Governance is everything: Simo Ahava on making GTM a bridge, not a back door

Speaker A: Foreign.

Speaker B: Welcome to another episode of Couch Confidentials. Today I'm taking another tangent out of the, the Martech realm just to kind uh, of um, share my world. Ah, and this, um, person who's sitting here with me, Simo Hava, straight from Finland, has uh, played a big role in my career and understanding the technical side of collecting data, especially from uh, from websites, uh, platforms like Google Tag Manager. So Simo, super welcome, uh, to the show and uh, thanks for taking the time to have a chat with me.

Speaker A: Well, thanks for having me. It's always fun to talk about things that are fun to talk about.

Speaker B: Absolutely, absolutely. And I have to say, I have to say, uh, I think it was around, uh, I started out with digital analytics and at some point I kept on running into poor setups and that was absolutely the moment I said things have to change. So I got into the technical side and I'll admit, uh, I don't have this little pedestal or photo of you, you here. I do have the dog though, but I don't have a little photo of you, but in a different world maybe, because I have to say you were one of the, uh, kind of the, the key people in the Google Tag Manager community and tag management in general, uh, to uh, be active on the Internet, to share your experiences, share a lot of script examples. How did you, how did you end up in that? I mean, can you tell us a little bit about your background and how you got into the, uh, kind of the ambassador role for the world for Google Tag Manager?

Speaker A: Uh, well, there's a lot to unpack there. But um, I guess that originally, um, you know, I've been uh, geeking out with websites and web design since I was a teenager. So like, you know, mid-90s, late, late 90s. And um, um, back then it was pre JavaScript analytics day, so we had those stat counters and I managed to create a couple of websites that actually, you know, got quite a bit of traction. And it was just so, it was just so infatuating watching that counter go up and click it and see where the visitors are coming from, like what their IP address is resolved to. So that's how, that's how I kind of got into the, into the measurement side of things. It was just a vanity thing, um, seeing how many people visit websites. Um, but just working on the web just followed me throughout. You know, I took my studies and my masters in a completely unrelated topic, so I had to keep myself active somehow. So just building websites and, and working with that. And then um, by the time GA, uh, became free availability in 2006 or thereabouts. I was, you know, I was among the first to, to jump in, um, just because I wanted an upgrade from those stat counters and I didn't have a budget so I couldn't, I couldn't start using Adobe or Omniture or any of the other other complicated tools. And GA just kind of swept me away. And then, um, in, in 2010 or thereabytes thereabouts, I, I stopped my, stopped with my studies and entered uh, digital marketing and, and um, just technology sector, uh, entered it basically. And that's when I figured out that I could actually start turning my hobby, or my kind of this fun little hobby that I had, uh, into, into something that I can actually work with. And then, um, so I started working with implementations of, of Google Analytics. Classic Google Analytics back then, or although it wasn't called classic, of course, it was just Google Analytics. And um, and then I ran into what you said, like implementations suck. Uh, that was just a common thing. And Google also realized, like Google also realized that so many Google Analytics deployments that they see have typos, they have mistakes, they have errors, they're, they're collecting bad data. And of course it reflects badly on Google because people blame the platform and not their implementation. And so in 2012, late 2012, October 2012, when uh, Google Tag Manager was released to public, um, it had been in a beta for a while and I was testing it because we, we worked at a, I worked at a company that had access to the beta. Um, I kind of just saw the incredible utility there. Uh, because for anybody working in marketing, um, already back then, of course, uh, one of the most difficult things to do was to get changes to your snippets and to your pixels. So just trying to get, get the developers to understand, like, what are we doing? Uh, why do you need this? And so Google Tag Manager, um, you know, I, I quickly realized this was not a good way to sell to do Google Tag Manager. But, but back then the cool thing was that you could actually get stuff done without having to go through the IT process. And so that's how I, that's long story short, how I got on, got into Google. But uh, the blogging thing just came organically. I, I'd be blogging the entire time I've been working with web, um, starting in the 90s, starting in, in bulletin board systems actually, um, and blogging in those. And it just came so naturally. Um, um, I figured out a cool solution and I, and I just wrote about it and published it. And that was the time of Google plus. So we had a very active Google Tag Manager community there and Google Analytics community there. And so I started sharing those blogs there and. And then they just took off and then, and then it became more like a. I think I was among the first people to focus solely on that interface between development and marketing. So technical marketing, as I call it now. Um, so Google Tag Manager blogs didn't really exist. Um, there were people who blogged about Google Tag Manager among other things. But I decided early on that that would be my entire focus. And I think it, it then just this very strict focus and the fact that I was every week coming up with new cool ideas to blog about, um, really helped, um, you know, take it off. And, and I think that I'm still, I'm still here. So obviously it had some. Yeah, still there. I mean, I don't blog as well. Yeah, I blog like of what I used to blog. But um, every now and then I still have a nice idea to write about. And then I do. Um, of course now I have a couple of other blogs as well to curate. But. But yeah, that's, that's how I let. Let me just say finish by saying that none of this was intentional. Like this was always a hobby. Um, it was always fun. And I was very, very surprised when I started getting invitations to events and Google started giving me attention and people started asking me questions. I was very, very surprised at that. And there was no, there was no method. Method to it. There still isn't like, there's still. I don't have a set schedule. I don't have any goal for blogging. I've decided that as soon as it feels like a chore, I'll stop doing it. And I've really, really slowed down over the last couple of years just because it started feeling like a chore. So it's just, it was good timing. I think that's, that's how I would, um.

