6-Step Engineering Approach to SaaS Content Marketing

Omer (00:11.600)
Welcome to another episode of the SaaS Podcast.
I'm your host Ober Khan and this is the show where I interview proven founders and industry experts who share their stories, strategies and insights to help you build, launch and grow your SaaS business.
In this episode I talked to Martin Gontovnikas, the VP of Marketing and Growth at Auth0.
Auth0 is a platform that makes it easier for developers to implement authentication and authorization for web and mobile products.
Martin, or Gonto as he's more commonly known as, started coding at the age of 12.
He was a software engineer most of his career, but then a few years ago decided to move into a marketing role, or what he calls the dark side.
At Auth0, he developed a six step engineering approach to marketing.
Using that framework, Auth0 has grown from $200,000 a year in revenue to to an eight figure business in less than five years.
In this episode, Gonto shares that six step engineering framework with me.
We discuss each step in detail and walk through a real life case study from Auth0.
It's a great interview with a ton of really useful insights.
If you're looking for a more methodical way to do your marketing, then this episode is for you and I hope you enjoy it.
Gonto, welcome to the show.

Martin Gontovnikas (01:48.520)
Thank you very much for inviting.

Omer (01:50.440)
I always like to ask my guests what, what drives or motivates them.
Get inside their brains a little bit.
So is there maybe a quote that you can share with us or maybe just in your own words, like tell us like what drives or inspires you?

Martin Gontovnikas (02:04.600)
Yeah, there's a quote that I love.
It's called.
It's be yourself because everybody else is taken.
It's from Oscar Wilde.
And the main reason why I like it is because at first, a lot of times I was afraid to be myself or I was insecure or having imposter syndrome or however you want to call it.
And I think that if you are your true self and you're genuine, you're going to be much better than everybody else.
And that's what I try to help my team with as well because I think everybody mostly has the same problems and we're all in this together.

Omer (02:34.410)
Dude, I love that quote.
And funnily enough, my wife used that same quote when she was talking to mom to our son a few days ago.
Oh wow.

Martin Gontovnikas (02:46.890)
What did she use it for?

Omer (02:48.250)
You know, it's kind of started the new school year and you know, he was trying to do something that he didn't feel comfortable with, but he felt like he should be doing.
I can't remember exactly what it was.
And she just said that to him.
She just said, you know, be yourself, because everybody else has taken.

Martin Gontovnikas (03:03.290)
And I was like, yeah, I definitely love the phrase.
And I think that a lot of people are afraid of being themselves and that blocks them or make them feel insecure.
I think that people always stays in their brain assuming a lot of things, and I think it's usually better to just say what you want or just ask questions and that's it.
I think that life and things that happen in general are simple.
We just complicate them.

Omer (03:26.460)
Okay, so let's talk a little bit about auth0.
Can you tell us a little bit more about the company?
Like what?
What problem are you guys trying to solve?
Who are your target customers?

Martin Gontovnikas (03:37.940)
Yeah, sure.
So basically, Auth0 helps you with your authentication and authorization needs for your applications.
Out zero basically provides thousands of enterprise customers with a universal identity platform for their web, mobile, IOT and legacy applications.
It's very developer focused on the fact that it's for when somebody's building a custom application.
And we basically have a lot of SDKs that it's very easy to integrate into our own applications.
And we also have a very extensive platform because we know that every need is different.
So they can actually customize how authentication pipeline works with code.
And the idea is that Auth0's extensible platform seamlessly authenticates and secures more than 1.5 billion logins a month.
And our mantra right now is loved by developers and trusted by global enterprises.

Omer (04:29.380)
Got it.
And so is it fair to think about auth0 as you kind of abstract out some of those other protocols like OAuth and stuff, and you kind of provide an easier way for developers to use those technologies?

Martin Gontovnikas (04:43.780)
Yes, exactly.
So now the most used one is OpenID Connect, but there's OAuth, SAML, et cetera.
And all of them are hard.
And it's even harder if you want to implement something that's enterprise, like Active Directory, saml, adfs, or something like that.
So the idea is that with an SDK and a couple lines of code, you can just implement it yourself without having to worry about.
Our main idea is you don't even have to understand exactly the protocol to use Auth0, because if it's not that simple, then we're not doing our job.

Omer (05:12.870)
You talked about enterprise customers.
Is that sort of the main focus?
Like, if somebody is kind of, kind of at an early stage building a SaaS business would auth0 be something that would potentially be a fit for them or do you guys kind of position the product sort of differently for different types of customers?

