Plausible Analytics story on the Changelog podcast
Last month we joined Adam Stacoviak and Jerod Santo over at changelog.com to talk about Plausible Analytics.
We go through the backstory of the project, why it’s open source, the details behind a few viral blog posts Marko shared to bring in a ton of new interest to the project, why privacy matters in web analytics, how we prioritize building new features, the technical details behind Plausible Analytics, and our thoughts on a server-side option.
You can listen to the full show below and read an edited and shortened transcript too. There’s the full transcript too and you can also listen to the episode on the Changelog website.
About Google Analytics
Jerod Santo: Marko you wrote a blog post which was quite intriguing - “Why you should stop using Google Analytics on your website”. We thought we’d start with you just giving us the hard sell. What’s the pitch? Why should folks stop using GA on their website?
Marko Saric: Google Analytics is not lightweight. And the official way to install Google Analytics - you use this second script as well, so you end up having quite a lot of things loading, even though it’s something that most people don’t really look at.
We know the GDPR in Europe, there’s the CCPA in California, and there’s one in U.K, and so on… And these regulations require webmasters or website owners to insert these different – like a cookie banner to ask for permission to store cookies, the permission to get consent to share data with the third-parties, and all these.
I agree with GDPR, it’s something necessary for the web to keep it healthier, but I also think that website owners that care about it can use a different solution that doesn’t use cookies, that doesn’t collect personal data, which means also that you can get analytics without needing to give all this prompt to your visitors and kind of annoy their experience.
Adam Stacoviak: I think it’s been hard to grok the dashboards and the data. It’s pretty complex; it’s not very clear, like “This is what you really need to know.”
It seems like one tool built for many different kinds of customers, and the kind of needs we particularly have aren’t exactly the ones that are surfacing easily. Just for that one reason alone it’s pretty difficult. I only really care about a couple pages in there, and even then they’re just difficult to (I guess) bend to my will, give me the information I actually wanna know. Analyze my actual information. It just seems like that’s what it should do well, and that’s the one thing it does, in my opinion, pretty poorly.
Marko Saric: When I did the research for this post, I actually went through Google Analytics and I basically counted all the different reports they have, and I counted more than 125 different reports in the left-hand side. All combined, these 125 reports or so, have about 300 different metrics between them. I’m thinking for my own site I probably use 5-10 at most. The other ones - some of them I’ve never heard about or have never looked at them. But still, I’m running the script there that’s kind of calculating these and collecting this data all the time… And I’m using it either never, or maybe once in a while.
I think I’m not the special case here. I think this is quite common. People install Google Analytics because it’s the most popular tool, or it’s something that they’ve told that I need to have… And they have this - this collects 300 different metrics from their visitors, and they use maybe five of them. It’s a waste in terms of the kb load on every website visitor… Thinking in the sense of the climate change even - how to decrease the Carbon footprint of a website, there’s a Carbon footprint calculator. If you decrease something like 50 kb of your site, your score gets so high that you’re something like in the top 10% of the sites that load fast. So it does make a difference, and it’s something people don’t use to look at; and if you don’t use it, then you can remove it. For me personally it makes sense to do that.
Jerod Santo: For me, if I said “What’s the number one reason why I don’t like it?” I’m just like, I log into the thing and I just wanna log out again. There’s so many features that I don’t care about… I don’t care about conversion tracking, I don’t care about AdSense things, and goals… It seems like the information I want is further away than I want it to be, and the information I don’t care about is right there, and then the filtering is convoluted…
And then the other thing is I don’t really trust it. And I don’t not trust it in a privacy way, although I think that’s there as well; I’d say my second reason is I feel like they have so much of our information, and we’re just giving them more… Like “Here. Here’s all of our website traffic information.” That feels naive to me. But I don’t actually trust their analytics, I think because of our audience demographic, and Changelog.com’s traffic demographic. Most of the people that read our website are blocking things.
Adam Stacoviak: I’m blocking things… [laughs]
Marko Saric: If you think about it, Firefox, Brave - they’re quite popular browsers, especially in the tech community, and these things block it by default. And then not even that - people that use Chrome, they have extensions on such as the AdBlock or uBlock, and they block it by default as well… So it’s not uncommon to see 40%-50% of a tech site going hidden, as in the visitors block Google Analytics. So using a different service perhaps also gives you more accurate data, because that different service is less of a target, as in a less popular service, and it might not be blocked by some of these services such as Firefox, and Brave, and so on.
Adam Stacoviak: I’m sure eventually Plausible though will get to a popularity point that you do get that visibility… And maybe we could talk about how you block scripts, or how you block tracking in a way that respects the user. Because if a user comes to our site and doesn’t want to be tracked, I’m not gonna force my way to track them; that’s wrong. So I want an opt-in world, and that’s what I really care about.
About Plausible Analytics
Jerod Santo: Let’s talk about Plausible Analytics a little bit here, and set the stage… Because this post that you wrote was a brilliant piece of content marketing for Plausible Analytics, which is a service and a tool and an open source application that you two are working on. So the pitch is uninstall GA – and I would say, why do we still use Google Analytics? It’s because, well, what else are we gonna do? It’s free, it’s easy to set up. What else is out there?
So this was a nice piece of marketing, because it’s like “Here’s this great post all about it, and here’s some alternatives… By the way, Plausible Analytics is something that we’re building, which is an alternative to that.” So I would love to hear all about how that works, and some of the stuff Adam’s bringing up.
Uku, when did you start building this, and was it because of the reasons that we’ve been discussing with regard to the status quo of tracking and analytics with Google and other large providers?