Speaker B: Yeah, I think so too. And I think it's a healthy curiosity what you showed. And um, it's interesting to say that you kind of got into this valley, uh, between technical and marketing and you opened up a whole new. I think in a certain sense you at least contributed to a whole new career path for many because a lot of people weren't actually considering sitting in the middle. And I still experience that to this day is, you know, especially when we talk about customer data platforms or even digitrant. Where does it sit? Does it sit with marketing or does it sit with uh, uh, with the technical side? And I remember the, the early days of uh, of Google Tag Manager. I was working with a tool called tagman at the time but the, the same premise was there. You get one JavaScript container onto a website, give it a UI where you can inject other scripts in and uh, and do magical stuff uh, stuff with it. It was, it was definitely an eye opener and uh, I think some of the wild western days, um, I remember just, yeah, like you said it was dangerous on one end. I actually um, this was before I actually knew JavaScript. I wasn't clear on what an alert was so I started testing with a uh, with I believe Tag Commander at the time this client was using Tag Commander and I wanted to see if I could scrape product data from the website, put it into an object and

Speaker A: then send it off.

Speaker B: But I was testing with an alert instead of console log and I'd accidentally published it. So it was live for about an hour as these kind of things. It's so much power to put into a tag manager. Have you. I um, mean I'll get back to the community part because I think that that's really what defines you. Um, but governance, does that play a huge role for tag management? Do we have taggers? We have State and a lot of their messaging is around, hey, use our adjusted version of the tag and you can bypass things. Where does that responsibility in your opinion sit in terms of managing a tool as powerful as a tag management solution?

Speaker A: Um, well yeah, it's. Governance is everything. Um, and like you I quickly noticed that this is ah, this can be a very dangerous tool in the wrong hands. Which was actually incentive for me to continue blogging, um, trying to kind of introduce those best practices. Um, but yeah, so, so tag management systems have always had this weird dichotomy to them that they are very, they facilitate uh, technical marketing, they let you do things that would otherwise be difficult but at the same time those things are typically difficult for a reason. Like there are reasons why it won't just on a whim really publish what you ask them to. Like there are reasons for that. They have processes, they have uh, certain build processes in place that they need to evaluate and they know it, knows that if something goes wrong, if there's a data security breach or a data protection breach, they are the ones, I mean like in the first line of getting the blame. Um, so of course they are skeptical about stuff like that. And then Google Tag Manager comes along and it's, it's really like A backdoor, uh, to, to implementing all this stuff.

Speaker B: Trojan horse.