Martin Gontovnikas (05:33.380)
No, it would definitely fit for them as well.
We actually have both self service plans as well as field deals where a sales guy will talk to you.
And for the self service plans you can just start.
You have the free trial on the website and you pay with the credit card actually using Stripe and it can cost from $13 a month to whatever the number is.
So it actually the idea is that we fit it all and that you pay us for the value that you're getting.
So it's not just for enterprise customers, it's I would say for both and it's for any developers.
Something that actually has happened a lot to us is that we have developers who do a pet project in their own free time and then two years later they join a company.
They're like, hey, I try out Zido, it's awesome, let's use it.
And then they talk to us.
Actually most of our revenue, 80 to 85% of our revenue is inbound and it's through the developer.
Love that they sometimes write small and then convince the company to use us.

Omer (06:30.600)
That's awesome.
So tell us a little bit about your role because you have an interesting background.
But let's start with what you currently do at Auth0.

Martin Gontovnikas (06:40.040)
Yeah, so I'm the VP of marketing and growth at Auth0.
So I basically run most of marketing and growth.
We have a lot of different teams.
We have like seven teams.
And as I was saying, we're a very marketing driven company just because most of the revenue comes through the marketing effort.
So that's why I think it's a challenge and it's definitely interesting.

Omer (07:02.040)
So the company was founded I believe in 2014.

Martin Gontovnikas (07:09.640)
Yeah, 2013 actually.

Omer (07:11.560)
And the two founders are from Argentina, but the company is headquartered here in Washington state.
In Bellevue.

Martin Gontovnikas (07:24.200)
Yeah, exactly.
One of the founders used to work at Microsoft and because of that the headquarters are in Seattle.
However we do have an office in Argentina as well as now we also have in London, Sydney and we're opening now one soon in Japan.

Omer (07:41.560)
Now tell me about your background because you weren't always doing marketing.

Martin Gontovnikas (07:46.990)
No, yeah, I was actually a systems engineer.
I started as a developer.
I actually started coding when I was 12 mostly because my uncle was a developer and he started teaching me.
So yeah, I always joke and I always say that I moved to the dark side now because I moved to marketing and all of my friends that are still developers are like marketing.
Oh, you're like all the smoke and mirrors and blah, blah, blah.
But I'm a big fan of it.

Omer (08:15.640)
How did you make that transition?
Because you were a developer for quite a long time, from what I understand.

Martin Gontovnikas (08:24.200)
Yeah, exactly.
So I was a developer for a long time, and I actually like doing a lot of open source.
At one point, I was doing AngularJS and I did a few repositories with Angular that got popular.
One of them was called restangular, and I also had a few others.
And because of that, I started to get invited to conferences.
And then what happened to me is actually that I broke up with my previous girlfriend and I had an existential life crisis of like, hey, do I want to continue coding?
And then what I realized is that it was actually more fun for me making restangular popular, because we always think that open source just becomes popular.
But for me, it was.
I wrote a lot of blog posts.
I went to the Angular conferences, and then in there I was having speaker dinner, meeting the speakers, et cetera, who are the Angular corcometers, and they were promoting restangular.
I was also going daily to Stack Overflow and checking if there were rest or restful APIs questions in there and answering them and telling them, hey, I'm the writer of restangular.
Maybe you like it.
It's about this.
So it was a lot of effort.
And then I realized that that's what I wanted to do more.
And I remember that I stayed up all night one night, and it was like, one what exactly do I want to do?
And I remember there was a guy called James Ward that he worked back then in type Safe, working on Scalia.
And I remember I liked what he was doing, what he was tweeting about, he was speaking at conferences, writing blog posts, et cetera.
So I looked for him in LinkedIn.
I saw he was a developer evangelist.
Then I went online.
I found a book about what is a developer evangelist?
By Christian.
Read it in six hours, all night.
And then I said, okay, this is what I want to do.
And.
And developer evangelist is the development part of marketing, because it's still marketing.
And that's how I started to move into the dark side.

Omer (10:15.950)
So break that down for us.
For people who are not familiar with that, what does developer evangelism mean?

Martin Gontovnikas (10:21.390)
Yeah, so developer evangelism, basically what they do is they help other developers be successful.
So when you have a product that is developer focused or they have an API, what they want to do is press the word so what I was doing is writing blog posts and content for that, going to conferences to speak about the product.
In my case, about zero again, I was writing the SDKs, writing some of the docs, writing some of the quickstart.
So it's all about enabling other developers to be successful in a way that it's marketing, but not marketing, because I always think that we developers have a very high bullshit detector.
I don't like when I feel I'm being sold something.

Omer (11:00.160)
And so how did you land at auth0?