Uku Täht: For sure, yeah. When I first started writing the Plausible codebase, I didn’t wanna use Google Analytics, but its – I didn’t have much of a problem with its UX, because I had never really used it; I’m a developer, I don’t spend much time in analytics… But I was working on a different project, and the marketing guy asked me to install an analytics tool… And he asked me to install the industry standard, Google Analytics obviously, and it just rubbed me the wrong way, for some reason… Because the year previous I’d become very aware of all these privacy issues, and I was trying to use less and less services by Google. I was just getting off Chrome, and trying to replace Gmail, and things like that.
So being in that mode of de-Googling my own life, I thought “Well, I don’t like installing Google Analytics for my project”, but I had to do it, because I also didn’t have a good alternative. There were some alternatives, but I thought some of them were just very simplistic, and quite expensive, to be honest. It’s hard to justify paying for analytics when there’s this standard solution that’s free for everyone.
But you know, you do realize that you end up paying with data, essentially… So I thought there’s room for an interesting alternative there, and I started writing something. I didn’t know where it was gonna go, but I had a proof of concept in mind, so I just thought “I’ll get started on it, and I’ll run it in parallel with Google Analytics and see how it works.” And it took about three months, I think, to get a simple beta going initially.
Adam Stacoviak: What was involved in that? What were some of the initial features you focused on? And even how did you focus on those initial features?
Uku Täht: Yeah, I just figured “Well, what are the basic stats?” I didn’t have really any experience in analytics before that, so I had to kind of learn about analytics, like what’s even a useful stat to have… So I just started using common sense. I wanna know how many visitors would visit my website, I wanna see how many pages they’re viewing, what’s the top content, what referrers they use… And I kind of started building things from scratch.
Obviously, I took a look at all of the other analytics tools and tried to distill the most useful UX and what features they’d surface on their dashboard… But yeah, the first proof of concept was just having a graph with the visitor numbers, how many visitors there were in a given timeframe, and giving the top referrers and pages for that time.
On raising awareness in the early days of a startup
Uku Täht: Before I wrote a single line of code, I wrote a blog post that said “Here’s what I’m doing. Here’s how the proof of concept looks like. I don’t know how long it’s gonna take, but if you want to join the beta, send me an email.” I shared that on Twitter and some communities, and I think I had – not many people; like 20, 30. But that was enough to give me the motivation to finish the proof of concept. So that’s something I’d recommend to everyone who is thinking about writing a side project, getting something going.
Jerod Santo: Did the early interest inspire you, or was it more the commitment?
Uku Täht: Yeah, both. I felt like the fact that people cared enough to send me an email about it, but also I felt like I have something to show to them in a few months… So I felt like I made a commitment to some real people. And that changed it. That actually kept me going for three months, to get it out there.
Jerod Santo: Yeah, absolutely. So you had at least that much intuition, and you had something going pretty quickly… But there are so many people announcing launches, and so many alternatives to things, and so many open source projects even. This is a common theme on this show; people we talk to are like “I built a thing. Now how do I get people to use the thing, or interested in the thing?”
So you have a nice one-two punch here with Uku and Marko. Marko wrote that post, which brought a lot of attention to Plausible, and definitely striking a cord with people that are already angsty against GA, or looking for alternatives.
Uku, how did you guys meet? Did you decide “I can’t tell this story on my own, and I need a helper”? How did that go down?
Uku Täht: Yeah, it was two reasons. One was, like you said - I’m good at writing code, but I’m not very good at writing blog posts, communicating stories and ideas to people. You can get good at anything, I think, if you put enough time into it, but at some point it’s also useful to focus on what you’re good at and try to bring someone else in who can complement your skills, so that you can both be just experts in your own fields.
I realized I’m not gonna build deep expertise in marketing and content writing. That was something that after doing it for – I was trying to tell the story, to get people to use it for about a year… And I had minor success, but nothing to write home about.
Jerod Santo: What was the stuff you were doing?
Uku Täht: I was trying to write blog posts, I was writing emails to people, “Hey, can you include me in your blog post about Google Analytics alternatives?” I was writing updates on Indie Hackers, I was trying to post stuff on Hacker News… Sort of haphazard. I didn’t have a strategy. I was just opportunistically trying to get in front of people with Plausible. But at the same time, I felt like it was taking time away from what I really enjoy doing, which is development.
I really wanted to involve someone who could help with marketing, with getting Plausible in front of people, and telling its story. But the other aspect, which is why I reached out to Marko, is that accountability aspect that we talked about earlier. I felt like working on it alone, I started going a little bit crazy sometimes. If you don’t have someone to talk to, if you don’t have someone to hash your ideas out, someone to tell you when you’re wrong, it’s so hard to make up my mind. I was going back and forth on a lot of decisions; I didn’t commit to a strategy… I didn’t like working alone.
It’s interesting - there’s a lot of talk about being a solo founder, and whether that’s a good thing or a bad thing… And I felt like it was a bad thing. I didn’t enjoy it at all. I think people are different, but for me - yeah, I felt like I was going crazy at times.
Adam Stacoviak: Yeah. When you have somebody there with you, it’s good to get direction from that person. They’re your litmus test in many cases too, but it’s also more fun.
Jerod Santo: So how did you find Marko?
Uku Täht: I think it was January this year, I was just scrolling on my Twitter feed, and I saw this post about de-Googling your life; moving from proprietary tech company tools over to more open source solutions. Marko was in that blog post talking about what you can use instead of Chrome, and also Google Analytics, and he had alternatives for all the Google products. I felt like this is awesome content. I just stumbled upon a great blog post.
Jerod Santo: Kindred spirit. Is that the AMP one? Adam, didn’t you put that on Changelog News?
Adam Stacoviak: I sure did, yeah.
Jerod Santo: Was that the AMP one or not? I know that was the December time range, Marko, when you wrote that one…
Marko Saric: Well, for whatever reason, I wrote several Google posts… [laughter]
Jerod Santo: Guy is like the Google killer. Google assassin.
