Growing Ecommerce – The Retail Growth Podcast

The State of MarTech with Daniel Smulevich

Smarter Ecommerce Season 3 Episode 6

Learn about the future of marketing from Daniel Smulevich of Jellyfish, as he guides us through the MarTech revolution in a landscape brimming with change. This episode isn't about chasing shiny new tech; it's about strategic problem-solving and making tech work for your marketing needs.We dissect the tech FOMO that grips the industry, using platforms like Roblox as cautionary tales, and emphasize the importance of thoughtful tech adoption. Daniel describes the imperative for realignment of data and analytics roles, which need to shift from mere tech support to strategic business advisors. We also discuss choosing between creating your own technology or investing in off-the-shelf solutions. either way, you must ensure that your decisions are informed by deep industry knowledge and the right balance of manual and automated processes.This episode is a must-listen for any marketer looking to stay ahead in a dynamic, data-driven world.

Speaker 1:

Welcome to Growing Ecommerce. I'm your host, Mike Ryan of Smarter Ecommerce, also known as SMEC. Today we're joined by Daniel Smulevich. He's Executive Vice President of Data and Analytics at Jellyfish, and we talk about the state of MarTech, the adoption of AI, build or buy decisions. In that context, we talk about the ways that measurement is and is not changing, the ways that the death, the agonizingly slow death of the third party cookie is incredibly drawn out boy who cried wolf-ish and in other ways, it already happened years ago and we're just feeling the pain of a phantom limb. Remember, if you enjoy this podcast, please leave a review and share it with friends, co-workers, on your social media networks. We really appreciate that. Okay, let's get into it. So, Daniel, thanks so much for joining us today, my pleasure. Would you get us started with a quick introduction to yourself? What are your skills? What themes interest you?

Speaker 2:

Sure. So I am in data and analytics at Jellyfish, I'm the executive VP and what I do? I focus on data analytics and the intersection with ad tech and CRM, which is pretty much what the whole of my tech is these days. It's a bit tricky to say about skills and what interests me, because I guess the red thread throughout my career is that I just like solving problems. What I found that in marketing increasingly at a rate I did not expect when I first started in marketing increasingly marketing is my tech. So where, despite the fact that I actually started with a marketing management background as a degree, eventually got more and more interested in the tech to support those marketing outcomes. And I guess what I've always tried to stay true to is focusing on the marketing problem rather than the tech, and that's where it all comes together and I think the intersection of the two is what I'm the most passionate about.

Speaker 1:

Okay, yeah, maybe we can get back to that Like focusing on putting the problem first instead of the solution first. I think is good advice, like so you mentioned, yeah, you've been in data and analytics. You've been doing that for nine years and you've been I mean, I think you've got at least nine years in data and analytics. So how has that role evolved? And you mentioned that marketing and technology have kind of married or the story they've become one story in a way. But how has your role evolved? And tell us more about that evolution of the landscape?

Speaker 2:

Yeah, and I guess also you know, as a data analyst I can tell you my view is slightly biased, has a selection bias in terms of also being quite junior. When I started just over, I think, about a dozen or so years ago in data analytics, before Jellyfish, and eventually at the start it was more on working on the solutions rather than the problems, and it's a bit of a digression. I've always found it a bit wrong the other way around and I felt that the more junior you are, the more you should actually focus in learning around the problem and challenge yourself, rather than diving deep into the tech and then try to look for the problem that the tech solves later because it takes so long to unlearn. I think that's something that consultancies do really well, like the Deloitte extension of this word a problem thinking. I think in the rest of the industry we obsess a bit more on the tactics and solutions rather than the problem. And why that digression? Because I haven't really seen the problem changing much in the last thousand of years or so.