Speaker A: Yeah, so, so, you know, I quickly, at least I tried to pivot, um, away from, you know, everybody just use Google Tag Manager because it's so cool to. Let's try and make Google Tag Manager, uh, the tool that unites your organization, that brings the developers and the marketers together because developers can actually appreciate what Go Tag Manager does because it offers a standardized way of deploying, tracking and tagging. Um, but it's of course it's the same thing that I'm in with server side tagging now, uh, where server side tagging seems like an amazing way to bypass all sorts of stuff. And the companies that you mentioned do advertise this as their usp. And from the beginning I've been trying to push this narrative that it's not about bypassing, it's about taking control. Um, and bypassing is actually bypassing user agency, bypassing ad blockers, bypassing tracking protections. Um, again, we're back in this, in square one with the governance problem where these tools might look like they were built specifically for that purpose for bypassing stuff, but they don't have to be used for that. And in fact, if you do use them for bypassing or for taking over IT processes, you are building a reputation for those tools that will make it all the more difficult to deploy in an organization that actually cares about these things. So it's a very difficult message to push, especially because I've become known for being such a strong proponent of tag management. And then it seems like sometimes I'm the only person shouting into the wind saying that okay, tag management is cool, but you have to use it responsibly. It has all these cool ways you can use it, but some of them are almost strictly black hat. Um, and I just, yeah, I'm very disappointed often in our communities, um, for spending so much time trying to come up with workarounds and ways to override user agency and ways to just get more data. As if just having more data is somehow a grail thing because it's inconsequential. If you get 3% more data, what do you do with that? Ah, who cares? Yeah, it's such a weird industry.

Speaker B: No, it is. I remember, uh, when gdpr, uh, went live, we had to make changes to Google Tag Manager to make sure it was fully behind the cookie wall. And then I think it was two days later, the marketers, they saw the drop in performance GDPR at work. We were respecting people's consent. And uh, it Went up the chain of command. They said, well, uh, we need to revert back as it was before gdpr. And I was like, yeah, well that kind of defends, yeah, that kind of be, uh, the purpose of this whole entire GDPR thing. So people were felt threatened in the work that they were doing since they were being assessed on their performance. And if their ad performance dropped, so did their performance. And I mean even I read today Shane was uh, fined 150 million for placing cookies that they shouldn't have. And in the US there's a rise of scalping, uh, lawyer agencies, or at least lawyers, uh, that just all they do is scan websites to make sure, hey, are you adhering to uh, the laws? And it's a difficult position to be in, as you say, um, if you have a rational mind, you know, you shouldn't be doing it. And it's always a tough discussion that uh, you probably have with uh, clients as well. Is it? I mean, you mentioned the community there. The community has definitely been a driving force, uh, for you and for Tagman. If you, if you look at Tagman or if you look at uh, the products that Adobe was serving up, it's sure they, they have communities, but nothing as big as Google. How have you, uh, like you said, you've been invited to, to speak with Google and I think at one point you were even, you were advising them or you were, uh, you were on a special program with, uh, to be pushing Google Tag Manager.

Speaker A: Yeah, the developer expert program. Yeah, yeah.

Speaker B: Oh, that was it. Yeah. What did that exactly entail? And is it. I mean, I personally haven't been following most of the changes within Google Tag Manager, but seems to be kind of slightly dormant.

Speaker A: Yeah. Um, so, yeah, so the community, first of all, um, has always been impressive. Um, like I said, there was already a big existing community around measurement in Google, in Twitter, um, and of course all over the place before those technologies or those services. And um, so I did enter, um, a place that was already well established. And the cool thing about the community, at least like until the demise of Universal analytics, was that it was all about intervention and people were just sharing cool ideas that they had. It was all about building and people were really encouraging each other. Google was very active in those communities. They were, um, positioning themselves as open to feedback and you could even see was quite electrifying because we might have a discussion in Google around, uh, it would be cool to see this feature in Google Tech Manager next release. We suddenly see that it's there because they were listening in the forums. So it was almost an immediate um, feedback loop. Um, you know as you said I'm sure all the other platforms had their own forums and of course the discussions we had were universal. So uh, we could have a very fluent discussion with Adobe practitioners about tagging and tracking as well. But Google was dominant because it was free, um, quote unquote free. Um, but um, so that was. The community was certainly very uh, very quiet. It's such a shame that we don't have that anymore. Google uh, has all but removed themselves from community interactions, especially since they killed off um, the evangelist team uh, of Daniel Weisberg and Chris Poseidon and Adam Singer and Louis Gray and all the others. That was such a wonderful way uh, of them interjecting themselves into communities. Um, but they don't have that anymore. So I think that uh, and that now we have Measure Slack, uh, which I think is the still like the biggest gathering. There's over 20,000 of us now, uh, a free to join. Yeah.

Speaker B: And it's a free Slack even So I mean, ah, you've got 20,000 people talking to each other on a free Slack account. I mean I don't even think it's paid. I don't even want to know what the bill for that is. Salesforce, uh, could uh, be, uh, could be, yeah. Could add another zero behind their revenue just off the measure Slack alone. But you still play an active role in that.