About Google’s control on the web
Marko Saric: I wrote one for your personal life, and one of your website. That was basically me, in my life, over the last two years or so, trying to figure out “What can I do to make the web a healthier place? What can I do to support the smaller players, and what can I do to get away from the ad tech, and so on?”
One thing was Google is obviously the target there, because – you know, Facebook is easy, because if you don’t use Facebook, you don’t see it much. But Google is like – any website has Google fonts, has Google Analytics… Google has everything - the AMP, and so on. And Google is so much more difficult to get away from, and this is maybe why I focus more on Google rather than something like Facebook, which I think is probably an even worse company. So just because of the fact that Google is so much more difficult, so much more ingrained into pretty much every website that we visit… That’s where some of these posts have come from, that was the motivation really.
Adam Stacoviak: Yeah. In a lot of cases, this reminds me of the Basecamp story, Jerod. Back when Basecamp first came around it was the power story; that’s what we were kind of talking around. They have a story of this David and Goliath, which is mentioned in one of the posts on Plausible.
But this whole idea of like when Basecamp first came around - its claim initially, which got it its headlines, was “We’re not Microsoft Word” or “We’re not Microsoft-this”, or whatever. It was this idea that these anti… You sort of knew what you didn’t wanna be, so it was fairly – not easy, but it was sort of easy to see what you don’t wanna be, and you can kind of see what you do wanna be, and people can grab a hold of that. But Marko, as you’re saying - you’re right. If you don’t log into Facebook, you kind of don’t see it. But Google is everywhere.
Uku Täht: It’s almost as if their business strategy is to embed themselves in the structure of the web, and then be the middleman for ad buys.
Jerod Santo: Right. And they’ve always provided more useful tooling. I mean…
Uku Täht: Oh, the tooling is great.
Jerod Santo: I mean, I complain about the interface, but Google has provided to developers, and just the techies, for years, very valuable – I mean, Google Reader was a hugely valuable tool.
Uku Täht: Google search is by far the best search out there, I think… But I don’t use it. [laughter]
Marko Saric: We were both big fans of Google. Look back 3-5 years, I was the one using 5-7 different Google products every day, and I was the one telling to my parents and my friends “Check out Google Inbox. It’s the greatest inbox for email. Check out this, check out this…” And I don’t know, over the last two years or so my personal opinion has changed about these things, and now my thinking is a bit different. And I’m not the only one; there’s kind of a growing movement, if you want, of people who want to de-google their lives, or de-facebook their lives, and support some different alternatives.
Uku Täht: Yeah, so I came across Marko’s blog post about that, and I thought he’s telling the story, he’s doing an awesome job at it. I read more blog posts, I thought the content was just amazing, and then I went to the landing page of his personal site and it said he’s a marketer. I was surprised, honestly. I thought he was a developer by the content that he was writing.
Jerod Santo: I thought he was a developer when I first contacted him.
Marko Saric: [laughs] You know, I got a compliment from Uku the other day. He said that my blog or my personal website doesn’t look like the typical WordPress site that he thinks about. When he thinks about WordPress, he thinks about all these banners and all these flashing stuff, and lots of stuff that are right in your face… And my one is more basic, like something that you might have a developer do on some smaller CMS or static site. That was a compliment to me.
WordPress by itself is not what you have the image of in all these marketing sites, with all these calls to action, and so on… It’s what people put on top of it that makes it so. So yeah, just a nice little compliment to a non-developer.
On the state of the web
Adam Stacoviak: I wanna go back to what you said though, Marko… You wanted to play a role in making a more healthy web. And this isn’t simply just “Google = bad. Plausible = good.” It’s not just simply saying that. It’s more this notion of power. How you make a more healthy internet is probably decoupling away from one person or one large entity controlling the data.
So in the aggregate, if GA has been freely available and is so accessible to many people - and I think the stat is somewhere above 80% of most websites out there are using this free tool… And that means in aggregate, over many years, potentially decades, you’ve got a lot of data. As an ad tech company - I’m not saying that they’re using it in bad ways, or they’re bad people, or they’re a bad company; there’s varying degrees of that. But the point is just like when you put that kind of data in one organization’s power or control, potentially bad things could happen. When you control your own data and you have your own data, then we don’t have to worry that some other organization has our data, whether it can be used against us or not. It’s just a matter of “You don’t know when you want privacy until you need privacy.”
Marko Saric: I think my thinking of this started to change with all the Snowden stuff, and Cambridge Analytica, and all these things happening a few years ago. I’m a marketer, and I was using these tools personally, I was using these tools in my profession as well, and that kind of made me aware of the issues. I was ignorant about these issues before, I guess, and it took Snowden and it took all this media campaign and all these people to talk about it for me to realize “Hm, maybe I should rethink what I’m doing here. Maybe this is not the healthiest in the long-run, to have Facebook and Google pretty much control everything we do online.”
I have my own blog now, that’s disconnected from everything. Plausible is trying to get at least some websites to choose the same, to disconnect in one way or another from these big companies.
Jerod Santo: Does that make you unique amongst marketers though? Because when I think about who wants more analytics generally, or more information, more tracking, it seems like marketing folks do, because they can then do better at their job. When Adam and I talk, and sometime she’ll put on more of his marketer’s hat, and he’ll start to say “If we knew X, we could do Y.” And I’m always like “Yeah, but X is gross. We can’t do that, right?” Not always, but you know, we have a balance… Because he’s thinking like a marketer. And then when he thinks more like somebody else, he’s like “Yeah, that’s not a good idea.”
Marko Saric: Many marketers for example use ad blockers themselves… This just tells you that maybe even within the marketing world this whole thing of collection of data and privacy invasion is not optimal, it’s not something that many people like. But yeah, normally - and this is what we discussed about Google Analytics - the fact that Google Analytics has 300 different metrics is because somebody wants as many metrics as possible about their users…
There’s an argument that the more you know, the better it is, but I think that doesn’t mean that you need to collect as much as you can. It’s better to limit it down and actually understnad what you need to know from your website, from your customers, and so on, how can you use it, and what’s the best way to get that without going overboard and collecting everything and making all these behavior profiles and all the other tracking across the web as people browse different websites and all that stuff… Kind of a balance, like we mentioned.
