, like we're startup, but more importantly than the money was the time. This idea that like we, we

Auren Hoffman (04:52.379)
Well, and it was a project manager. It was just not something you could just like a, an engineer just can't do it and, and just run with it or try it out or something like that. Like it becomes like a core strategic thing. Like the CEO has to manage or something at that point.

Jeff (05:05.554)
Exactly. And it was like, wait a minute, but like what in the world of software that we do is like years long roadmaps, right? I'm accustomed to software being agile, right? We have an idea that maybe takes a while to build out, but you build out that idea with customer feedback along the way sprint by sprint, making it better and better and better with customers guiding your way at every step. And so this idea that we have like, here's what we think we need.

Then we're going to go spend millions of dollars in multiple years building it out and then push it out into the world, fully complete. And then customers tell you, wait, that's completely wrong. That's not what we needed. Um, was ridiculous. It means like back to waterfall development practices. And so what we found was that we could actually bring communications into the realm of software developers and into the world of software period, like agility, sprint over sprint, like using all the software practices with this brand new thing in 2008 called, you know, web API space.

Auren Hoffman (06:05.167)
And the way...

Jeff (06:05.29)
And this was an insight coming from my time at Amazon, at AWS, of seeing like, oh, wow, infrastructure delivered to software developers as APIs that are pay as you go. I had an early window into this notion of like, this is going to create a whole new buying class in terms of software developers who previously were unable to buy a bunch of services because they didn't have the authority or the experience in like, you know, making big purchases of like storage systems or whole data centers or

communications. But now suddenly if you can pay a penny and just get a capability that you need to plug into your app, it's a whole different way of going to market. And you've opened up a whole new set of buyers that previously weren't able to buy these services. And so that's really the founding, I'd say, insights that enable us to build Twilio.

Auren Hoffman (06:54.043)
The way I think of Tulio is in a very similar way I think of Stripe, which is actually a pretty complex problem. It doesn't make sense for every company to reinvent the wheel of that problem. No company needs all the features. Every feature that Tulio has, let's say, is really a subset of the features that every company needs.

But every company needs a different subset of that. So you have to build lots of things. Now, same thing with Stripe. You know, maybe I don't need to pay someone in Malaysia, but somebody through Stripe needs to pay someone in Malaysia, right? And say, you know, maybe I don't need to text someone in Malaysia, but somebody through Twilio needs to text somebody in Malaysia. So you have to interact, you have to do all these types of things. And it's just, it's like a, it becomes like a common good in a way.

Jeff (07:39.33)
Well, and I think...

Jeff (07:43.374)
Maybe my shareholders wouldn't want us to be thought of as a common good, but rather a common need of the digital economy. No, it's interesting because I think the world of software, what has happened is that if you roll the clock back 30 years ago and you think about the world of software, there were very few companies who built software in the world. Everybody knew that.

Auren Hoffman (07:48.019)
Sure. Yeah, yeah, yeah.

Auren Hoffman (08:08.9)
Yep.

Jeff (08:11.274)
And those are the companies who knew how to build software and everyone else, you went to them and you went and bought what you needed. And then what happened is, around the turn of the century, or century and the millennium, right? You got this shift, which was, well, wait a minute, because of software as a service, the market fundamentally changed. In that first world, software was thought of as-

It was like the era of big software, right? You needed a new ERP system. You went to SAP and they quoted you $30 million and five years transition time. And only the CIO could make that decision because it was so big, so expensive.

Auren Hoffman (08:47.023)
Yeah, I need to hire Anderson Consulting at the time to come in for an extra $4 million to integrate it.

Jeff (08:54.354)
Right? And then what happened is with software as a service, the industry fragmented and said, no, actually, you know what? Every project isn't this big, it's this complex. You need the CIO to make the decision. Actually, we can simplify it. So the head of sales can buy the sales software application they need. They had a marketing, can buy the marketing software application they need. And they basically went after every persona in the company with like a department and budget. Okay, what's the software you need to do your job better?

And we're going to go after the budget you have to do that and provide you essentially workflow, the workflow of your job function as a software that helps you manage it better, right? And that's what the story of SaaS really is. And so that was the first decade of the millennium was all these SaaS companies figuring out the head of HR, the head of marketing, sales, whatever, everyone's got budget and everyone's got workflows, great. We're going to go in and we're going to sell to them. And so you exploded the number of buyers of software from essentially the CIO to now the heads of

all the various functions inside of a company. But the thing is, you still are in a world where the people who built software were the vendors, people selling you an application. And what happened around probably 2010 with mobile and with just so much commerce happening on the web is that every company started to become a digital company. And every company realized that if the experience I give my customers is a digital one, then...

I need to go build a digital experience that is differentiated from my competition. I can't just go buy that off the shelf. If I'm a bank, 30 years ago, my customer experience was basically the branch. I said I liked my bank if there was enough parking in the parking lot, there was a fresh coat of paint on the place, and if it was a right-hand turn on my way home from work, I really like my bank.

instead of a left-hand turn, right? I mean, things like that would make you loyal, but nowadays, of course, your bank is a mobile app. And so you like your bank if that app is fast, if it doesn't crash, if they adopt the newest face unlock feature very quickly, you're like, okay, I like my bank. It's an easy to use app. And in that world, every bank now has to become a builder of software. Every retailer has to become a builder of software. Every company has to become a builder of software if they're going to compete. And when that happens, now suddenly the number of software.

Auren Hoffman (10:43.77)
Hehehe

Jeff (11:11.63)
developers, the importance of the software developers, but the number of companies who are building software explodes and therefore the needs that they have, like a software supply chain fills in to help them accomplish those goals. And that's really what the API economy is all about.

Auren Hoffman (11:27.827)
And the number of developers has also gone up dramatically, both because we have developers now from all over the world and more people going into software development. And it's just becoming easier to get up to speed. There's more tools. I remember probably when you and I started doing software development, it was a huge slog to do anything. And today it's still difficult, but way, way easier than it ever has been.

Jeff (11:56.322)
You know what's interesting? I often think about this, the fact that we've used Moore's law basically to make it easier for developers to write code versus like end users benefiting from it. And end users do benefit from it because they have many more things that they can now do on computers. But like, I think about like 30 years ago, we had Microsoft Word as a word processor and you could type in it and it had spell check and all these other things, right?

And now here we are 30 years later, right? That was on my 16 megahertz 386 computer, right? So now you've got computers that are.

Jeff (12:35.894)
just let the dog finish its.

Jeff (12:42.154)
Okay, so now you've got computers that are so much, like obviously like a million times faster, but we still have a word processor. Like the word processor itself didn't get a million times faster. Like there's only so fast you can type. Yeah, it's basically the same thing that 30 years ago, but the way we deliver it is totally different. It's delivered in a web browser with like so many layers of abstraction between like the user and the computer itself that...

Auren Hoffman (12:54.659)
Yeah, or even a million times better. Yeah, yeah, it's...