Speaker A: Yes, I do. Yeah. It's so it's premium Slack but uh, you know we don't pay anything for it. It was sponsored to us. Yeah, it is premium and it was, it was sponsored to us by Slack themselves. So because it was such a, I think it was because it was such an active community and such a big, you know, it's of course a great promotion for Slack. 20,000 people using their tool and seeing how good it is. Um, yeah, I am very active, uh, and also a moderator there. And it's like we said, it's free, um, and we can add to the show notes how to apply to join but it is right now like the de facto. And the cool thing about measures like is that it kind of brings people together from all different things. It's not just about Google. There's Adobe, there's piwick.

Speaker B: Vendor agnostic.

Speaker A: Yeah, it's completely vendor agnostic and such a lovely space to be in. So that's now I think um, where the discussion mostly is. Uh, of course LinkedIn gets its fair share. But LinkedIn is just such a, it's, it's Such a terrible, terrible um, place. It's such bad energy. Most of the time. AI generated slop.

Speaker B: Yeah.

Speaker A: But unfortunately it seems to be the only existing marketing um, channel. From a social media point of view, that makes sense.

Speaker B: Something new at one point.

Speaker A: Yeah, I think so too. Yeah. And now I of course forgot already what your question was. But um, when I switched down the memory.

Speaker B: Yeah, no, just around the community. But you did touch about the AI slob and I'd like to go into that because that was one of the original reasons I reached out to you because I uh, think you posted uh, or you replied to an article uh, around that AI will not take uh, people's job. There's a lot of fear uh, in that lower uh, level developers will lose their job to AI, et cetera. But at some point you're going to get this gap. If you ignore all the junior developers now once you know the generation moves up, you're going to m, get a huge gap in um, uh, expertise. I haven't really been following AI within uh, analytics and tag management. I think the only solution that I found so far was Piwick Pro and that you can actually, I think they demoed it. I don't think it's live yet but what they told me at one of their um, get togethers was you'll be able to put in a URL for your website, it'll scan it and it'll kind of determine what the best way is it would be to tag it. Are there any other kind of new features that you've been hearing about or seeing, uh, in terms of AI, uh, and then we'll get back to the question about you know, the importance of humans in that, in that sum.

Speaker A: Well there's a bunch, I mean um, Google Analytics released an official MCP server uh, not so long ago which lets you kind of do agentic stuff with Google Analytics data. So you could do um, you know, uh, run AI agents through Google Analytics APIs. The quality is iffy. Uh, but of course it will get better. State released an MCP server for Google Tag Manager which would technically let you run the programmatic API for Google Tag Manager using, using chatbots and agent prompts. Um, so you know, I think uh, and I'm definitely not in the deep end with AI. There are people far more well read than me, uh, able to talk about that. But my feeling from being active in the communities is that there are certain things that LLMs and generative AI is very good at when it comes to technical marketing and digital marketing. For Example, writing documentation, um, building structured data constructs such as Google Tag Manager containers. It's well documented, it's structured JSON. There's no problem in doing that, um, doing diagnostics, um, trying to give, giving you ideas how to improve an existing setup. There are things it's not very good at, but it tries to do a good job. For example, scanning a website, um, for patterns, offering tagging ideas. Uh, so idea generation around digital marketing, it's not always good, um, simply because the whole idea of measurement, tracking and tagging is so tied to what the website historically has been doing, what, what its market is. You can't just generalize these things, uh, but you can get a nice bullet point list of ideas. And then there are things that I think where it's severely lacking, which is, for example, if you try to delegate actual analysis to an AI, just dropping it some numbers and telling it to generate insights, it's just not very impressive. Um, and just this idea of delegating marketing processes to a machine, um, sounds like the best thing the machine can do is just be average. Uh, when you would in marketing certainly try to look for those breakthroughs as well. So, um, I haven't been very impressed. I think there's a lot of things it can do, especially around those menial development tasks like uh, documentation generation, uh, uh, project organization, structuring, uh, building those method, um, signatures for API connections, for example. It's very good at that, um, and I use it liberally for those kinds of things. But I have, um, been quite disillusioned with its ability to actually generate application code. Um, I've been doing this quote unquote by coding a bit, um, and I find that as soon as I get past the first couple of releases, I just have to jump in and start coding myself because then it starts making like mistakes. Um, so I'm not an agnostic, um, but I'm certainly not in a hype train. Um, and I think that the more time passes, the less impressed I am with the iterations. And it almost feels like generative AI is plateauing already. Um, and we've kind of seen its limitations. But uh, I'm a very bad futurologist in this sense, so I'm sure that somebody will prove me along quickly.