Martin Gontovnikas (11:03.440)
So for auth0, I realized I wanted to look for something and I was actually interviewing with Mozila back then as well.
But then one of the advisors of Auth0 is called Guillermo Rauch.
He's the creator of Socket IO and now he's actually the founder of Sites and now and all of that stuff.
And he's a friend of mine and he said like, hey, there's a company that's starting, I think an advisor.
There are like five, six people.
Why don't you talk to them?
And I started interviewing with them actually for programming when I had this middle existential crisis.
And then I remember sending an email to the cto, to Matias, saying like, hey, I know I've been doing the interview.
I don't want to be a developer anymore.
What I want to do is developer evangelist.
So if you're looking for one, that's awesome.
If not, I'll continue looking somewhere else.
And they were actually looking for one.
And that's how I ended up in Alt zero.
So I think there was a lot of serendipity.

Omer (11:56.920)
Yeah, that's an interesting story.
And so, you know, one of the things that I was kind of really intrigued by when I was kind of doing research on you and kind of looking at your website, which is at G to you, you kind of talked about this, an engineering approach to marketing and the rise of the ctmo.
What does.
What does CTMO mean?

Martin Gontovnikas (12:20.760)
So to me, CTMO is the chief of the Technological Marketing Office.

Omer (12:25.860)
Where does that term come from?

Martin Gontovnikas (12:28.820)
I made it up, so I don't know where it came from.
I think it was just an idea in my brain.

Omer (12:34.900)
That's awesome.
So on your site, you kind of laid out a methodology for sort of an engineering approach to marketing.
And I'll kind of just read the introduction in this.
In terms of, you know, marketing is all about sending the right message to the right audience to get an expected response.
Even though this sounds easy, it's certainly not an engineering approach.
Approach lets you Decompose this big problem into smaller pieces that you can tackle by running continuous experiments that are validated by data.
And you kind of laid out like a six step methodology for sort of approaching marketing.
And I thought it would be really interesting for us to sort of deep dive into this and learn from you in terms of, you know, what is this methodology, what are these six steps and maybe kind of walk through some examples so people who are listening to this can kind of understand the methodology and if it kind of, you know, resonates with them, maybe they can go and try it with what they're doing and, and maybe have a maybe a maybe different sort of methodology.
Methodology to sort of approach their marketing.
So does that sound good?

Martin Gontovnikas (13:47.450)
Yeah, that makes sense.
Some background.
I think that in general before marketing was very similar to what people can see in Mad Men.
It was like whiskey and cigars for me.
So you had a whiskey, cigar, you were drinking.
It's like, this is the message I want to do.
This is what we're doing now.
And it was very gut feeling or let's try something and that's it.
And I think that now we have so much data, like Harari speaks about that in sapiens and Homo deus about how there's so much data right now that why shouldn't we use it in marketing?
So that's why I started thinking of this.
And mostly actually I have to say that I started using an engineering model for marketing because I knew nothing about marketing.
So I was given the opportunity to run marketing at AL0 and I have no idea of what to do.
So because of that I said, let's try an engineering approach and a process because maybe with a process I can get better ideas.
So talking a little bit about the process, the first part is about decompose.
It's decomposing your problem into smaller pieces.
And when I started running marketing in Alcedo, the biggest problem we had is that we were flat on signups.
We were having very small signups every month.
It wasn't growing, we didn't know what it was.
And that problem seems very scary and you don't know exactly what to do.
So I need to decompose the problem.
The first thing that I thought is how can I decompose the problem?
To me, decomposing the problem is around understanding the problem more.
So what I started doing, even though I was a developer, I started doing some qualitative interviews on developers around what do they do?
What are their habits, why?
So I just found some of the signals that we got in Alzero some other people as well and start doing calls and asking them, hey, what happens when you have an authentication problem?
Have you ever had any of those?
What have you tried?
What do you do when you get blocked?
And a lot of them were saying like, hey, I search on Google, okay, and what do you search in Google for?
And they were giving me specific things that they search.
I was also asking them about what do they want to learn, how do they search for SaaS or for technologies.
And also talking to them about what do they do during the day, like when do they use the computer, when do they use Twitter, why, how, what are their habits?
And just trying to understand a little bit more about them.
By doing that, I started understanding a little bit more about why we weren't growing on signups.
And it had to do with the fact that we weren't reaching to them in the right way.
Back then we were doing some ads on Google and Facebook by talking to them.
What I realized is they all have adblock, they all have ghostery, so of course they are not going to see the ads.
However, they were saying, for example, that they do use Twitter because they like learning things using Twitter, using the app.
And for that maybe we thought, okay, maybe we can do ads for that.
Other things they were saying is that most of the time they spend it googling when they have problems.
So then maybe that could be another thing that we can solve.
So we basically started separating this into different stuff and different problems that we saw and different things that we could try to experiment with something.
And the first one that I focused with, because it was the one that was repeated the most, was when I have a problem, I Google about the problem.
So that moves us to the second step, which is, okay, let's formulate a hypothesis for each.
And my hypothesis for this case was, okay, if every time they have an authentication problem, they Google, what if we write content for exactly what they are looking for?
I was seeing that the competition was creating content for single sign on, which is like another definition of authentication.
But when I was talking to developers, none of them said single sign on.
They were saying authentication controller, angular react redux authentication, I'm blocked, or something like that.
So we said, okay, let's start formulating a hypothesis that maybe if we write content for the blog for those specific searches, so then we're going to start getting more people to use the product and use auth0.
So then it was around, okay, let's define a metric, a time frame and a goal for this.
So what we said is, okay, let's focus first on AngularJS.
Let's write three articles during the span of one month around how you can implement authentication in AngularJS using Auth0.
And each of the blog posts would focus on different parts and different searches that I was getting from the developers.
The metric was going to be signups because the problem was that we were flat on signups.
So what I wanted to see is an increase on signups.
However, the secondary metrics what I picked was page views.
So how many people are seeing this page on their blog and what is the conversion rate?
So how many people who see this page eventually convert to a signup?
And then the goal was actually just increasing a little bit based on the baseline that we had from other articles and the fact that we were blocked.
Once that was defined, the idea was, okay, let's run the experiment.
Running the experiment means writing the three pieces of content, waiting the month to see if the signups come up, and then verify.
So what we decided, as I said, was we wrote three articles about how to implement authentication in AngularJS using Auth0 and for three different use cases.
So we waited and then the next step was, okay, let's now verify the data.
So we checked and actually, experiment failed.
It was like a huge failure.
But that's why verifying data is important, because I think that an experiment that only fails if you don't check at the data and you don't try to think about what this means, if you do that, then it doesn't fail, even if the numbers doesn't match.