Jeff (13:11.142)
All Moore's law basically in the speed and the speed of computers and networks and everything else is basically there to make it so that we can deliver advanced software with layers of abstraction that make it actually pretty easy for those software developers to go build those experiences. And, and so it's lowered the bar for what it takes to go build those experiences and it's made it more accessible to more developers. I mean, I think it's fascinating that, you know, technically speaking, I think when you, when you, when you, when the

the experts who like survey job categories. I think according to those folks, there are about 30 million software developers in the world. Yet, I believe there's more than 100 million GitHub accounts. Right, so you also have this notion of like, the citizen developer, the hobbyist developer, and those people actually not even just doing it on their own because I think it's like interesting or fun or, you know, job skill, actually building things for their work, despite the fact that technically speaking,

They're not a software developer. And I think that's amazing as well.

Auren Hoffman (14:11.035)
Well, you think of, I mean, what is really a software developer? You have all these people who like can write very sophisticated pivot tables, um, in Excel or, you know, Google spreadsheet or something like that. You have people who, you know, build Zapier type of functions, right? Which is, um, you know, the, uh, I mean, you, you bought, you guys bought, uh, a segment. They're generally selling to like a sophisticated marketer, kind of a technical marketer, not necessarily software developer.

maybe even doesn't think of themself as anyone who can code, but essentially is kind of writing code in a way. They're putting together very, very complicated building blocks, kind of very, very complicated, if then statements essentially, right? So, nowadays there's like a lot of technical people, or a lot of analysts are kind of software developers.

Jeff (15:00.406)
You know what's interesting? I think those types of folks, they've always been. It's just their main tool, in fact, almost their exclusive tool was Microsoft Excel, right? And now like, in fact, the most successful IDE in the history of software is Microsoft Excel. It's got billions of users and more has been built on top of that than any other developer tool that exists. And so, you know, the evolution of those technical users. I like to think of it this way. What Twilio does,

Auren Hoffman (15:03.803)
Yes.

Jeff (15:29.182)
is we sell technical products to technical customers. And those customers are the ones who want to go build. They want to see their competitive landscape and have a better idea than their competitor. And when they have that idea, be able to roll up their sleeves and go build it. And that can be obviously the software developers themselves. Those are the product managers. Those are the technical buyers inside of an organization, like the marketer who's like, I am not content with just whatever capability someone gave of like some batch and blast from like,

I actually want to go in and build and design journeys and analyze them and understand if they're working and add new data in to make those journeys smarter. And that is the act of building. That is a technical act with all sorts of tools that enable a wide variety of folks to be able to partake in that. And that's really the competitive landscape that we live in now today, which is companies who think that way, who see a customer problem and roll up their sleeves and go build, they're the ones who win.

Auren Hoffman (16:25.401)
15 years ago you started Twilio. A lot has changed in 15 years. What do you think are some of the biggest go to market changes when you think of selling to developers, selling to these technical buyers?

Jeff (16:39.978)
You know, I think one of the early insights that we had along with say, AWS, Stripe, and a few others was that developers would become, would become a buying center. And I remember raising money in 2008 or trying to raise money unsuccessfully in 2008. And one of the, there are two pieces of feedback I got. One was there's a financial crisis going on, stop asking us for money. But number two, I think that was more in my control was they were like, well, developers aren't on market.

Auren Hoffman (17:01.515)
Yeah.

Jeff (17:07.874)
They don't have budget, they can't make buying decisions. And so investors were like, go build an app that you can go sell to someone who's got budget. And if you want to add an API to it later, have fun, go for it. But you need to go find a buying center. And that was the conventional wisdom of software go-to-market. And what we observed was that if you take something that used to cost, I don't know, a million dollars to go build out a data center or

Auren Hoffman (17:21.777)
Yeah.

Jeff (17:36.994)
or anything like that. And you make it now available for pennies per use. You've just taken it from a large capex, therefore it's a big budgetary item, a big decision, et cetera. And you've taken it into something that goes on a credit card. Well, now you're gonna get all these developers who build upon it. And the neat thing about software and the neat thing about developers, when they build on something, that something starts out as a prototype, that's no big deal, that's on a credit card, it costs five bucks to do a proof of concept, fantastic. But when...

Developers make progress and build some experience. Guess what happens in the meeting room the next day? Everyone gets really excited. Oh my God, that's amazing. How quickly can we get that to production? And so the prototype very rapidly turns into something that is deployed in front of customers and then just grows. And so for a company like ours, the whole goal is to get in there with essentially the design win.

Auren Hoffman (18:16.391)
Yeah.

Jeff (18:31.618)
The moment when someone's making a decision about, we need this capability, aha, this is the tooling I'm gonna pull off the shelf and use. If you're there at that moment, you've got a decent chance of turning it into a scaled production use case. And really, it's just a matter of, it does that use case like the end users like it, and are they gonna use it, right? And that's something that is like a probabilities game. So you've millions of developers, some number of them every day are gonna have some new idea that they or their team are working on.

bring you in and bake you into this idea. And some percent of those are actually gonna be duds and no one cares and so they don't go anywhere. And then some percent of those are gonna be the next big thing for a company that they themselves have millions of users and that generates a ton of revenue. And I like to think of this developer market. I like to think of this developer market is, if you look at the number of developer focused companies, I think the ones that have achieved revenue scale.

like real revenue scale. They are one of three business models. Do you wanna hear them?

Auren Hoffman (19:37.017)
Yes.

Jeff (19:39.182)
So I think the three scaled developer business models are number one, business development as a service. So if a developer was.

Auren Hoffman (19:48.775)
Get you customers like Facebook or Google or something like that. Just gets you customers. Amazon gets you customers to your marketplace or something like that.

Jeff (19:58.014)
Actually, a little bit different. If a developer in the previous world was not empowered by their organization to go cut a deal with a carrier because those deals are complicated, they take time, there's revenue commitments, a developer would not have been empowered to go cut a deal with a carrier or open a bank account on behalf of the company. The developer just didn't have the power or skill set or desire actually to do those things, but now suddenly they can sign up for Twilio

Auren Hoffman (20:10.117)
I see. Yeah.

Jeff (20:26.602)
and get the result of having done business development, i.e. a text message delivered to a user, or I can go sign up for Stripe and suddenly be able to take payments. You empower the developer to actually take on a role in the company of establishing a business relationship with another entity that they wouldn't have been able to do otherwise. So that's business development as a service.

Auren Hoffman (20:29.903)
Yeah, okay got it. Yep

Auren Hoffman (20:42.363)
So, and you sure you would put, you put Stripe, you put Twilio, you put Plaid, you put like these other ones in that category. Yep.

Jeff (20:47.606)
Yep. Or, or, yeah, or like even AWS in terms of like a data center. So that's the thing that creates a buyer where there wasn't a buyer earlier. The second thing that actually does has a similar effect is, uh, Opex as a service, right? Some take something that used to be a CapEx and turn it into an Opex. So the data center or telecom hardware, right? Things like that.