Speaker B: Oh, you're called Simo, not Nostradamus. Ah, exactly, Nostradamus. I'd be asking you for stock at advice too, but, um. No, no, I hear what you're saying. I mean fire coding. Uh, I've tried two applications now. One to generate Synthetic data to push to endpoints like uh, uh, Google Analytics through measurement protocol. Um, like you said, you can get very far, but it feels like you've got the finish line in sight. And then at that point it just stops. It doesn't want to take you further. And um, after I'd finished one application, I said, okay, let's do a security scan now. Uh, we had API keys exposed there. We had, I mean there's so many things just doesn't take into account which can, yeah, take you over or create such a risk for you that it can definitely, uh, endanger your business. So one of the things you mentioned was the kind of the MCP for Google Tag Manager. I can even imagine there's. I mean, to what degree can you actually push that, uh, the trust in the MCP to. I think it would make changes or configurations or maybe add tags, uh, et cetera. To what degree can you really trust what it's outputting? And uh, I think that's what really scares me. So your argument is still humans in the loop or humans in the lead kind of thing? I think those are some of. I think human in the loop is one of the big terms you hear.

Speaker A: But yeah, well that's how I would approach it. And I would personally not delegate any processes strictly to AI. I would always have a human in the loop. But that's my own approach. Um, and I'm very careful, um, not to change it without good reason. Um, and I just haven't been that impressed. Um, but it's also because, uh, maybe I have higher expectations when it comes to Google Tag Manager. I really do expect that, um, there are just so many little catch alls and so many like catch 22s and so many little gotchas there that the AI just doesn't navigate around and it tends to generate quite lazy outputs. And then we have this funny phenomenon where, um, just another example of how AI is messing up with community. You know, you have people on the other hand coming to measureslack or LinkedIn and saying, I asked ChatGPT for some um, JavaScript code for could. Could somebody please look through it to see if it's okay? So why would you do that? Like why, why would you go to ChatGPT in the first place? If you then need to go back to the community and ask for help, why not just ask directly from the community when you'll get like verified experts saying this is exactly how it should go? So I think with ChatGPT or, or any of those other tools, the problem is verifiability. You know, it's, it's generated from a corpus. It's not generated by somebody who knows their stuff. Yeah, so there's a, There's a stark difference there. Um, and then, um, the uh, other example is just um, um, when I audit a new container, for example, I can directly see which bits of code were generated by Genai. And not always in a positive sense. So that's another thing. It leaves a certain signature. Um, and do you really want to leave such a breadcrumb trail, uh, behind you that you use AI to generate these things? Because first of all there's a risk that it's wrong. And second, you are just basically making yourself redundant, uh, because the customer themselves could just have used the AI for the same thing. So it's just, um. Yeah, um, uh, I guess there's like a Venn diagram of what AI should do and what a human should do and the overlap is just um, becoming bigger and bigger, uh, for the human side to overtake the AI side. Um, but we are strictly talking about technical marketing, tagging and tracking, which I think is a, it's just a realm that has so many offshoots in regulation and compliance, um, client side performance, things that you just don't want to mess with. But then there are other realms, um, where AI could have a much bigger, uh, role. Um, this junior developer thing for example. I think there are, um. There are definitely many people working in digital organizations who should rightly feel at risk, uh, not because um, AI is taking over the roles, but because they are allowing the AI to take over their roles. They are working on such things that you shouldn't really be spending time on anyway, like um, you know, setting up projects or refactoring code or writing documentation. Like it makes no sense to spend time on those things. You should focus on the, on the bigger level stuff, the higher compliance, the higher complexity stuff, and use AI to um, take that workload away from you. So rather than feeling threatened, think of it as somebody who's actually allowing something that's allowing you to focus on all the right things.