Omer (19:57.750)
How did this fail in terms of.
The whole thing failed or you weren't getting enough?

Martin Gontovnikas (20:07.670)
We weren't getting signups from the article, we weren't getting signups.
So we didn't understand exactly what was going on.
And that's why we digged a little bit deeper.
And what we saw there is that we were getting actually a lot of page views, but the conversion rate to sign up was very bad.
So we digged even deeper into page views.
The problem was that the page views came from people who already were out 0 users.
So they already knew of the product and they were trying to find better ways because they were blocked with authentication in Angular and they were trying to understand, hey, I'm using auth0, I'm blocked, what do I do?
So based off of that, we decided to actually iterate, which I think is the next and final step of this process.
And the iteration was, okay, this means that the people who are seeing the vlog are already out 0 users, but we did increase the page view.
So content does work for developers.
So let's try another type of content.
And what we decided to try was what's called greenfield content.
So instead of writing about how to implement authentication with Auth0 in Angular, we said, next experiment is let's write about three use cases about how to add authentication in general in Angular and then have a small aside about, hey, if you don't want to do this, you can use auth0.
That was the next experiment that we tried.
And that was basically, again, we needed to formulate, define, run, verify, and then continually to iterate over those.
So when we did greenfield content, I will fast forward.
And what happened was that we had very few page views, but we had a very high conversion rate.
So signups, it still failed because signups grew, but only a little.
But what we checked was conversion rate was great, we had very few page views.
So the problem was that SEO wasn't enough.
So then the next iteration was, okay, what if we start sharing this?
So we started creating a distribution list and start sharing this on hacker news, Reddit, JavaScript weekly, Ruby weekly, etc.
And once we started doing that, the third experiment, it finally worked because now we were having more page views, still a very good conversion rate, and we were increasing signups.
Once that worked, we said, okay, what other problem?
What other thing can we do?
Can we iterate and make this better, or should we focus on another part, another piece of the problem, which, for example was instead of doing ads with Google and Facebook, let's try them on Twitter.
And that's how we try to do this more and more.
And that's actually how we built the marketing team in general.
What I always think about this is we have a team that is always doing experimentation.
They experiment, they experiment, they experiment, they experiment.
When something starts working, we spin up a team for that because that works.
So let's add more.
Our first hires were technical writers.
Now we have a team of 10 technical writers that are continually writing content for and trying new ideas and trying new experiments.
And then once that works and we have a new team for that, it's like, okay, let's try another experiment with another experimentation team until it works.
And then again, let's spin up another team.
And that's the process that we actually still use at alt zero.

Omer (23:12.820)
To summarize the six steps, number one is to decompose, which is what you really talked about in terms of breaking down the problem into smaller pieces.
Number two was to Formulate, which is formulate a hypothesis for each of those problems.
Third step is define, which is define a metric and a time frame and a goal for whatever experiment you're going to run.
Step four is run the actual experiment.
Five is go back and verify the data.
Did you actually prove or disprove your hypothesis?
What does the data tell you?
What new insights do you have?
And then step six is iterate and pick your next problem.
And kind of the example you described, you kind of basically went through that process a few times before you were able to get the desired result in terms of increasing signups.