Uku Täht: Marko, in my view, you’re quite unique, being a marketer who thinks that the web is a little bit broken and wants to fix it. I remember telling to my friends before I stumbled on your website, “I wanna find a marketer who cares about privacy and open source”, and I couldn’t find one. I was trying to find someone who could help me with this, but someone who wasn’t just a marketer as I thought.
No offense, but it’s also my understanding that it’s marketers who want more and more data usually, and from the marketing departments - it’s where some of this data collection issues are coming from, and the previous issues… So I thought it was very unique to find someone who can do marketing, but also is in that same sort of mindspace in terms of understanding what Google is about, and trying to fix that.
Marko Saric: Yeah. Maybe I’m just a bad marketer, that I don’t collect all the data. [laughter]
Uku Täht: Well, you don’t need it.
Marko Saric: Yeah, I think in general it’s something that you can argue that there’s a need for it, but it’s not something that’s necessary. I think the more important thing is being able to have a product that people want, and then being able to communicate about the benefits of that product to people that are interested in it, to people that that product solves issues for. And that thing you cannot really do as well just by knowing a lot of data; you actually need to speak to real people, you need to get into their shoes and understand them better in order to communicate with them, or in order to create a product that actually solves the issues they have.
On deciding which features to build and which not to
Jerod Santo: How do you guys decide what to build and what not to build?
Uku Täht: That’s actually really easy. I decided that I wanted to make it as open, and the development as Plausible is gonna be as open and transparent as possible. The reason is easy - we have a public roadmap and we have a public forum for feature requests… And pretty much people upvote on what they want, and I just go in order.
Jerod Santo: But have you ever seen the Homer Simpson car? Do you know what a car would look like if Homer Simpson designed it, and it’s just got like knobs and widgets and horns sticking out the side? If you just give people what they want, you end up with a monstrosity two years from now, don’t you?
Uku Täht: That’s a good point. You have to weigh it against some of the values or the vision that you have for the product…
Uku Täht: So there’s a feature that I’m planning to add that no one has requested, for example, which is being able to just query ad-hoc anything, basically, by clicking on whatever you want in the UI. There’s no request for that, and I’m gonna do it at some point… And I’ve said no to things for sure, but more or less - I’d say 80% - the prioritization comes straight from users.
Adam Stacoviak: I think it’s great to have an open forum to invite people to. How do you gauge the judge on – is it simply just votes? I suppose, at this point, to some degree… But what else do you weigh it against?
Uku Täht: I think I weigh it against my own vision for the product, what I’m seeing in about a year’s time, and then thinking “Are these the things that will lead us to that point?” That’s one of those things. I just want to build a good product.
We didn’t have a thing that would differentiate us, that would give people a really good reason to switch, for the longest time, and I really wanted to have one. So being able to do analytics without cookies is now a big one. It wasn’t the highest-requested feature, but that was one that I brought in to the top, basically, because I really wanted to have that reason why someone would look at “What do I get by switching to Plausible? I’m gonna switch from a free tool to an analytics tool that basically gives me the same stats, and I’m gonna pay for it.” So it wasn’t a great value proposition for the longest time, but now that we do analytics without cookies, which I don’t think Google Analytics is able to do, I think that gives people a good differentiator.
Jerod Santo: What are the tradeoffs? What do you lose going cookie-free?
Uku Täht: You lose some accuracy. In my testing, I was running both approaches. Currently, the unique user tracking is based on the amount of IP addresses that access your website, anonymized. The numbers are very similar to the unique counts that I was getting using a cookie. So I was actually surprised. I thought that I was gonna see numbers that are quite a bit off, but they’re actually very similar. And unique user tracking with a cookie is not accurate completely either. Everyone has three devices now. You’re basically tracking devices, not visitors.
In some cases, an IP address might actually be more accurate when you have multiple devices on one IP address. So there are interesting trade-offs between cookies and the IP addresses.
Marko Saric: But remember, if you take the whole concept of the product, then it makes sense to make these slight tradeoffs here and there…
We are catering to people that care about these things, just as we do, and they will also want to remove Google Analytics from their site; they also want less tracking, but they still wanna see something. This is what allows us to make these decisions, because we know we’re pretty much on the same page as a lot of our audience.
If you go to the roadmap right now, which is on our site, a lot of those requests - they’re pretty much fit with what you’re thinking about. Now it’s just about prioritizing them and getting time to do them, and doing them right, so it fits with the product. But it’s not like we have completely different people asking for something that really does not fit. It’s a niche product that people that come to it are actually interested in this thing, and then they kind of think in similar ways, which really helps in building a product that’s kind of unified and that makes sense for this audience.
Uku Täht: Talking about prioritization - there’s a really interesting new input, which is Marko. Now that we have a marketer on board, we are dogfooding our own products more and more. Previously we didn’t have much traffic, and it was me looking at the stats. “Just 50 people today. Cool.”
Now that we have real traffic, and we have more of a marketing approach and focusing more on selling the product, I think the dogfooding aspect will really start feeding into the product as well.
Marko Saric: And I know the tools from my past experience, working for different companies. I’ve spent hours tracking heat maps, and looking at what people click on on the funnel to sign-up, and so on… So I know what is out there and what is useful, so now it’s like me feeding back from that side, and then Uku feeding back from the, you know, “This is not possible because you cannot do this without cookies” or “This is not possible because we would have to track/identify people…” So the bounce back and forth works in that sense.