And so again, you've taken something that used to be a large expenditure of a company, but it was a capex enough where there's a process that companies went through and you turn it into an op-ex and now developers can actually buy it. So that's the second.

Auren Hoffman (21:19.131)
You can even put like WeWork in that category or something like that.

Jeff (21:21.758)
Yeah, in a lot of ways, if, if a IC could go and essentially put it on their credit card and maybe they can't, I don't even know. Um, but in particular though, if it's like, it's gotta be well below the spending threshold for it to like actually work. Right. So for Twilio, a developer can go and sign up and on their credit card, build a prototype or something and literally spend a couple of dollars. Right. And that's something that again, would have been hundreds of thousands or millions of dollars in the old world.

Auren Hoffman (21:25.509)
Right.

I have no idea.

Jeff (21:50.57)
now is a few dollars, therefore it can just fly into the radar. And guess what? Instead of people getting angry about it, they're so excited. They're like, Oh my God, you were able to build that in, you know, a week. Holy moly. Let's let's get going. Right. Um, so that's the second one, a CapEx, StopEx. And the third is arguably the hardest, uh, to do. And this is the rarest, but I would say is, is a, um, essentially an algorithm as a service.

but some kind of algorithm that is so hard, that is so complicated, that the developer really can't replicate it themselves easily or even get it in terms of like open source. And so if you look at that category, it's like, you know, I think Amazon like Dynamo or Google BigQuery are kind of like that, right? Like running a scale data warehouse that can accept terabytes and terabytes of data, like that is just really hard to build. Or now you look at it and you'd say large language models.

Auren Hoffman (22:35.248)
Yep.

Jeff (22:45.13)
Right? Something like that running late, like building language models, obviously, but even running them at scale. It's like, that is a set of algorithms and capabilities that it's just so hard to replicate that your developer would be like, look, I'm not even going to try to go build that myself. That would be a fool's errand. I'm just going to use what's.

Auren Hoffman (23:01.411)
I mean, there are a lot of these like matching engines or kind of data co-op type things as well where, you know, where it might fit that algo as a service.

Jeff (23:11.854)
Yeah, but a little less developer, right? So there are, you know, there's other business models for other buyers inside the company for sure. So like, you know, if you've got a data asset that is unique, then obviously that's a great business to be in selling it to probably marketers or other folks like that, absolutely.

Auren Hoffman (23:17.744)
Yep.

Auren Hoffman (23:29.747)
Okay, interesting. I like these three. This is, and is it like, do they generally like start at number one and then move to number two? Instead of number, or how does it kind of, it can flow any of the ways.

Jeff (23:43.01)
Well, what's interesting with number one, I think number two is actually the most common, but if you think about how many developer focused companies are at say a billion of revenue or more, there's not that many, right? You've got the major infrastructures of service players, you've got Stripe, you've got Twilio, and that might be it.

Auren Hoffman (23:47.142)
Yep.

Auren Hoffman (23:55.812)
Yep.

Jeff (24:01.194)
And so when I think about the like biz dev as a service and OPEX as a service, those I think are the biggest drivers here. And so you look at it and you're like, well, what are the major categories of spend that you can essentially do that to or the major relationships a company has that you can go do that to? And you've got, okay, you got banking relationships, communications relationships, data center relationships. And I think there's opportunity out there to go look at and say, well, what other major relationships do companies have?

And can they be disrupted by turning them into as a service as opposed to, you know, a one time big deal, you know, contract signing effort.

Auren Hoffman (24:36.347)
One of the nice things, as you mentioned, of having a developer just sign up and all of a sudden it can get huge is, well, it's easy to sign up the customer. Maybe it's cheap to sign up the customer or they get big. But maybe the bug of that is it could lead to some customer concentration where, I know when Twilly went public, you guys had WhatsApp, you had Uber. It's very, very large pieces of your revenue, which I can imagine.

Um, is, uh, is, is quite difficult to, to manage some of these customer concentrations. What advice would you give to companies that have a lot of customer concentration? How do you deal with that? How do you think about that?

Jeff (25:20.418)
You know, in particular with the usage based business model, it does lead down a path to potential customer concentration because some customers are all gonna be growing at various rates based on the success of their business model. And it's a win, like if you've got customers who are growing like a weed, but it does lead you, you know, you're happy, you're happy, you're happy. And then one day you're like, oh, wait a minute, this could be a problem, right? And we've had, you know, that.

Auren Hoffman (25:44.679)
Hehehehe

Jeff (25:48.166)
for the history of Tulio, there was one point where GroupMe, if you remember that group texting app that was founded by Jared and Steve back in like 2011, I think, there was one point where they represented about 40% of our revenue for a very short period of time. And I remember, I'll tell you a quick story that hopefully Jared and Steve can laugh about at this point, but I remember one time like we were taking a walk and they're like, I don't understand it, Jeff.

Auren Hoffman (25:51.737)
Oh yeah.

Auren Hoffman (26:00.827)
Whoa.

Jeff (26:13.57)
we're such a big customer of yours, like shouldn't you be bending over backwards and like making your roadmap, everything we ask for and all this kind of stuff. And I was like, honestly guys, no. Because that actually kind of deepens the hole, right? If you will, of customer concentration. So the thing I tell my team is to take really good care of you guys. But in particular, we need to focus on building our business around you with all the rest of our customers because I can't make a bet on, you know, one customer to carry the company.

Auren Hoffman (26:19.132)
Ha ha ha.

Auren Hoffman (26:25.092)
Yeah, yeah.

Jeff (26:39.806)
And I look and I think they like, you know, a few months later, they sold it to Skype and Skype folded it in and then, you know, GroupMe is not a thing today that anyone's using. So it's like the right thing to do is to treat that customer really well, you know, as you would any customer, but do not torque your company to like serve their needs at all costs, because you can find that you, you know, you wrap your company around one customer and then something happens to that customer. Either they decide to stop using you for whatever reason or they.

Auren Hoffman (26:47.025)
Yeah.

Jeff (27:08.619)
They go out of business or their business model changes whatever and then you're just screwed and I will tell you I was just chatting with

Auren Hoffman (27:09.713)
Yep.

Jeff (27:30.478)
I was...

Jeff (27:33.642)
I was just chatting with an entrepreneur who's got a fairly well-known company with a usage-based business model who was asking me that very question. He was like, we've got a small number of customers who drive a ton of revenue and I'm worried that it's a problem. And he was like, some of my investors are like, this is great. How do you keep these guys growing? And how do you find another one? And he was like, I don't think I want this. I think I want...

Auren Hoffman (27:56.997)
Yeah.

Jeff (28:01.11)
you know, 100,000 small customers, not like five huge ones. And I was like, that is a very good instinct for you to have. And so I think that that's the best way to manage it. But this usage based model yet does lend itself to customers growing at like their rate based on their business, as opposed to growing. At your rate, based on your business, like most SaaS companies grow as a function of how many sales reps did I hire? Not as a function of how much did my customers spend on sales and marketing this month?