Martin Gontovnikas (24:18.690)
Yeah, exactly.
And I think that sometimes a lot of people get frustrated because they try something and it doesn't work.
And I think that most things don't work the first time you do them.
They never do.
So it's about being antifragile and continuing to work on that and iterate and iterate and iterate until you either find that definitive results that it's not working, or it starts working somehow.

Omer (24:41.890)
Yeah, I'd love to dive into this a little bit more.
So when you talk about, like step one, decomposing, what.
What's the process that you or your team go through to try and decompose a problem?
Like, what are some of the questions that you ask yourselves to kind of break that problem down?

Martin Gontovnikas (25:04.220)
So in general, we try, I try to.
We try to divide this into two different parts, like qualitative research and quantitative research.
So every time there's a problem, and we know that signups are not working, people are not using the product or something like that, we need to start focusing on what specific things are breaking in the middle.
So one part of the analysis we do is quantitative analysis.
So, for example, we write all of the steps that somebody does to sign up, so they go to a certain page view.
Then what is the conversion rate to sign up?
They sign up.
What is the conversion rate to using X feature, how much time, what happens next, et cetera, to start finding exactly where it's broken in the middle, and we can find something that is easier to tackle rather than a very generic problem.
And then on the qualitative side, it will basically depend on the problem.
But I think it's asking who is your target market?
And people who have both succeeded and failed using your product and.
And asking questions depending on what it is.
So for us, for example, if it was top of funnel, like I was saying before, for getting signups, it was about how do you search for technologies?
What do you like learning what happens if you get blocked with authentication.
How do you look for solutions?
What are your habits in your day to day?
Because I think marketing is basically tapping into people's habits.
And then, however, another problem we had was people were signing up and then they weren't able to get their application running using Altito.
So some of the questions there were both for people who didn't get it, about what happened, where did you get lost?
Did you find this documentation?
Did you read it?
Did you not read it?
So it's just like this idea of detective work of trying to understand in a qualitative way what happened.
And then when we do a lot of qualitative interviews and we find that a lot of people mention the same thing, we try to convert that to a quantitative problem.
So for example, for the second one that I was saying, we have a lot of people that were saying that they just weren't able to implement out zero because they didn't read the docs because they were too long, they didn't have bullets or whatever.
So we said, okay, how can we convert this to a quantitative reason?
So what we searched was, for example, how many people bounce, how many people don't scroll, how many people move very fast between scrolls?
And those were the three metrics that we said, okay, this is the specific problem that we need to figure out and work on right now.

Omer (27:37.640)
Got it.
Okay, totally makes sense.
Okay.
And then so the formulate part, in terms of formulating a hypothesis, if, if anybody has done any kind of experimentation or AB testing, they'll probably be familiar with this.
But for people who aren't, so a hypothesis would probably be something like, for the example you described might be

Martin Gontovnikas (28:03.550)
more

Omer (28:03.870)
people will sign up to use auth0 if we can provide content that they're searching for on Google.

Martin Gontovnikas (28:10.430)
Exactly.
That's exactly right.
And I think it's like one part like what is the problem statement?
And then it's what will happen then.
And what you said is exactly right, is more people will sign up.
That's like, what is the problem that we're trying to fix if we write content that helps them?
And that's the then.
So I think a hypothesis always have these two parts.

Omer (28:32.630)
Yeah, that's great.
Yeah, thanks for clarifying that.
And then the third step is just in terms of defining a metric and a time frame and a goal.
What is the process that you go through to figure those out?
Like if somebody is saying, okay, yeah, this sounds good, I want to try and implement this myself, what kind of guidance would you give them to when they think about their own metrics or time frames and goals.

Martin Gontovnikas (28:59.760)
So I think that first of all, sometimes the problem already has the metric.
Like we need more signups.
That makes sense.
It's just a sign up the metric.
However, if it is something like people are not getting the product set up, maybe we need to find something more specific.
But I think a lot of times it's either in the problem or, or based on qualitative research.
We need to dig deeper into the funnel and getting analytics on the product or on the page views to see exactly which one of them means what we think the problem is.
But I think the metric is something that is, that will always go up if we fix it or not go up, but it will go to the goal that we want if we fix the problem.
And the other thing we need to do to set a goal is once we have the metric, we need to find the baseline.
The idea of the baseline is how is it working?
Now that is not working.
So what is the number that means it's not working with what number would we think that it is working?
So for signups, for example, for us we have a funnel which was like, okay, we have X people sign up.
X people will use the product for free.
X percentage of them will convert to paid.
This is the end number that we want to get for revenue every month.
So, so then we went back and we said, okay, for getting $100 more every month, we need to get this number of signups every month.
So that's the goal that we have.
And then for time frame, this is actually the hardest one.
At first I remember for blog posts we tried two weeks and then it wasn't working out.
And then we saw that articles that were older were still getting signups and maybe they were increasing.
So we said, okay, we need to increase, increase the time frame.
But I think that that in order to get the right time frame, it does require a lot of times, either a lot of trial and error or something else that I do is I try to talk to other companies and other people to understand what are their time frames and see if they fit to us.