About self-hosting analytics
Jerod Santo: So you’re in an interesting spot and a niche where your customer base or your perfect user is privacy-oriented, but not so much so that they have to run all their own things… Because you’re still hosting the data. It is open source, it’s an Elixir app, I assume there’s lots of moving – we also host an Elixir app. There’s things to do to host that yourself. Maybe you can make that click-a-button deployable at some point; whether that’s in your interest or not I’m not sure. But you’d think the real privacy-oriented people - they don’t wanna host their data with you. They wanna run their own plausible.io. Have you run up against that? Those that are really “I would love this, but I’m not gonna host it on your guys’ stuff.”
Uku Täht: Yeah, and I wanna offer that. It makes sense.
The main reason why I haven’t done it so far is because the product is still fairly early-stage. There’s a lot of not only moving parts in the infrastructure requirements, but a lot of moving parts in terms of just upgrades to the database schema, for example. I’m now working on adding a second database to the infrastructure. It’s work to just upgrade my own servers, but having to upgrade a hundred other ones, and having the documentation and the click-a-button convenience for that - it’s a bit too much for me right now.
Jerod Santo: Yeah, it takes away from other things you’d be working on.
Uku Täht: Yeah. But I have nothing against people – it’s released under MIT; that means you can do whatever you want with it. You can start your own company running the same code, if you want. But recently there’s been more and more interest in self-hosting Plausible, and there’s a GitHub issue with now three people involved - well, excluding me, so four people involved - and I’m trying to offer my own help as much as I can to make it self-hostable.
Adam Stacoviak: You mentioned it was MIT… What was the thought behind making it MIT? Why is the transparency important? Obviously, Google Analytics is not, so…
Uku Täht: One of the things I thought about was “Well, what would stop a company from becoming Google?” I think one of those things would be “Well, if you don’t trust the company anymore, you can just take the code and build another one using the same product, for example.” Having the code in the Commons rather than trademarked I think is valuable to the community, because it stops a company from going haywire, I think, in terms of what they do with the data.
You can’t lose customers’ trust. You have the threat of forking, and the threat of forking is what keeps a company in check when their code is open source.
Jerod Santo: Well, I’d like to take this opportunity to announce our brand new service. It’s called More Plausible Analytics… [laughter] And it’ll be up here real soon. Just kidding.
Marko Saric: Let’s see if you can do content marketing as good as we can.
Uku Täht: There you go. That’s the secret sauce.
Adam Stacoviak: Well, that’s the question - are your blog posts Creative Commons, ShareAlike 4.0?
Jerod Santo: [laughs] Then we can just rip off everything you’re doing.
Marko Saric: You know, one interesting aspect that I’d never experienced before - the other day I got an email which was just from GitHub, an automated email. Somebody was telling me he was reading my posts before I published it. And I didn’t realize that we’re now using a CMS which goes through GitHub, so every time I do my drafts and I save something, it’s there, and people can see it and read it.
It’s just a funny aspect of it that I did not consider… But it just makes the whole thing more open - from being open source, to having this open roadmap, to even our silly little blog posts being on GitHub while they’re being written…
Jerod Santo: …before they’re published. That’s one webhook away from easy plagiarism. Every time Marko pushes to GitHub, just webhook that sucker and publish it.
Marko Saric: Don’t start me about plagiarism. [laughter]
Jerod Santo: Oh, are you getting ripped off?
Marko Saric: Yeah. And maybe I should not talk about it; I’m trying to be quiet. But you know, it’s been going on since this successful story of ours… Yeah, it’s been going on, unfortunately, from even bigger companies, if you want… Which is funny.
Adam Stacoviak: Well, things happen when you ruffle feathers. When your story makes sense to a large majority who have been paying attention…
Jerod Santo: And when you’re in the limelight, people are like “I want some of that limelight.”
Adam Stacoviak: Yeah. Or they get threatened.
Marko Saric: I think this is the way I look at it. I’m looking at Google Analytics as a competitor, if you want, and we’re trying to find some of those hundreds of millions of sites that use Google Analytics to maybe consider Plausible… And being threatened by someone of similar size, I’m not. So yeah, it’s just the way some people react to it differently, and then they see somebody having a little bit of success – and because we’re so open, people can see that we’re having success… It backfires in that sense. Some competitors can actually try and steal our limelight from us.
Uku Täht: Remember, imitation is the sincerest form of flattery, right? It means we’re doing good.
On server-side analytics
Jerod Santo: I wonder if you’ve considered server-side logging, or if it was JavaScript all the way. Because your JavaScript is not doing very much at this point.
Uku Täht: Yeah. This is one of those things that I don’t think it’s been requested on our feedback page… Or maybe it has, and it only has one or two upvotes. But this is one that I’m raising the priority up by – yeah, I wanna provide that, for sure.
Jerod Santo: Sign me up.
Uku Täht: I think there should be a module for NGINX or Apache, whatever you’re using; if you’re running Elixir, I can just write a plugin, and send that over to Plausible from the backend. And you will miss some of the things… I can’t get the clientWidth, for example, of the browser.
I think it’s nice to do device detection based on the actual width of the browser, rather than the user agent.
Jerod Santo: Right. Viewport size is important, yeah, for responsive design.
Uku Täht: When you’re writing CSS, you care more about how it’s displayed… And there are things that I’m planning to add to the product that might require JavaScript, but the general idea that the basic things, like getting visitor counts and refers and user agents - this doesn’t need JavaScript. And I would recommend everyone do that once it’s available with Plausible.
Jerod Santo: The nice thing about JavaScript is at least you know you have a device that runs JavaScript, right?
Uku Täht: Bot detection is hard, and it’s one where Google has so much data, and they’re not releasing that.
So it’s tricky… I’ll have to see what I can do. Currently, the bot detection is very basic, based on user agents. And the fact that the client has to run JavaScript for it to be counted… That counts out a lot of bots.