Auren Hoffman (28:26.448)
link

Auren Hoffman (28:30.583)
Yeah. And is it, is it like in this kind of product led growth motion where somebody signs up, you mentioned the developer, they spend five bucks or something like that and you kind of watch him at some point, like it makes sense maybe to have a human talk to them, but that could be like quite a long time before you do that. Like how do you think about like when to get like the human rep involved?

And obviously you might want the human to help them because they may have some questions and maybe your docs aren't as good and you know, et cetera. But like when you want the human to sell.

Jeff (29:05.702)
here's the thing you really have to think about. So a developer will bring you in the door and they will get it started. But at the end of the day, like those investors, I mentioned, they told me, hey, you know, they're developers on a budget, blah, blah. Like actually those investors are right. The developers bring you in, but at the end of the day, the part of the company that is paying you is someone else. And if you don't know who actually is signing your checks every month,

you're at great risk because at some point that person can wake up and say, well, I don't want to pay this anymore or why am I paying so much? And that's kind of what happened to us at Uber. We, you know, the developers brought us in and Uber was on a $60 million annualized run rate with us, 60 million, yet they didn't even really want to talk to us. You know, we'd, we'd ask the developers, Hey, you know, can we come by? And they had a sales rep assigned to them.

But you know, they kind of play, I know we're really busy, you know, no need, it's fine. And then until one day, the priorities of the company changed and they got very focused on cost savings and efficiency and all that. And, you know, the Twilio spend popped up as an area where they could, you know, focus some energy and we didn't get a heads up because we didn't have that relationship. We didn't have someone walking the walls there, walking the walls, walking the halls there, getting the heads up of like, Hey, management's really focused on efficiency. Like, can we work with you? Like.

where are places where we can save money, which we had a lot.

Auren Hoffman (30:32.407)
sounds like you tried like you had someone calling them you're calling they didn't want to meet so what would you have done differently

Jeff (30:37.442)
So here's the deal, right? What you have to do with your sales team is as the accounts start getting to be a meaningful spend, what you need to do is, it's not like it's up to the customer. If they want to talk to your salesperson, great. But if they don't, okay, well, I guess we don't have to spend any time with them. And I early on thought that it was like the customer benefited when we had a salesperson with that relationship. But if they're already spending money on us, it wasn't like I'm not necessarily trying to sell them something. So, but what I came to realize was that it's not the customer's.

goal to have that sales relationship, it's ours. Like we've got to have that relationship because you need to like be really close to that customer. You need to know everything that's going on inside the company, what their priorities are, how you align to those priorities so that you can help your customer, you know, the buyer inside the company navigate their company's priorities and make sure that your attachment things are important to them. Oh, by the way, if some sentiment changes about like

what they're spending on you or their goals change or like use cases change. Like you wanna be there to help their customer navigate it and really help yourself navigate it. And I think that was something that we screwed up in the early days of the company because we didn't work hard enough to go kind of climb the ladder and figure out who is the actual check signer here and do we have a really good relationship with that person? Yeah, we knew and we had a relationship I'd say with those people.

but it wasn't good enough given like the importance and the amount that they were.

Auren Hoffman (32:03.879)
And is it like a playbook? Okay, now we have this playbook for the customer success, or we just have to hire a different type of person who knows how to, like, how do you do that? How do you implement?

Jeff (32:17.666)
I think it was a matter of focus. And I think it's a matter of like the questions that the sales leadership is asking of those reps. Right, so today, if you had an account that was that big and the rep was like, yeah, I haven't seen them in six months, right? The sales leader would not accept that. Yeah, the sales leader would not accept that as like, at the time we probably said, well, like, and we, you know, some pushback, like, come on, seriously? They're like, yeah, they keep declining the meeting request. And I think today you'd be like,

Auren Hoffman (32:34.474)
Be code red. Yeah, yeah, okay.

Jeff (32:47.358)
I don't care. Like you need to show up with a basket of candy and force the door down and get yourself in there. You're gonna send them donuts every day until they actually let you in the door. And say, we're not gonna accept that outcome. And so I think there's, but there's a lot of mechanisms you would use. The other thing I would say is like, you really need to align the sales incentives. So if you've got a salesperson with 10 accounts and that one account really allows them to,

Auren Hoffman (32:51.539)
All right.

Jeff (33:17.122)
They don't have to do anything. So they can focus all their time on these other accounts where maybe if they did more work, it'll result in more sales. And if their incentives are aligned to that outcome, then they will be willing to say, okay, well, this account doesn't need my attention. I'm gonna make a whole bunch of my quota based on, okay, great, I'm gonna focus my attention on the ones where maybe it's not as sure of a thing. And so you need to think about the sales incentives to make sure that rep can't take for granted that big account. In fact, they have to go work it in order to make their number.

Auren Hoffman (33:24.499)
Mm-hmm.

Auren Hoffman (33:35.079)
Yep.

Jeff (33:45.886)
and prove that the work that they did resulted in revenue growth to the company.

Auren Hoffman (33:51.159)
Interesting. Now you've you guys have made a lot of some big acquisitions like segment and SendGrid and a bunch of smaller ones like Authy. What have you learned about how to do acquisitions well? Like what are the what are the kind of the learnings and if you could give advice to other people who are going to be in your shoes, what advice would you give them?

Jeff (34:12.898)
You know, first thing, of course, you always have to go into these things with eyes wide open, which is that most M&A like fails to deliver value. I think that's just a fact of the world. So in some ways you would, you would, you'd be surprised that businesses do them at all. Right. Because the track record of really good M&A is so low yet.

Auren Hoffman (34:20.911)
Yep. That's the base rate.

Auren Hoffman (34:31.503)
Well, sorry, obviously more acquisitions fail, but some deliver more than 2x. So you fail, you lose a 1x, but some deliver more than 2x. Does the average fail?

Jeff (34:45.718)
Yeah, I wouldn't be surprised if somewhere there's academic research about this. I'm sure some B-School professors made a career out of studying M&A. I don't know it off the top of my head, but the conventional wisdom is that like most M&A is, is doesn't work, right? Because cultural integrations, like there's so many ways in which it can go wrong. And so the first thing is go into it eyes wide open. And I think you're right. Like say, like, Hey, well, if it doesn't work, what's the worst thing that happens. Right. And having that in your head, it's like almost like the baseline, because when you go into these deals.

Auren Hoffman (34:51.484)
Yeah.

Auren Hoffman (35:00.337)
Yep.

Auren Hoffman (35:12.593)
Yep.

Jeff (35:14.486)
You know, most CEOs, you know, you go into it starry eyed. You go into it like, oh my God, guess all the stuff that we can do together, this is gonna be amazing. But you know, when rubber meets the road, a lot of these things are, they're so much harder than you anticipate in terms of like just literal stupid things to get in the way. Like at some point it was really hard for us to merge our Google infrastructure, like our Gmail, like our corporate Gmail. In fact, Twilio, Segment, and SendGrid, and those are the two.