Omer (30:56.900)
Yeah, and do you kind of try to sort of run experiments time frame wise where you can kind of get some sort of know, statistically significant result or kind of is it just based on, you know, we're going to try and do this for like a 30 day sprint and we expect this kind of result at the end or is it a combination of both?

Martin Gontovnikas (31:18.430)
That's a very good question.
So to be honest, in the beginning, for the examples I was giving, I didn't even know that we.
I had to do something out that had statistical significance.
So in the beginning it was only a time frame.
However, now we do check for numbers that need to have a statistical significance and for that now what we use is there are different ways for measuring that.
There is one that is more accurate, that is the Bajesian way, using Baesches or I don't know how to pronounce it in English, who is a very known statistian.
And we actually use that.
And with that now we can calculate the time frame.
So for example, we're now running an A B testing in the dashboard where we're giving people a better onboarding.
And based on how many people now come to the dashboard and see the experiments, we know that it's going to take us 17 days to know if there is a difference of at least 10% better or worse.
And that is what we are looking for.
So that means that we only need to run the test for 17 days and then cut it.
And then if we see that there is no result, that means that we need to iterate.
But now we actually use that.
That was the right question.

Omer (32:31.600)
Got it.
And there are a ton of calculators out there that can help you kind of figure out this sort of information, right?
In terms of statistical significance, yes.

Martin Gontovnikas (32:41.960)
There's one that I actually like a lot that's called the A B test sample size calculator from Optimizely that is actually pretty good.
However, it's based on a simpler metric around like what is the sample size that you need?
The one that we are working on, that we are using is basin.
As I was saying, we're actually planning to open source it in the upcoming months and share it with the rest because I think it's something that's awesome.
All of these tools will always tell you what is the sample size that you need.
And then the only thing you need to do is if this is the amount of people I get every month or every day, how many days or how many months do I need to wait so that I get to the same precise that I need.

Omer (33:23.470)
Right, Totally.
And then the fifth step, verify the data.
And I guess primarily you're looking at in terms of what the goal was, what the metric, what the primary metric that you set and how you did against that.
But if the experiment isn't considered a success, like what are some of the things that you try to look under the sort of surface in terms of the data that you have, I think

Martin Gontovnikas (33:52.050)
it's decomposing the metrics that you picked by different segments.
So what we always do for that is let's say a signup didn't work and then we say, okay, didn't sign ups from, I don't know, a particular industry didn't work, or sign ups that came from a particular source, or signups that didn't from a particular landing page.
Because when you start segmenting them, maybe what you see is that this works for a particular segment but not for others and that can inform you for the next experiment.
So I think it's always key to start thinking of that.
To give you an example, the one that I mentioned in the beginning, I was saying that we did get a lot of page views, but then we didn't get any signups.
So we dig deeper into page views and we segmented them by source, by landing, by returning user.
And then what happened is that all of the page views that we have were from returning users.
And that's how we knew that these people had an Alcedo account before.
And that's what informed us for the next experiment.

Omer (34:50.470)
Got it.
Awesome.
That's a great kind of methodology.
Very similar to anyone, as I said, in terms of anyone who might be doing a B testing kind of thing, maybe for conversion rate optimization or something like that.
But I love the way that you've kind of applied this to the whole marketing approach and kind of the way the whole team works.
I'm curious, I mean, maybe it wasn't so much of an issue because you started in the very early stages of the business, but since then, like, have you hired people from kind of more, let's call it traditional marketing backgrounds?
And if so, like, how have they kind of responded to this sort of approach into marketing?

Martin Gontovnikas (35:36.080)
So I think that in general, sometimes it's hard to get people that have the more traditional approach to go into experimentation.
What I saw that worked with that is actually showing them how experimentation can help them be better.
However, I actually do like that we have traditional marketers in the team because I think diversity is key.
If everybody thought the same as me, like we couldn't have any new ideas.
So I think it's key that we can actually work together and talk to each other and have a more diverse team to start getting better ideas, to experiment.
Because even if they don't accept the process, maybe at first, they still have very good ideas that come from our background that now we can test.
To give you an example, I remember I had one person from product marketing that they said, no, it's all about storytelling.
So this is the story that we should write.
And it's called going to work.
And they actually had really good ideas about what was the study that we needed to do.
But I wanted to test them out.
So what we ended up doing is actually ab testings on different patients with having the study we had versus the study they were proposing.
And it works.
Their story was much better.
I was just not comfortable with going forward with that without experimentation.
But then by continually experimenting on what they were recommending, we were actually able to even improve the message that they were saying even more than they thought.
And that's when they started accepting that this process is a very good process to use.