So there are challenges there. And there’s a challenge of convenience. The frontend is standardized, but the backend isn’t… So there will be many different modules and libraries for people to hook into when you do offer backend tracking. I mean, it can start with a simple HTTP API, for you to just shoot requests from your backend. It’s a hard problem, and it’s not one that I have spent too much time on right now, but it makes total sense to me. That should be available.
Jerod Santo: It’s the ultimate differentiator. It goes the complete opposite direction. And I haven’t seen – because I’ve been looking for it, and I was like “Is this something that we should build open source, or something?” I’ve looked – there’s obviously open source log analysis tools, and a few things our community members in Slack have pointed us to… Nothing’s quite been what I thought it should be, so it would be a differentiator. No JavaScript. That sounds nice. That feels nice.
Uku Täht: Yeah. I’ve actually said publicly to a few people, “If you feel like you wanna deal with the log analysis tools and manage that - I think that’s better than including our JavaScript.” In terms of privacy, it’s great.
Adam Stacoviak: How does that impact real-time metrics?
Jerod Santo: Yeah, implementation details in terms of “Are you deciding to have a standardized log format that Plausible (or whatever tool) reads, or are you streaming data directly into an API? Are you batching that as real-time?” There’s lots of things to decide. You guys have probably thought about this more than I have.
Uku Täht: Honestly, I haven’t… Because it’s one of those things where when I decide “This is what we’re doing”, then I’ll do a bunch of research on all the different options out there… But log analysis is its own little world. I don’t have much expertise, but what I do know is that the biggest challenge is probably bot detection.
Uku Täht: Yeah. There’s stuff you can do with JavaScript that you wouldn’t be able to do from the backend.
Jerod Santo: Like event tracking. In-browser event tracking.
Uku Täht: It’s much more convenient to track events, yeah. You can do more accurate time on page. If you want scroll depth… You can do interesting stats like that, which require JavaScript.
On being open source
Adam Stacoviak: Well, so Plausible is open source… Is it open source in the fact that you’re taking contributions and looking for contributors? To what degree can someone get involved in Plausible?
Uku Täht: Yeah, I think once we figure out the self-hosting aspect in the coming weeks, it will be very easy to take the code and run it yourself. That’s where things can really get going in terms of contributions. I’d love for people to get involved.
Marko Saric: About the contributions aspect on GitHub - this is again the first time that I’m dealing with this kind of GitHub and open source aspect of running a startup… When this blog post went viral a few weeks ago, I think the day after there was a new thread on GitHub - quite a long, beautiful thread where a guy said “You’re using unnecessary code”, or I don’t even know what he wrote; Uku is much more familiar with that. But basically, a guy that really knows his stuff - he came in there and he wrote pieces of code, how we can improve our tiny 1.4 kb script to go even lower, to perhaps get it to 1 kb or even under 1 kb.
I thought that was amazing, that somebody would take 1-2 hours to write quite a long GitHub thread to help us, without us asking for it, without us knowing him… Just because he read that post perhaps on Hacker News, and he thought “Okay, these guys don’t know what they’re doing. Let’s get it from 1,4 kb to 1,1 or 1,0.” I just thought it was amazing that this kind of contribution can happen.
Uku Täht: Yeah. I know from experience that running an open source project is a lot of work as well. So far, I’ve focused on the product aspect and the business aspect of things… Because one of the trickiest subjects is “How do you get open source funded? How do you get people working on open source in a sustainable way?” That’s kind of what we’re trying to figure out here… It’s definitely a direction I wanna go in, is more community involvement and more of a community project where we charge for hosting it.
On results we’ve achieved
Jerod Santo: Let’s get back to the blog post. How big did this blog post go, and what did that result in in terms of users, trials, people giving you money?
Marko Saric: Yeah, basically the post was published on the 8th of April, I believe it was, and our stats for April were just over 60,000 visitors. Almost 63,000, which is two-and-a-half-thousand percent increase compared to March. That’s one number to look at.
We offer a 30-day free trial, so people can sign up to test us out before actually deciding if it’s worth it… 272 people signed up in April, which is six times more than sign-ups in March. And actually, all the sign-ups in April - they were more than the previous nine months combined. So just that aspect - we got a huge boost in visitors and a huge boost in new sign-ups for trial; it was great.
Now we’re kind of reaching those 30 days or so from that first day, which means that some of these trials are expiring. Now we basically went over 100 customers, so we have seen a 70% increase in paying customers, from the day of the post until today. We’ve seen MRR as well increase by 80% from the day of the post till today.
These are very concrete numbers, which you can go and say – doing this type of marketing, it can work as well; you don’t have to go to Facebook and pay Facebook. You can do some content marketing, you can reach out to your audience organically, and you can actually still achieve results. This is at least one that proves it. Maybe we were just lucky, I don’t know, but it can be done.
Jerod Santo: Well, you’ll find that when you write your second big post, and you’ll see if you hit a home run again, right?
Marko Saric: Yeah. I wrote in the follow-up to that one – I was like, “The only way from here is down.” [laughter] Because I don’t wanna promise people that – like Uku was saying before, he was writing posts before, but that doesn’t mean… You can create the greatest post, you think it’s an amazing post and everyone should read it, but only your mom comes to your site and reads it maybe.
So this kind of thing of publishing content works in general in the long run if you’re consistent with it and you produce value, but you also need some luck here and there in order to get a spike. This is why I mentioned the only way to go is down, as in “Don’t expect us to have a spike every time we post something, or even every month.” It’s something that rarely happens. But the important thing is that you’re building up the content, you’re building up the value you’re creating, you’re answering people’s questions, you’re kind of building up the authority and the name of the brand… And this kind of slowly rises from day to day, from week to week. And then, if you zoom out and look at the long-term picture, “Oh, there’s actually a big achievement. We’ve actually gone gradually up”, and now in the last six months there’s a huge increase, rather than one spike and then nothing again.