Auren Hoffman (35:39.469)
Mm-hmm.

Jeff (35:43.722)
biggest acquisitions we've done. We are all using Gmail for our, you know, email calendar, right, basic stuff. And when we did the deals, we're like, okay, how do we just marriage it all together? And Google was like, yeah, you can't do it. They're like, we have consultants who figured out ways to do it, like using a bunch of obscure APIs. And they referred us to those consultants. Those consultants were like, yeah, we have a year long backlog of companies asking the same question. And it was like, are you serious? This is like,

Auren Hoffman (36:08.595)
Oh my gosh, it's crazy.

Jeff (36:12.726)
this is the best. And so literally you go stupid stuff that you did not think about on the day you signed that deal and everyone's excited and toasting like, there's gonna be this amazing future together. Oh, but it'll take us seven months so that you can see each other's calendars. You know, you didn't plan for that. And so and that's a stupid little thing yet, employees would like in it kills morale because everyone's super excited, like, let's go do it. And then like the stupid stuff cuts away at the will.

Auren Hoffman (36:19.048)
Duh.

Auren Hoffman (36:28.24)
Hehehe

Auren Hoffman (36:36.08)
Yep.

Jeff (36:39.658)
and the desire and the idea of like, yeah, we're gonna go build something valuable too. Like, yeah, it'll be a big win when I can like see my calendar again, you know? And so those are the types of things. So my advice to folks is one, go in eyes wide open. Like you're signing up for something that is a track record of failure. But you'd say the same thing about entrepreneurship. Entrepreneurship is a track record of failure with the occasional success that motivates everybody. And so you look at...

Auren Hoffman (36:46.682)
Hehehehe

Jeff (37:08.758)
successful M&A, you look at like, YouTube going into Google, I mean, that's probably one of the best M&A deals in history, I have to imagine. And you're like, that's what I'm aiming for, great. But you have to be as maniacally focused on getting that outcome as probably you were in the early days of your startup saying like, I'm gonna make it, I am gonna be the next, whatever. And be as focused and determined in terms of getting that outcome. Because there's a lot of impediments in the way of success.

Auren Hoffman (37:14.447)
Yep.

Auren Hoffman (37:37.391)
In the B2B world, in many, many cases, when we've seen acquisitions, they've changed the brand. And so it's like, it now becomes the Salesforce marketing cloud or something like that. And like, you forget the old brand. In your case, like there is still the SendGrid and Segment brand. And those brands obviously had a ton of value when you bought the company.

and you decided not to kill the brand. Like what was, was there an actual decision making process that you did throughout that to get to that place?

Jeff (38:13.29)
Well, so you're absolutely right. Most B2B businesses, the go-to-market hinges largely on the sales rep walking in the door with a brand at their back. In that case, you've got a large company with a large brand, SAP, Oracle, whatever, right? And a sales rep walking in with their Oracle business card saying, yeah, you already spent 100 million a year on us, I got something new to talk about, right?

Auren Hoffman (38:37.491)
They're selling everything. Yep.

Jeff (38:43.798)
And so it makes kind of sense. He would just very quickly strip the old brand and like just put it in the bag of the sales wrap and call it the new Oracle, whatever club. But in the new world of like the developer world, you do have this more consumer like motion at the beginning of a lot of these customer relationships, which is customers coming in, signing up in a self-service way, getting started on their credit card, yada, yada. So in that world, it actually feels a lot more like a consumer model than it does like a B2, a traditional B2B model.

at least the beginning. And so you end up with a lot of focus on like SEO and building equity in the domain and the brand and all that kind of stuff that is actually different than I think traditional B2B. And so that's why we've kept the brands of Segment and SendGrid and both those websites are still up as independent websites. Now I think over time we will reconcile these things more than we have historically. But when you look at a website, when somebody Googles for email deliverability and SendGrid is like number one.

because we've got the best product, the best SEO, the best brand in the space, you're like, do I really wanna tear that down? No, I don't think I do. I think I wanna leverage that.

Auren Hoffman (39:53.767)
When you think of the acquisitions that have worked well, there's a high correlation with them keeping the brand. If you think of the, you mentioned YouTube, Instagram, booking.com, even my acquisition with LiveRamp, they kept the brand. And there's something about what you said where like there's a sales motion there, but there's also some like cultural thing that the, putting together these two cultures is hard.

And if you all of a sudden say like, Hey, we're going to like, you know, all that old swag that you owned is now like, doesn't even work anymore. You can't even wear it. You have to wear this new swag. Um, kind of, it kind of hurts people's, uh, um, morale a bit, or do you think that has, could have anything to do with it?

Jeff (40:43.074)
You know, it's interesting. I actually think it's kind of the opposite. One of the things that we have found with all of our acquisitions is you know, you acquire and you think about like, Oh wow, here's some startup that's invested a lot in its brand and getting its employees to care about the company and its brand and all that kind of stuff and you acquire them and you want to be super sensitive. Like, Hey, we're not trying to like squash or whatever. And that's not all companies like that. I think Oracle comes in and they, you know, like they've got an army that comes in after an acquisition I've heard, and they like, you know, burn your, burn it all to the ground. Right. And, but like.

Auren Hoffman (41:05.863)
Oh yeah, they tear you up. Which they've been very successful in doing acquisitions. So it's worked for them. Yeah, yeah, yeah.

Jeff (41:13.898)
Yeah, right. Yeah. So maybe that is the way to do it, but, um, we've, we've gone in and tried to be very sensitive to the fact that, oh, okay, well, you've got a lot of pride in what you've built, so we're not going to stop on that. And what we found is actually the employees are often eager. They're like, we want to be accepted as like part of the new company. Like we are excited. Like if you did your job, they're excited about the acquisition. And so therefore they're like, we actually want to email addresses. We actually want to paint the office walls red.

instead of whatever color they are now. We actually want, and it's not out of anything like, you know, get out of like, oh, we don't like our old company or whatever, but it's more like, no, we're excited to roll up our sleeves and become part of the bigger thing. Let's do it. And actually those things that slow you down in that end up being impediments and they're demoralizing more than anything, as opposed to, you know, motivating.

Auren Hoffman (42:03.387)
Now I checked my old emails and you and I met back in 2010, which are kind of the good old days of SaaS. Like what are some of the surprising ways the industry has changed since then?

Jeff (42:14.806)
You know, I often think about those days. I can't remember, what were you doing at that time, Oren?

Auren Hoffman (42:18.615)
I was at live room. Yeah.

Jeff (42:20.338)
You want a live room. Okay. So that was already the live room. So, you know, it's interesting. You know, I just think about, I'm going to reminisce for a moment about those days and then I'll, and then I'll answer your question. I just think it's fascinating. Cause I think back to 2010 when tech was like all of this, especially mobile and all this tech was kind of small in 2010. You know, I remember Uber was an office, a few doors down from ours with like five people in it, you know, and, um, uh, Instagram was three

Auren Hoffman (42:39.769)
Yep.