Speaker B: Yeah, that's true. I mean it's uh. Now that you kind of point that out, I guess in a sense. So the way I use AI, uh, um, is, uh. And I try to be very open about this. Um, I feel like my brain is always on. I'm having like 20 different discussions at the same time. Oh, this is an interesting topic. I need to remember to tell my wife that or I need to make sure that Dog Gets them at the medicine. And I, I feel that AI in a sense is helping me structure all that information uh, in a, in a better way that I can uh, as you might notice while I'm discussing here, I'm thinking while I'm talking. So the words that come out is um, they seem to go all over the place as well.

Speaker A: The train is already on the next station.

Speaker B: Exactly. Yeah. And, and, and, and yeah, that's a good point. I like that one. Train's already at the next station. But what AI helps me does is really sit down and focus. So what I uh, use a tool. I think it's called Super Whisper or Whisper. I don't know what it's called but I just talk to the computer. Yeah, um, no, no, pretty cool tool but um, um, it's based on whisper from OpenAI and I just talk to it and then it will just structure my thoughts for me. And I have to say that that is helping me to focus while I'm talking. It'll split out different topics and say hey, you were talking about this, you were talking about. And it's uh, yeah, there are good applications.

Speaker A: One thing I, it's a good PA

Speaker B: noticing though is that. Yeah, like it is. Yeah. I, and even if you use tools for akiflow, I use that and I can actually talk to now say hey, I need an, I need some time for a meeting on Thursday. What's my availability? And it'll do, it'll do that. I don't know about you but I always, the great best thoughts I always have is when I'm not behind the computer, I'm in the car, uh, I'm walking and uh, I need to talk to someone. So I just put the voice to text and I think uh, uh, that's absolutely brilliant. So Google Analytics, I recently heard that Google Analytics, uh, they don't want to call it Google Analytics 4 anymore. They just say this is just uh, ga, uh. And like I previously mentioned Google Tag Manager, it feels like it's kind of dormant. Um, a few months ago or maybe a year ago there were some discussions like with people, people from, I don't know, I think it was ah, Rick Drunkers, um, who was maybe pushing a narrative or suggesting a narrative that Google Cloud in combination Google Tag Manager was pushing towards a customer data platform. I just want to see if I can pick your brain about. Have you heard anything on the grapevine um, that kind of verifies what Google's trying to achieve now with Google Tag Manager and Google Analytics, like the the demo team's gone or the. What's, what's happening in the space?

Speaker A: Uh, that's a good question. I don't really know. Um, uh, so Google Tag Manager, it hasn't been dormant like they push releases in server side GTM quite frequently actually. Uh, what we are missing, what we've been missing for a long time are those small incremental updates. It doesn't feel like Google Tag Manager has been getting any new features really um, other than uh, you know, related to Google Tag which is a bit of a shame. Like there's a long list of things that I would like to see in Google Tag Manager and I don't, I just don't feel like they're coming there anytime soon. Um, and then there are bugs that are Almost you know, 10 years old now that haven't been fixed. But you know that's kind of understandable uh, because Google Tag has overtaken M, the Google's marketing organizations. Like the, the only thing they are focused on right now. Um, kind of understandable. They want to consolidate the technologies. Google Analytics certainly gets a lot of love. Um, it's, and it's improved. Um, I, uh, those that have been following me know that I've been very, very critical of Google Analytics over the past years. Um, it's still, I still get a icky feeling when I open the ui. I operate almost totally in bigquery and data form now. Um, but as far as CDP goes I don't think, I don't, I don't think that Google itself um, is pushing forward with any related technology. But what they are certainly doing is opening up the APIs and opening up the connectors and improving the connectors to make like community driven CDPs around Google's tagging technologies a reality. So server side GTM itself can be used as both an ingestion and a throughput system for uh, channeling data from CRMs from um, transactional databases. It can certainly be used for that already in a very, very CDP type, uh, of type of way. Um, Google Analytics itself is building integrations directly with um, third party platforms that might allow them to better. Yeah, exactly. And there's going to be others as well. I'm sure that will let it make it easier to do like audience transfers uh, from one place to the other. I think just the deeply embedded skepticism of um, kind of relinquishing your customer data to Google is what's holding them back. And I think there's a lot of like regulatory problems down the line in case they would Start hoarding customer data themselves, which they of course do already with all those identifiers. But turning into a cdp, for example, I don't think that's in the horizon. I think what they'll prefer doing is to become a key component in um, a customer cdp. Um, so that would be certainly something that they're probably facing. But yeah, that's also one of the reasons why I haven't been blogging that much is because they just haven't been cool updates that invite me to get the kind of wow, I can't wait to tinker around with this. Um, um. So it's, it's just that you've actually

Speaker B: been teaching that, that, that as well to uh, to produce courses.