About coming up with ideas for blog content
Jerod Santo: When we first spoke, you said “Why to take GA off your website” was like the post you already had written in your head when Uku first approached/met you. That one was obviously successful, and now you’ve had your follow-up, like what traction looks like… That’s an interesting sophomore album… What’s some other stuff you have in the hopper?
Marko Saric: I think it’s important to understand who you’re speaking to. If you can do that, if you can see it from their perspective… So one of the things that many companies go wrong with is they think of themselves first. They’re like “I’m gonna sell my company, I’m gonna sell my features and my product.” They’re not thinking about who they’re trying to sell their product to, or they’re not thinking about what those people actually want.
My experience is if you actually think of them first, and you actually provide value to them first, and only then indirectly you can talk about your product and what your product can do even better than whatever solution you described earlier on, then you can actually see better results. But the majority of sites go like “Me, me, me, me” rather than “You, you, you, you.” And there’s no love at first sight in online marketing. You actually need to provide real value to people for them to actually understand you, to take you seriously, to actually spend a few minutes of their time to actually explore your posts and your product… Because people are busy, people are impatient online.
Jerod Santo: So Uku shared his roadmap… Can you share your Drafts folder? What do you have – I know you share it on GitHub, but do you have some titles? What is the kind of stuff you’re working on and you’re thinking about writing, things that are gonna go out?
Marko Saric: You know, because of this success here we’ve had, I’ve been doing more interviews and speaking to other people… Because so many people are curious now. They heard about Plausible, and they wanna either feature us on their sites, they wanna ask questions, and things like that. There’s many questions on social media, email…
If you look at our blog, it hasn’t been that active since then. We’ve published twice or three times… But basically, one good example of what I’ve published since – obviously, I’ve published one with the results and how one blog post can make a difference… But the second one I published is actually six or seven different people have asked me pretty much a similar question, like “This sounds great to me. I also don’t like Google etc. but would I lose my search engine rankings if I remove Google Analytics, or does Google actually use Google Analytics and the data they get from there to help me rank better for a search, and get more visitors?” That was something I didn’t consider; that was not on my roadmap to write about… But I was like “I see. I understand the opportunity, I understand what people are asking for and what they’re curious about, and what questions they have.” And then I did my research and I wrote whatever I could find directly from Google on this topic, and I published that.
Again, you’ve gotta put yourself in the shoes of people that you’re targeting and just understand what kind of questions they have what are they thinking about, and these things… And then you basically roadmap according to that.
It’s not like I came with a roadmap of the next six months content and I said “We’re gonna do this, this and this.” I might have an idea or two here, but then it’s about listening, monitoring, being part of the conversation… I spent I don’t know how many hours looking through hundreds of Hacker News comments, and Lobste.rs comments, and people on Twitter, and Mastodon, and so on.
I spent hours on reading, taking notes and trying to figure out what the situation is, what’s the feel and what are people talking about. This then leads us not only to create more interesting content in the future, but also to improve our product.
Jerod Santo: That’s true.
Marko Saric: Basically, it’s really important to be connected to the people, to the community you’re talking to, and react. Don’t have like “These are the set rules. We are gonna do this the way we want.” No. Be part of the conversation and be flexible enough to understand where the conversation is going and work according to that.
Adam Stacoviak: There’s one word to describe somebody who does something like that. Well, a phrase technically, but one word that describes it. They care. You show up, and you participate if you care. You solve problems because you care. You can write these blog posts because you have empathy, because you care.
About the future and sustainability
Adam Stacoviak: So where will you all be at in a year from now? …considering the success you’re at right now, the adaption rate you’re working on, your open source roadmap, this idea Jerod’s given you that you also agree with on server-side… Where will you be? What can we expect?
Marko Saric: I think we’re gonna still be on your podcast, talking about the only spike we ever had in traffic, and the only time Hacker News talked about us. [laughter] Like “Back in the day, that was a nice day…”
Jerod Santo: We should have a one-year retrospective, and we can just reminisce… I’ll bring the champagne.
Uku Täht: From my side, I think my goal for the year would be to be sustainable as a business. I wanna make sure that we can do this full-time, without worrying about the runway. I love the product so much, and the project… It’s really important to get to the point where we can just work on it without worrying about it. We’re bootstrapping it, so it’s difficult financially, but we don’t want to have anything to do with VCs, or that kind of stuff.
Jerod Santo: Are you all-in at this point? Are you still working?
Uku Täht: No, I’m not. So I’m all-in, I guess… [laughs] I’m totally comfortable sharing the MRR. We’re at $800/month, and we’re very open about that. I write a monthly journal where I share that as sort of where we are in the journey to get to a sustainable business. That’s the most important thing, I think… Because if we don’t reach sustainability, then there just won’t be anyone working on it, I guess.
Adam Stacoviak: Yeah. Well, in April’s post about March, the recap – so one, I love these recaps you do. i think they’re really awesome, and very transparent in terms of your growth… And you do mention MRR publicly; it seems March was $415, so you’ve doubled based on what Marko said earlier, and what you’re concurring against here in your blog post.
Jerod Santo: What’s that number look like at sustainable? What’s the threshold?
Uku Täht: We haven’t talked about it with Marko. I don’t know… I live a simple life. I can live on a minimum wage here in Estonia.
Jerod Santo: Which is what.
Uku Täht: Which costs the company about 700-800 Euros a month. I’m fine with that. It’s a cheap country. [laughter]
Marko Saric: There for one of us. [laughter]
Jerod Santo: Yeah, you’re halfway there, right.
Marko Saric: We’ve gotta double up.
Uku Täht: The servers and the databases are quite expensive…
Jerod Santo: Yeah, what’s your monthly expense?