Jeff (42:50.178)
people upstairs from us in our first incubator. Right? And it was like, it was just the stakes were small and it was just people exploring new ideas and things you could do like with mobile in particular. And then, and everything's gotten so big and so serious and high stakes and, you know huge market cap companies. And it's just like, it's kind of amazing how those tiny seeds back then were, you know when we moved into our new office we threw a kegger in like 2010 and like probably all of.

Auren Hoffman (42:53.651)
Crazy.

Auren Hoffman (43:16.241)
Hehehehe

Jeff (43:18.482)
the founders of these enormous companies today who were all over just drinking beer. And yeah, you were probably there, right? It was just a cool period of time when you look back at it and say, wow, it was just this infancy. But to answer your question, like how has things changed other than getting bigger, in particularly around SaaS, I think it's fascinating is that the last 15 years of SaaS in particular have really been about

Auren Hoffman (43:23.683)
I was probably there, yeah. Ha ha ha.

Jeff (43:45.998)
going after every pocket of budget in a company. And we keep slicing it narrower and narrower and narrower in terms of like, it used to be, okay, you had the head of people, so you had a work day. You had the head of sales, you had Salesforce, and you had the CFO, and so you had NetSuite. And like that was SaaS. But now you've like every job persona inside of a company, people have asked, okay, well, what's the SaaS they need? And you get really fine grained things.

Auren Hoffman (43:53.052)
Yep.

Jeff (44:11.678)
in terms of like various buyers and various budgets. But at the end of the day, like you still have to go after like big budget items inside of a company to build a big business. But I think that's a fascinating trajectory. But the problem is it creates enormous problems for the buyers because, you know, you have empowered, you know, the, the head of, uh, I'm trying to think of the most obscure part of the stack I can think of, but you know, I'm struggling to come up with it. And even if I did have it in my head, I don't want to like, I don't offend anybody, so I'm not going to say anything. But.

Auren Hoffman (44:37.557)
You don't offend anybody.

Jeff (44:41.482)
You know, when every function in the company buys a different piece of software to run as part of the company, you create all these silos. And I like to talk about like the biggest enemy to great customer experiences in most companies are the silos that exist inside of companies, right? And so when you take the natural human tendency to when we organize ourselves into teams and companies.

to therefore create silos, right? And so, you know, at some point in the earliest days of a company, you know, you've got very few people wearing many hats, like it's all very easy. Like I remember in the early days of Twilio, on any given day, I might be a salesperson, the marketer, the product manager, a developer and the office manager, right? Like that would be a typical day in the early days of a startup. Well, of course, as the company grows up, those all become independent departments of their own that then start scaling themselves. And so the, you take the...

the silos that form and then you buy software for that silo and you actually firm up the silo because now the software systems enforce this lack of collaboration and everybody promises like, oh, well, with APIs you can integrate stuff. Everybody knows that it doesn't actually happen. And so you take the natural human tendency to form silos when you organize large groups of people and then you reinforce it with technology and it creates.

Auren Hoffman (45:47.507)
Hmm.

Jeff (46:04.578)
really bad outcomes. And now companies are all trying to figure out like, how do we fix that problem? And that's what we look at is this customer engagement platform we're building as a solution to that. But I think that's what interesting is you think about the trajectory here, which is it was really good for SaaS to go figure out all these customer personas with a problem with a workflow that I can go solve and tap into their budget. And then at the end of the day though, it creates a whole new set of problems, which now have to get solved.

Auren Hoffman (46:12.742)
Yep.

Auren Hoffman (46:29.743)
And the number of vendors that let's say technology vendors that a company has since then has skyrocketed. Let's say the number of maybe the spend that a company spends is 4X, but the number of vendors has more than 10X. Um, and so there's just like, there's just like, everyone has like 17 vendors to do this one little marketing piece over here and another 17 vendors. So, you know, if you just look at your dev stack, like

You just have to like, I'm sure your own developers have all these different developers to do each thing, each little thing, each little feature sometimes can bring in a vendor in some ways. That's like great because like that one vendor is going to be the best at that one little thing. But it also can be this, a morass of different types of things coming at you, which is hard to integrate and hard to like get a better solution.

Jeff (47:19.59)
I think that's where certain standards start to emerge, which is super helpful. So for example, you remember the old Lampstack, Linux, Apache, MySQL, PHP, right? And it's like, they're all independent. Those happen to be open source technologies. But when put together, they work really well together. And so I think what happens is you do have a proliferation of options, but then natural groupings of them start to emerge as a winning combination. And then an ecosystem forms around that to make it even better.

Um, and then acronyms form and everything else, but you end up with like, okay, just put the, like, I've got a bunch of vendors. I actually can start to build a world where that makes sense together. Um, because the world kind of figures out, yeah, these things work really well. And I think that's the, that's the key thing. Like, and we see it right now with, um, for example, data warehouses, like segment and what we're doing with our data platform that is exclusively customer data.

Auren Hoffman (47:51.803)
Yep.

Jeff (48:15.17)
So we are really good at understanding customer data, real-time signals about customers that come from your website, your mobile apps, all the various SaaS systems you have that have a piece of information about that customer, whether it's your sales CRM, your marketing system, your customer support systems, all that kind of stuff. We're really good at pulling that data out of those systems and forming them into a profile of that customer that's actionable by the entire enterprise. It helps you understand that customer, the two predictions about that customer.

But what we're seeing is that, oh, the data warehouse is another area where companies are turning to, to like solve this category of problems, which is like, hey, this is a place where we can actually put all of our data, not even just customer data, data of every variety, IOT data and everything else, and put it all for later analysis. And so it turns out that, you know, the data warehouse vendors and segment go super well together. And so now you start to say, great, how can we actually collaborate even better? So we've announced partnerships with Snowflake.

Auren Hoffman (48:54.94)
Yep.

Jeff (49:10.47)
and Databricks and Redshift in just the last six months that allow us to then jointly go serve our customer better to make it easier for those customers to get value out of the fact that they're using multiple vendors. But it continues to allow us to specialize in what we do, the data warehouses to specialize in what they do, but together deliver more value to customers than if every customer had to go figure it out on their own. And I think that's the answer to how you kind of solve that.

And so the great companies in each category tend to start working together. And I think the mediocre or poor companies then start to fall off the map because they can't unlock the value for customers in the way that like the leading CDP and the leading data warehouses can join together and have a leading solution for those customers.

Auren Hoffman (49:50.483)
Now, one thing about segment I think is interesting is, just managing customer data is hard. It's also very sensitive because you're talking about data about people. So you really want to keep that. Could you, because this is a data podcast, could you imagine a scenario where like, you could have like second party data sharing, let's say segment manages the data for McDonald's and it manages the data for Starbucks, and they want to do some sort of joint collaboration with one another.