Speaker A: Yeah, yeah, that as well.

Speaker B: This is definitely not a plug. It's. I'm not advertising here. I did, I think I, I did three, three courses with you. What I, I think it feels like a natural progression from your blogs to be, to go into uh, uh, into building these, these uh, courses. Um, you recently released a whole new bunch or at least updated um, courses. Uh, at least for the uh, bigquery, uh, course I believe with uh, Johan

Speaker A: Van van der Berk.

Speaker B: Yeah, yeah.

Speaker A: Um, well it was kind of natural. Um, I have been doing trainings uh, both online and in person through companies like 121Watt in Germany and um, um of course training my clients. And, and I guess at some point this uh, was covet times when companies were. This was literally the first Covid year when companies were stopping uh, in person, in person trainings and live trainings and wanted to move online. And um, my wife was stuck in a kind of a dead end job and I was kind of getting very bored with consultation or just you know, the kind of consultation that I was doing. And so we just, we thought that let's just. You know I've been doing online courses for CXL as well. So I know how to perform in front of a camera and how to build those um, materials and we just thought with my wife that uh, hey, let's set up, let's just do a quick proof of concept and see if people are interesting. A server side tagging was just released. So it seemed like a good um, good, good first course and then we published the course and it took off and then we just decided okay, this, this looks like a. So it was again a very good timing. Same with the blog. It was very good timing because we released a course with a technology that had been previously unvisited. Uh, nobody had been doing anything with it because it was just released and that companies were slowly moving to, or quite fast actually moving to online trainings. Um, so again it was very, very lucky timing and it's been a lot of fun. It's hard work, uh, trying to come up with, with courses that um, people are actually looking for. But server side tagging and the bigquery stuff and the JavaScript have certainly been the kind of evergreen topics that uh, I think many people are looking for. So that's all happening on teamsimmer.com where we, and we also like the big push a couple of years ago was that we released this free, uh, kind of course for technical marketers called the Technical Marketing Handbook, um, which started again as something that, hey, we've got this extra capital floating around. What should we do with it? Well, let's take time away from actually generating more capital and let's invest it into building a resource. Um, yeah, so it's, it's, it's handbook, um.teamster.com which, it's like an online manual for technical marketing, goes through all the different disciplines. So if you're like entering digital marketing for the first time or if you want to make sure that you have a good understanding of the different technologies, it's a, it's a good place to start. But that was just a very fun experience and I'm experiment, I'm very, very proud of it. Um, even though it's not directly generating anything for us, we don't, we don't push anything through it, we don't force

Speaker B: you to subscribe or anything that's worth something. Yeah, but how do you feel the con, I mean, what kind of people are using this? So I, I work sometimes with uh, with startups and I feel that a lot of this technical experience is, is missing. Do you see that back in the, in the data that you're collecting from the people who sign up for. And I mean what kind of topics are the highest in demand? If you maybe look at the course topics that are uh, uh, that are in there.

Speaker A: So, um, first of all, we don't actually do analytics at Sibir. Um, we collect some data, uh, but we rarely use it. Uh, because that was also an experiment. Um, can we build a successful brand without actually doing anything with data? And we did manage to do that and we didn't even do any advertising. Uh, only very recently did we.

Speaker B: It's like a chicken egg story in marketing. Can, can you, can you market without the data or can. Yeah. Oh, that's, yeah, exactly.