Speaker 2:

The tech has evolved a lot. The one thing I did see changing is that I felt when I started and then again I don't know if it's my bias there is there wasn't necessarily a clarity on what was possible, so you were playing with the technology and often imagining things you could and couldn't do, and there wasn't necessarily a shared understanding in the category of what was technically feasible and what you could do. So you felt like you were stretching it and often, again, you had the solution, but not really a problem to solve for. But there was that excitement in trying and finding that. I find that over the years, what changed the most, especially in Europe, but I would argue most of the technology industry that came along is, I think, the level ground of what actually can be done is fairly understood by most, something that I don't remember seeing that a while ago. So now it makes even more sense to focus on the problem rather than the solution, because you know what arguably can and cannot be done.

Speaker 2:

The challenge is how you apply and that's where, again, I try to focus more and more of my time on really not saying we want to increase profit as a goal, but say, okay, what levers do you have at your disposal? What does that that mean to you? What is the action you're going to seek to drive? Who is going to do the action, what do you have at your disposal, what is already available in the tech you're using, and so on and so on. So a bit of a long answer for a short question, but I guess my role has become more and more of a true advisor and consultant to businesses, in a way that I think a lot of the media folks are currently changing their role and evolving too, because first, 10 years ago, even more senior folks were just configuring and changing a lot of platforms, and now it's more about the strategy and the thinking. Do we do PMACs, do we not? What is the split between that and the other?

Speaker 1:

Okay, I've got maybe a follow-up on that, but I do really appreciate this discussion about focusing on the problem. Like, I come from a product management background and there's kind of a truism that typically teams are spending like 20% of their time thinking about the problem and 80% of the time working on the solution, and it should really actually be the opposite, just focusing a lot on getting the problem right and then thinking about the solution. So I really take that note. I appreciate that. And now, yeah, we've talked about changes occurring in Europe and also you're talking about businesses starting to move in more of a consultant-like role and getting a little less technical Something maybe we've discussed on this podcast before.

Speaker 1:

But I feel like there is this kind of advice, this sort of generic advice like oh, ai is happening and platforms are becoming less fast in some ways, the things that we're supposed to do there. In some ways, there are fewer levers to pull. As we lose user level targeting over time, there's going to be a loss of really detail of the things that you can do and we're all told to move upstream. But the thing that is left unsaid there is I always have to think of fish swimming upstream, and they do it for two reasons. They do it because when they lay their eggs up there, then the baby fish only have to swim downstream. But it's also kind of like a built-in selective pressure, so that only the strongest fish are going to make it up there and lay eggs in the first place. And I feel like we're all being asked to swim upstream and it's definitely a selective pressure.

Speaker 2:

It's not like I think there's less room upstream, isn't there? Yeah, I think the challenge you have is, effectively, ai is in everything. Right, and I mean it's now a bit more of a buzzword in the last couple of years thanks to a subset of it, but at the end of the day, the majority of the work carried out in AI realistically is still the same ML type of workload that you had five years ago. They're just a bit cheaper to execute. You have more standard libraries, it's more integrated into platforms, there's more training, data sets for that to work better. But I consider and I try to say that AI, at the end of the day, is your marketing technology, because I challenge anyone to look at their enterprise tech and say that none of that uses AI. You find in Google, you find in Meta, you find in all CRM these days. You find in your analytics platforms arguably nearly anything you can think of analytics platforms, arguably nearly anything you can think of. So I share the thought that, hey, we have so much to focus on and learn and understand, and I agree on the fact that marketers, especially the CMO of this world, they do really they cannot ignore tech. They cannot ignore my tech. If they want to look at AI. They need to look at the enterprise stack.

Speaker 2:

Where I challenge a bit is the thought that we really have to keep up with the latest GPT model, the latest meta model, anything that comes under the sun, because otherwise we become like those people on LinkedIn the first they were crypto experts, now they were metaverse experts, now they were metaverse experts, now they're Gen AI experts. All they spend their time on is just regurgitating and summarizing information already available out there. Yeah, and we already got something for that. It's the internet. You're never going to beat the internet at regurgitating information and summarizing.