Omer (37:05.750)
Yeah, yeah, totally.
I mean, one thing I think about, with maybe people thinking about this approach who are not used to doing this, it sounds like a lot of work in terms of analytics and data and, you know, just spending a lot of time figuring out the data before you can even actually run an experiment.
And in terms of when you were in the early days, how much of an issue was that for you guys?

Martin Gontovnikas (37:33.370)
It was a little bit hard, to be honest.
I think it's hard sometimes to sell the idea that this will take a little bit more time.
But there's another phrase that I love that is that to go faster, sometimes you need to go slower first.
And the idea is that if you don't have a framework, I don't think you can continually improve because then if you do something fast and it works, but you don't have a framework or a reason on why it worked, you can't reproduce it.
To me, using this framework is all about reproduce, it's about reproducibility.
So if I understand the process and the process works, I can continue to use the process to improve ideas that I have.
However, the other way I cannot.
What actually helps, to be honest, is that both the co founders are engineers.
So I think that that made them appreciate a little bit more of the process.
But it's a tough sell and I do think that you need to get buy in from the rest before doing it.
Another way that I used is actually showing that other companies were doing the same.
Once you can start, I think that similar to case studies, like when you see that somebody else is doing and they are doing great, then maybe you can do great as well.
So I think that showing them that, for example, Twilio was using this sort of process, Mongo, even Facebook, Facebook has their growth team and they have this idea of you need seven friends in three days to succeed and then if you do that, you will continue.
They iterated a lot to get to that number.
So once you show that big companies are doing that, that can help you get buy in in the company.

Omer (39:04.260)
So how has that sort of played out?
Like what numbers can you share with us about Auth0 in terms of the size of the company?

Martin Gontovnikas (39:12.510)
Yeah, sure.
So basically when we started when it was flat, it was 1500 signups.
Now we are actually more than 10x that number of signups every month and we were able to grow them little by little every year.
I can't share the revenue number, but we are now on eight figures revenue.
And when I started I think we were only at like 200k or something like that.
Yeah, exactly a year.
So I think that something like this, this model definitely works.
And I was saying Alcedo is mostly marketing focused and most of the revenue comes from marketing.
And I don't think we could have done this if we wouldn't have used this process.

Omer (40:03.030)
Yeah, no, that's awesome.
I mean, especially to go in that sort of period of time to an eight figure business, you know, it's clearly working for you guys.

Martin Gontovnikas (40:13.590)
Yeah, I'm excited about it.
And what I like is that considering we're growing fast, the challenges and the problems we have every year, every month, every quarter, they change so often that it still makes it fun because it's a challenge, right?

Omer (40:28.800)
Yeah.
The only constant is change, right?

Martin Gontovnikas (40:32.280)
Exactly.
100%.
Right.

Omer (40:37.000)
Awesome.
Okay, great.
So it's time for our lightning round.
I'm going to ask you seven questions.
Just kind of answer them as quickly as you can.
You ready?

Martin Gontovnikas (40:47.880)
Yeah, let's do it.

Omer (40:49.880)
Okay.
What's the best piece of business advice you've ever received?

Martin Gontovnikas (40:54.850)
I think one of the best ones is around focus.
Even though we're doing experimentation, in the beginning we were doing so many experiments at the same time that it was great to start building up and creating the company.
But then as we started to grow a little bit, I think that having focus on specific areas to make experiments can make you improve much faster.
Because as I was saying, experimentation is all around iteration.
So if you can iterate faster, you're going to get better faster.
And to iterate faster, you need to focus and choose exactly what area or what are you experimenting on?

Omer (41:32.960)
What book would you recommend to our audience and why?

Martin Gontovnikas (41:36.240)
So I'm a big fan of two books.
One of them is Thinking Fast and Slow.
It talks about how the brain works, how we have System one System two.
And it talks a lot about how we make decisions and why.
I think that if you're working in a company and targeting human beings, it's very interesting to think about how they think, to be able to send the correct message to them.
And then the other one that is one of my favorites now is Homo deus from Harari.
I think it makes us be reflective of where we are going.
I feel that Harari puts himself in the position of a prophet where he predicts that future.
So it doesn't happen, but it makes you think a lot about where are we going.
And it's very philosophical books.
I'm a big fan of philosophy in general.
That's why I like it.
I don't know if it's useful for something in particular.
I just enjoyed it.