Speaker A: So we tried to be and it was so, so based on personal brand and our newsletter. Um, and we did an experiment with Google Ads actually just this summer and it just didn't work out. Like we're too niche um for. And, and we've we've chosen some very unfortunately popular topics like Learn BigQuery and even our brand name Simmer um is a popular like doordash kind of uh, food delivery service uh in, in the U.S. i think so it's um. So we haven't made things easy for us but, but we, we I also um. So, so we don't look at data that much but what I can say like generally is that well almost all of our um, buyers are companies or business. Business to business. Um, partly because the VAT rules in EU because VAT is added to if a individual ah, buys a digital service then their VAT is added whereas companies get a reverse charge. So it's actually cheaper for companies to buy those courses. Um, I know, sorry. Ah, yeah, yeah, it's complicated but actually fascinating. Um, um, the majority of our buyers is from the US Um and there's certainly like when I look through the names of companies that people at it's big uh digital marketing companies that, that want to educate their, their students. But in terms of like um. And then of course we know what's most popular where the BigQuery courses are our most popular courses. Server side tagging is a close second and then the rest come after and based on how much people ask questions those are the things that are, are the hottest topics. Um, yeah but they're also kind of frustrating to do courses around because they are third party user interfaces. So anytime Google makes changes to them we have to be very careful to see if our content becomes outdated. And it happens maybe once a year that we have to go back and redo a lot of videos which is time away from anything else. So that's why I actually preferred for example our JavaScript course and our CSS selector course and our DevTools courses are really fun because they're almost evergreen material like those interfaces can change but the fundamentals stay the same. So hopefully um, in the future we could focus more on evergreen content simply because it's still, it's still a two person company. It's just me and my wife so we don't have the luxury of every day um, you know, working on 10 different courses and having like a full pipeline. Um, so it's out of necessity.

Speaker B: I know there was this uh, um, there were two German guys they were doing programming courses for JavaScript and they did a few and they React and they were quite thorough and they saw all their stuff on Udemy. But the thing I question, and this is just for me personally because a catamind is what they're called. Um, but they throw out so many courses and I actually did the VUE and the React one to be able to work with them and then do the tracking. But I think is important to know for people watching or listening is about for simmer. Um, what I see with a catamind is it's very high level. This is the program. There's little context to business. They're always doing the same kind of tools, uh, uh, what do you call it to do list kind of things. And there's hardly ever any context. And that's I think like you said, it's a niche and it might be a small market, but I think the people who are taking these courses benefit the. Sorry, benefit uh, the most from it because there you can't find anything else. Technical, uh, marketing is great, but how do you apply that to your, you know, to tag management, to data collection for further down the road. Now it's pretty cool. I didn't. I thought you guys were a little bit bigger than two. Uh, but I think you sometimes work with uh, with other authors like Johan. Uh, but uh, yeah, even with that, the Evergreen story, changing user interfaces, uh, it's still pretty impressive, um, the quality of the courses you put out there. I know we have a hard deadline, so I want to take this moment to thank you for your time and uh, yeah, it's been interesting to hear a little bit angle from uh, where tag management is. Been out of it for a while, but uh, yeah, it was good talking to you and I wish you all the best.

Speaker A: Thank you. Thank you so much. It was fun talking to you as well.

Speaker B: All right, and let's uh, plug it. Well, we need to plug your own podcast of course.

Speaker A: Oh yes, yes. So, um, uh, uh, actually there are two. Uh, but let's start with. So the, the. The act. No, actually I had a one for a long time and then that's a bit dormant now. But it's also called the Technical Marketing Handbook and you can find on teamsim.com we have. I had two seasons, I think. Or maybe just one. I can't even remember, it was years ago. But I'm, I'm actually thinking of going back to it. Um, maybe turn it more into like an audiobook format, uh, around technical marketing topics. But then the active podcast is called the Standard Deviation Podcast where Juliana Jackson is the host and I'm just tagging along the ride. Um, so we are, uh, we are, um. Active is maybe the wrong word because we maybe push an episode out maybe once every two months now. It's been a very slow year for both of us, so, but it's full of brainless banter and, and insubordination. And we're actually doing a live show in Hamburg at ah, the um, Track and Analytics Summit. Um, it's the first and will probably be the last live show that we'll do. But it's, it's fun to give it a, Give it a test as well to see how.

Speaker B: Never say never. I mean, as long as you avoid certain topics, I'm sure your, uh, your uh, uh, your show will not get banned or uh, get. But live shows that. The live shows can be pretty fun. So is that going to be on stage or.

Speaker A: It's going to, uh, be on stage. So the reason I generally myself hate live shows, I think they're noisy and it's difficult to follow along and they're full of like, um, visual cues that you can't get through just listening to the podcast. So we'll try uh, to do it in a way that doesn't get other people to hate live shows as well.

Speaker B: All right, leave some room there for me in the future then. But uh, hey, thanks for uh, taking uh, the time and I'll, uh, see you around soon.

Speaker A: See you.

Speaker B: Bye.