Uku Täht: We’re going – just the Postgres database right now is like $160-$170. The server itself is $7/month. So altogether we’re running about $200/month in terms of expenses, for now…
It’s one of the first products I worked on where the cost efficiency is a big deal. Previously I’ve only worked on business applications that are just glorified spreadsheets where it doesn’t matter. It’s gonna run on any machine. It might get a little bit slower and you just throw more metal at it and it’s fine. But with analytics…
Jerod Santo: So where are the needs coming from? Just from the data storage aspect?
Uku Täht: Yeah. That’s one aspect to think about when it comes to sustainability… But we still have some way to go. A year is a good timeframe to reach financial stability with a company.
Marko Saric: One good headache we’ve had over the last few weeks since this post is that several bigger sites have come to us and said “You know, on your site you mention only (whatever it is) a hundred thousand as the top level… How about three million, how about four million?” And now Uku is in the process, like he mentioned earlier, of scaling Plausible, and making it more efficient, so we can actually at some point in a couple of weeks perhaps start beta testing some of these sites that have 4-5 million per month of users/visitors.
Jerod Santo: Do you have to introduce enterprise pricing at that point? I haven’t checked out your pricing very closely. Does it scale up pretty well? If somebody’s gonna have massive data needs, are they gonna be paying more?
Uku Täht: I think it’s gonna be fairly – not linear, but the more you use, based on usage, the more pageviews you have, the more it’s gonna cost. I wanna keep the pricing as simple as possible, just based on traffic pretty much. We’ve talked about some pro features, but personally I’m not a fan of having complex pricing, where you try to navigate the features to get people to upgrade. I’d prefer to have a simple, like “Here’s the product. As much as you use, you’re gonna pay for it.” I like that.
Jerod Santo: Cool, guys. Well, we hope you get there. We will be rooting along, we’re following the blog. We’ll be trying out Plausible. I’ll definitely be checking out that roadmap and waiting for the server-side stuff to hit.
On hosted vs self-hosted analytics
Uku Täht: Well, just an interesting question - are you comfortable with us hosting the data, or would you also want to run your own instance of Plausible?
Jerod Santo: I’m not interested in self-hosted whatsoever.
Adam Stacoviak: We use Google Analytics right now, so… I mean, we’re already comfortable with the beast, let’s say.
Jerod Santo: I’m not a privacy far-ender, and we talked about where our stuff has to be all self-hosted, at all. I like the server-side idea. I like the simplicity of it, I like not shipping any JavaScript to my users… I’m more about the user privacy than the hosting thing. I guess at the end of the day – well, I think you can probably host it more securely than I can, probably… [laughter] So I guess that is probably a pro, user privacy, versus us having to maintain that. But I do like the server-side idea, and I like not having to ship any JavaScript whatsoever. Those are the main things. I like the simplicity in terms of the user interface, and you already have that going for you in the product. The way you display information is nice and neat, and it doesn’t have a bunch of the stuff I don’t want.
There are a few things I do want that aren’t there, but like I said, I’ll just check out that public roadmap and maybe I’ll give a few thumbs up to some things… But yeah, that’s interesting to me, that aspect, and really just kind of return to the old way of doing things, server-side versus in the browser.
Uku Täht: I’m with you. I think allowing people to self-host makes sense for Plausible. We’re in that sort of space.
Jerod Santo: Absolutely.
Uku Täht: But personally, as a user, I’d rather pay someone to take care of that for me.
Jerod Santo: Same.
Uku Täht: I wanna minimize my headache as a developer.
Jerod Santo: There are people that love setting this stuff up though. They love self-hosting, they love running servers, they’ve got their Ansible scripts, or they’ve got their Docker files, or they’ve got their Kubernetes clusters. There are people that love that stuff. I’m just not that person. I am interested in server-side stuff.
A feature request
Jerod Santo: So now that I have you off-air, let me give you my feature request real quick.
Marko Saric: Okay.
Uku Täht: Yeah, go for it.
Jerod Santo: So what I’m interested in most of the time – so our site is pretty well designed from an information architecture point of view, if you think about our different podcasts, if you think about our news and then our posts. We think of these as almost like little verticals inside of our media company. So Go Time is its own thing; it’s a show for the Go community, it has its own episodes, it has its own stuff. And what we’re trying to generally track is the success and growth of our portfolio shows…
So what I mostly wanna know - and this is hard to get even inside Google Analytics, but I think it’s a pretty basic thing, is I wanna know the performance of paths. If I can just see /gotime and everything underneath it… Think of it like your top little nav. If I can show just the stats on this path - how is JS Party doing, how is News, which is a whole section, how is our Posts…? Is Posts growing? Not the individual posts, but cumulative, over time - are they dropping? That’s beautiful for me. And that’s so simple – that plus what you currently have is basically BOOM.
Uku Täht: Yeah.
Marko Saric: We are in progress of doing that, no? That’s what’s gonna happen, to be able to dig deeper into these levels.
Uku Täht: Yeah, but what really sucks is I can go into the database and I can run that query…
Jerod Santo: Totally.
Uku Täht: But surfacing that as a feature is something that I’m not very – I’m not a UX guy; we get a lot of compliments on the UX…
Jerod Santo: It does look nice.
Adam Stacoviak: Well, I think if you knew the routes, you could put that in the users’ hands, and say “Which routes matter to you”, instead of saying of saying “Here’s all your routes”, or even top-level routes… Like, “Which top-level routes matter to you - build a dashboard for you on that” kind of thing.
Uku Täht: Yeah, we need to think about some sort of way – I can think easily of a way to add a filter on top, where you select the properties to filter by, which would be the path name, and then add a regex that you can filter by… And then maybe we could have some sort of stored filters…
Jerod Santo: Stored search, yeah.
Uku Täht: And you can have just tabs with different stored filtered searches, or something like that. I can think of it, but building is gonna take me ages.