Could you say, okay, hey, well, you could give like some sort of permissioning that could happen where I could help each other out. Could you see these kinds of collaborations happening over time?

Jeff (50:28.602)
I think those things can happen. And I think we do have an opportunity to help companies to build those kinds of collaborations. However, I would obviously very carefully caveat that with, look, I'm not gonna tell in your example, Starbucks and McDonald's, obviously fictional, fictional case study here. I'm not gonna tell them you're not allowed to go do that. It's their business and it's their privacy policy and it's their customer relationships. So if they're...

Auren Hoffman (50:45.872)
Yeah, yeah.

Jeff (50:56.226)
customer relationships allow them to do that, then we can be in a position to help them do that. But I think the two things I would think about is number one, let's say, you know, we all have finite amount of development resources. And so you have to think about where's the world going? And am I gonna invest in things that are gonna be valuable five years from now? Or the software my teams are writing today, is it gonna be obsolete by the time I'm done writing it? And so you would have to ask yourself, in a world of privacy,

Do I think those types of partnerships of data sharing are gonna become more and more and more prevalent? Therefore, when I build a feature to enable it, it'll get used by a lot of companies? Or, nah, do I tend to think actually those types of data sharing relationships are going away because of all the privacy changes that have gone on in this world? Therefore, it's probably not worth me investing a huge amount of roadmap in it. And I think generally speaking, I'd say we're probably in the camp of the latter. Now we do allow companies to bring in data from data sources, whatever data sources they have access to.

Auren Hoffman (51:49.766)
Yep.

Jeff (51:55.778)
But in general, I would say I think a lot of those probably data sharing types of things are not the future. However, anonymization, like true, provable anonymization of data does enable some use cases that maybe if it was not anonymized, you wouldn't do. But with anonymous, you can actually unlock a lot of value. But again, that's kind of hypothetical at this point in terms of what we do. We don't do things like that today. But the other thing that I would say that I think people should think about.

If you're building products, you might say, do we want our product to enable this? And if the world knew this was happening, and you know, pass the New York Times test or the Wall Street Journal test. If the headline in the Wall Street Journal, you know, the next day was, you know, your company enables like shady data sharing practice, like, are you comfortable with that headline? Or would you be like, Yeah, I don't want our name associated with that. And

use that test to say if the world knew what was happening and knew that we were powering it, would we be okay with that?

Auren Hoffman (52:56.547)
Yeah, exactly. Exactly. This kind of mom test, like, is it good for our collective mom or is it something that could creep her out in some sort of way?

Jeff (53:04.394)
Well, and I'll tell you a story, which is the founder, one of the founders and CEO of segment, Peter, they early on had a pretty strict policy against only first party data. And it was because of his mom, because he was explaining to his mom how internet advertising worked. And, and when he explained the details of it, she was horrified.

Auren Hoffman (53:18.982)
Hahaha

Jeff (53:27.598)
And so that led him to decide, you know what? Like, yeah, we're not gonna support certain data sharing things, especially the kinds of stuff that were happening on the internet 10 years ago in terms of the data sharing that went on. And so that became the kind of Peter's mom rule and inside the company. Now we've loosened that up a bit since, but I actually think now you've got much greater privacy frameworks in place, the EU obviously, and even California, many jurisdictions. So now you've got some notion of governing

Auren Hoffman (53:57.266)
Yeah.

Jeff (53:57.672)
the data privacy, which 10 years ago, you just didn't have that. And so nowadays, let governments regulate it.

Auren Hoffman (54:00.667)
Well, it's much easier to, it's easier to operate today because you know what the rules were like 10 years ago, you almost were like companies were almost responsible themselves for coming up with the rules. And so you were, you were the legislator, the enforcer, the, which didn't make sense. It wasn't really fair for the companies to that today. There's, there's more standards. Like, here's what you, there's what you do. Here's how you should follow it. This is how you comply. Um, and that makes it a lot easier for companies. You just follow the rules.

Jeff (54:32.138)
Yeah, and I think people are being a bit naive. I think people have gotten accustomed to, this is how the internet works, right? Like the data is there and this notion, oh yeah, consumers will readily give up all their data for one free taco a year. And I think the people had gotten conditioned to believe that that's just how it was and that's how it was going to be, even though if you rolled the clock back 10 years before that, they probably would have been horrified by a lot of those things.

Because I remember, you know, 1999, people talked about like, very mundane data sharing things and people were horrified by what the internet could become. And the internet did become that thing everyone was afraid of. And you know, everyone was kind of asleep at the switch on it and everyone got, you know, boiled like the frog. And I think when regulators really got their hands on it, looked at it, they took action on the things that I think we all thought were bad ideas. But

Auren Hoffman (55:05.381)
Yep.

Jeff (55:26.774)
But it came to fruition anyway. And I think that's good. And I'm glad that these privacy regulations have come into place, because I don't think we like collectively where the world ended up in terms of privacy. My only ask of the various regulators of the world is, can we actually make it even more well-known and consistent? Like, can we have a federal standard for privacy in the United States instead of leaving it up to state by state? Like, I think that would be really healthy for the ecosystem.

Auren Hoffman (55:48.292)
in the US. Yep.

Auren Hoffman (55:53.743)
Yeah. And then the other thing is then, you know, can you make sure that, that the often regulations are good for the very big companies, but they're not necessarily good for the new innovative companies, um, because they're, there's, there's good things that come with the regulation, but there's also just like all this extra compliance stuff, like a sock too. There's like some good things in sock too, but there's also just like a whole bunch of process things that maybe don't really help. Um, and how do you make it both?

good for the consumers and protect the consumers while still also allowing companies to innovate, not just being good for Microsoft or some really big company that's out there.

Jeff (56:31.97)
Well, I mean, I think that the devil's in the details of a lot of these regulations. And sometimes you're right. Sometimes regulation is essentially protectionist for the existing regime. And I think there are risks of that, especially like things that cement the role of the existing, say social networks or things like that. The flip side is it's a ton of work when regulations change, existing companies have to like completely retool the roadmaps and they've got a lot of work to do given the number of processes, customers and data they have.

Auren Hoffman (56:53.765)
Yep.

Good point. Yeah.

Jeff (56:59.638)
that now have to all go change. So it's not like it's easy or a great thing for big companies either. And in many ways, I always say change is great for innovation, right? If innovation or say startups, if the root driver of like the existence of startups or the motivator for people to start companies is there's an unsolved problem in the world and I can go solve it, that's the root of innovation. Then when the world changes, you by definition have new sets of unsolved problems.

Like I saw this in the early days of COVID, right? COVID barreled down. And in the early like weeks of March 2020, when the world was shutting down, I remember there were a lot of companies that were like, you know, I'll turn it to corporate world for a second, like freezing hiring and like, and batting down the hatches. And I remember we were debating as a leadership team, should we do those things? Should we freeze hiring? Are we, you know, and I remember looking at the data in terms of like new signups, new things getting built on Twilio, and it actually was

Auren Hoffman (57:45.286)
Yeah.