Omer (42:28.480)
What's one attribute or characteristic in your mind of a successful entrepreneur?

Martin Gontovnikas (42:33.600)
So I think that the main one is embrace being uncomfortable.
There's a book that I also love that's called the Advantage.
And they talk about how companies only focus on numbers in general, which I think they should, but they don't focus sometimes on organizational health, how to have better teams.
And that's because a lot of people hate being uncomfortable.
And I think that if you are uncomfortable, you grow a lot.
So I think embrace being uncomfortable.
Another one that I think is a great attribute is being genuine.
And that's why I love the phrase that I was saying before.
And of course it's around endurance and being energetic.
I think.

Omer (43:09.900)
What's your favorite personal productivity tool or habit?

Martin Gontovnikas (43:14.300)
It's booking myself time in the calendar.
Back then I used to have a to do list and I hated it because having a to do list means I need to do that, but I don't know when.
So I still have this cognitive load in my brain that I need to do it.
I don't know when it gives me anxiety.
So now what I do is I book myself time in the calendar.
So that gives me focus on income.
If it's booked there for one hour, I'm going to work on that for that hour.
I don't have to think when I'm going to do it because everything is in the calendar.

Omer (43:42.780)
Yeah, that's so true that even if you use something like GTD or Getting things done, you feel like you have a good list of things to do.
But that's exactly where I fall short as well.
You start the day and I probably spend more time going up and down the list, all the time figuring out what I'm going to do.
Whereas if you can kind of commit in advance and say, okay, tomorrow I'm going to break down my schedule and I'm going to do this at this time and this at this time, I don't know.
It's a good hack.
It kind of makes it easier to not waste time figuring out what you're going to do.

Martin Gontovnikas (44:18.010)
Yeah, I agree.
And I think that to me it's such a cognitive load of being worried, being anxious, checking the list, that it's definitely not worth it.
And the other advantage that you have is expectations.
So sometimes somebody gives me a doc and they say, hey, can you review this doc?
So I book myself time and I say, I can do it next Wednesday, is that okay?
And if they say no, I can see what other times I can book with myself and then move them around so that I can give better expectation of when I'm doing things to people.

Omer (44:47.550)
Yeah, that's awesome.
What's the new or crazy business idea you'd love to pursue if you had the time?

Martin Gontovnikas (44:54.940)
So it's not a business idea, but something that I would love to do is be either.
I would love to be like a stand up comedian.
That was something that I would definitely love.
Or something that I would like to do in the future is doing more coaching to people.
Something that I know that makes me feel great is helping other people succeed.
So if there's a business idea around that, I would love to do that.

Omer (45:16.540)
What's an interesting or fun fact about you that most people don't know?

Martin Gontovnikas (45:20.150)
It has to do with stand up comedy as well.
So I was bullied a lot when I was in primary school because of being fat.
So I was really, really shy.
So I thought like, maybe if I do stand up comedy, I can stop being shy.
So I did two years of stand up comedy just to be less shy.
And I think that I'm still shy, but now I embrace being uncomfortable.
Like now, like before we started the podcast, I was very uncomfortable because it's about sharing, but I think it's, it's stand up comedy has helped me be better at just doing it.

Omer (45:50.880)
I love it.
And finally, what's one of your most important passions outside of your work?

Martin Gontovnikas (45:57.360)
So one of them is definitely eating.
I'm a big foodie.
Every time I see I now have like the list of the 50 best places to eat in the world.
So when I go to a city and there's one, I just book it.
Another one is when I say reading, but I don't really read.
I use audible, but I listen to two books every month.
And then the other one is that I'm a big fan of meditation.
I try to meditate every day.
And of course, going out with friends.
I love going out.

Omer (46:24.780)
Awesome.
That's a great list.
Okay, so if people want to find out more about Auth0, they can go to auth a u t h0.com and if they want to learn more about you, they can also check out your personal website, which is g o n to.
And if people want to get in touch with you, what's the best way for them to do that?

Martin Gontovnikas (46:50.450)
So usually it's either Twitter.
In Twitter, I'm mgonto so m g o n t o or if they go to my personal website, the one you mentioned, g o n to, there's a contact form.
So if they contact me there, I'll get an email.
But either of the two is awesome.

Omer (47:09.320)
Gonta, thank you for joining me.
It's been a pleasure.
I enjoyed this conversation and I wish you all the best with auth0.

Martin Gontovnikas (47:20.600)
Thank you very much and thanks again for inviting.
I enjoyed the conversation as well.
Hopefully we can talk soon.

Omer (47:25.800)
It was fun.
Thanks, man.
Cheers.

Martin Gontovnikas (47:27.440)
Thank you.
Cheers.