Jeff (57:59.042)
growing, it was accelerating because there's a whole host of new problems in the world that developers and organizations are picking up their tools and going to solve. And I was like, I don't think we should slow down. In fact, I think that the world's about to get even more complex and more accelerated for us. And that turned out to be right. And so I always think when the world changes,

Auren Hoffman (58:04.678)
Yep.

Jeff (58:21.718)
that creates opportunities because it creates a new set of generally unsolved problems. And so regulations are a similar thing, right? If you are a startup and you can help companies actually navigate a regulatory landscape that is changing, that's an opportunity.

Auren Hoffman (58:34.483)
All right, this has been great. Last question, we ask all of our guests, what conventional wisdom or advice is generally bad advice?

Jeff (58:45.538)
You know, the internet is so full of advice. I'm gonna answer meta question first. My favorite category of blog posts, and I say that fisticiously, because I think it's funny. There's all these blog posts that are like, how to make your first sale, for example. And it's like the person starts off with like, well, we started a company and I just made our first sale, so I'm gonna teach you how to do it. I'm like, wait, we should learn from someone who's done it exactly once?

Auren Hoffman (58:47.708)
Yeah

Auren Hoffman (58:56.971)
or yeah.

Auren Hoffman (59:11.224)
Hehehehe

Auren Hoffman (59:15.39)
Hehehehe

Jeff (59:16.666)
You know, like, no, we should not learn from people who are like, also figuring out, we should learn from people who've done it, like at scale, refined it over many, many years of their career. But the internet is full of people giving advice on things that they barely know themselves. And I just find that actually hilarious. It's an interesting filter.

Auren Hoffman (59:22.472)
Yeah, yeah.

Auren Hoffman (59:33.743)
Is that actually so bad? Like, I mean, like if somebody like, if someone does something so well, like, you know, someone who like is a, the best piano player in the world, like they might not be able to teach me how to play piano, but if they just learn how to do it in the last two months, it may actually be fresh in their mind. They may actually be able to really teach me how to play because they, they kind of just went through it themselves.

Jeff (59:56.386)
Well, you know, they say like, you know, mastery is the ability to not just do a thing, but actually to explain it and teach it. Right. And so I think there is a hump, but you're right. There's probably a period of time of that, like learning curve where you are good, but not yet good enough to explain it. Right. So I think you're probably right. I just think it's funny. Cause when, when you have a, like, if, if someone was like, I just made my first sale and I can write the blog post about you, right. There's like explaining your learnings is nice. Like that, there's value in that saying, here's a bunch of stuff I just learned. And like, I'm going to share it with the world, but at the same time, when you're sample size one.

Auren Hoffman (01:00:01.971)
That's true. Yeah.

Auren Hoffman (01:00:08.528)
Yeah.

Auren Hoffman (01:00:15.401)
Hahaha

Auren Hoffman (01:00:26.221)
Totally.

Jeff (01:00:27.063)
Is it really like, are your learnings generalizable to the whole internet? You know, I'm not so sure about that, right? Wait till you have a sample size 10,000, then you can share your learnings because they probably are more applicable. But my

Auren Hoffman (01:00:38.213)
I got 33B on Southwest and that actually sat next to the buyer and that's how you do it. Just get 33B on Southwest.

Jeff (01:00:44.246)
Yeah, that's how you do it. So 33B, that's the answer to enterprise sales. But what advice is generally, what conventional wisdom, and I think is bad advice. You know, there's one that I always talk about, which is the conventional wisdom is like, focus, find a niche, get rich. Like, I think that's the phrase that is always touted as like, you know, if you wanna find an opportunity, you just have to find some niche somewhere that no one's exploited and like.

Auren Hoffman (01:01:04.848)
Yep.

Jeff (01:01:10.922)
You do that one tiny thing for a tiny part of the world really well. And like, that's the key to success. And there are a lot of examples of that, no doubt. And there's a lot of examples of that where someone started in a tiny niche and then reasonably grow and grow and grow and grow out of there. So it can work. However, I think that two things, one, if I'm going to spend my blood, sweat and tears building a company, I want to do it for maximum impact. And so.

I don't want to tell myself that my job is to go find a niche and go succeed there and then maybe expand one day. I think maybe if you're trying to increase your probability of making some money for yourself, that can be maybe a way to do it. But if you're trying to actually build a meaningful company, I'm not sure I buy into that wisdom. And the second thing I would say is this idea of focus, focus. I am a big believer in experimentation. So

I think companies should always be experimenting on the edges of what they do with new ways of doing it and even disruptive things to themselves. So this idea of focus, focus and only do one thing and do it really well. I understand why that is conventional wisdom because it's so easy to get distracted. However, if we had done that, Twilio's first product was voice, phone calls. Very quickly, customers pointed to us this opportunity in text messaging.

Auren Hoffman (01:02:26.224)
Yeah.

Jeff (01:02:31.178)
And if we had said, no, focus, voice calls, voice calls, voice calls, like we would have missed the enormous opportunity in the market, uh, to help companies actually build relationships over text message, which has turned out to be now a much larger product than voice, given that voice has declined in importance and text messages have taken off. So if we had said focus, and by the way, voice was growing nicely at that point in time in 2010, I think when we launched SMS, voice was growing really nicely. We've been easy to say, look, I get it. There's another opportunity over there, but

we're going to focus and we're going to do voice better than anyone else. I think we would be a much smaller company today and a much less successful one. And so our, our willingness to say, yeah, let's go experiment. Let's follow our customers and where they're taking us. Let's go understand the next thing we can do for them and follow, take their lead. Really has served us well in terms of multiple product expansions we've done through the years. In fact, even acquiring segment, which we did in 2020.

Auren Hoffman (01:02:59.825)
Yeah.

Jeff (01:03:28.714)
And now we see here in 2023 and everyone is talking about CDPs. We were way ahead of the curve. Why? Because we were following our customers and customers were painting this picture of a big problem they needed solved that required their communications applications and their data to get married so they could do all their communications better. And that again, that was customers painting the way for us. I think the wisdom that I would give folks is follow your customers. And the thing you should be religious about is not like

Auren Hoffman (01:03:32.535)
Yeah, yeah.

Jeff (01:03:57.266)
only do one thing or focus, focus. It is actually the thing you should be religious about is really understanding your customer's problems so that you can actually build a company and a roadmap that is like two steps ahead of the industry. And that's how you're gonna differentiate and that's how you're gonna do a better job because you have a better handle on your customers and their problems than the average company does.

Auren Hoffman (01:04:20.659)
That has been awesome. Thank you, Jeff Lawson, for joining us at World of Deaths. I follow you on LinkedIn. I follow you on Twitter. I definitely encourage our listeners to engage you there. This has been awesome.

Jeff (01:04:30.594)
Thank you, Arron, appreciate it.

Auren Hoffman (01:04:33.135)
All right. Thank you again.