Speaker 2:

Where they need to understand is how the tech, how the levels at their disposal, are actually functioning. Whether it's done by AI or not, it doesn't really matter. So my argument to them is understanding any moment in time, having a really good view of the business problem. That's what they should focus on and make sure there is a constant research on the tech that better serves the problem. Then whether it's done by AI or not doesn't really matter. The challenge should be is my customer support bad? Spoiler alert often is Is my product information complete enough? Most likely isn't. Do I have a good differentiating experience online in my retail website? Most likely no, because they all look kind of the same. So, again, those are kind of the worries they should focus about. And then AI is just a lever at their disposal, which is usually embedded in one of the other levers they've always had.

Speaker 1:

I appreciate that. I think it's all around, as you said, crm everywhere. I mean, salesforce has great AI We've been surrounded this for a long time and also there's so much scarification that goes on with AI. Like you mentioned, the formerly crypto, then metaverse, then AI hype, train people and they'll be out there hitting you with messages like if you're not aware of these tiny new changes, then you're falling behind, um, and like everyone else is doing it and you know. You look around and you're like, actually no, I don't think that's probably true, but, like you know, I have a couple of follow-on questions here. Maybe how can, how can we distinguish that between something like this is just going to be a fad or this is a more fundamental change that actually deserves my attention?

Speaker 2:

That's a hard question but it actually has a simple answer and I would say it's not just about fads or fundamentals but it should be applied to anything you do. I've been in now my tech for a dozen years or so and I think I can count, maybe on one hand, the amount of business cases I've seen justifying the deployment of a new tech or an important project, meaning a quantification, solid quantification, of the business benefit that any solution would have. And I think we have a bit of an obsession. As an example, in our industry we're targeting and always focusing on the better and more advanced workflows. You know the classic right message, right person, right time. But with the idea that once we're going to be able to serve some content, whether it's on our own or paid media channels, and expect a fantastic uptake, the reality is we don't. We can most likely serve part of the use case and there is value, again, in a CDP, there's value in personalized communication, but it is never as valuable as some people assume. If you go through that exercise of quantifying and sharing your assumption with the rest of the business on why the value is going to be high enough versus the cost of researching, implementing, configuring it, maintaining it, enhancing, transitioning to a new platform. Then you will find actually, maybe about 20% of the solutions that go out there are even valuable or will even get you to break even.

Speaker 2:

And I see the same approach for me is to any fad. When I look back at the metaverse I go back to maybe it's a boy in 5, 10, 20, 30 years, but let's look at now or the next one to two years. I could never see any data. I know it's always that easy to say in hindsight, but I could never see any data to justify and say this is now a channel where I have millions of people in my markets that I can engage with, that they want to be engaged with, but they have also the money to spend on. So I'm looking at Roblox probably not my target for Gucci or D&G clothes and so on.

Speaker 2:

So again, it is about doing those calculations and then you get a sense of whether it makes sense or not. So when I look at AI now, I go back and say I see an opportunity, for instance, in the world of e-commerce, on having personalized virtual agents, but rather than the classic search functionality. But there has to be a solid case around. Okay, what are we trying to do? Is it about serving a part of our stock or inventory that otherwise never gets seen, for which maybe we have growth and less margin, because maybe we lay on the price elasticity to give them greater discounts? But because we need to get those products out of the door, we need to free up some of the warehouses. So I see as opportunity for additional channels, but there have to be solid cases behind and in most cases it will not be about building it but it will be about buying a solution.

Speaker 2:

Yeah, so like you're suggesting that you know like 80% of MarTech, maybe lacks a clear business case or slash a ROI positive business case. When you're saying that this new software is going to change your business, tell me the assumption in increasing click-through rate, increasing average order value, increasing, whatever that may be, number of products purchased. Just show me the assumption and then run it to the test and see is that even actually true? Why? Why are we saying that it's increased?

Speaker 1:

Yeah, it's a great point and like, yeah also, you mentioned Roblox. Like I think there's just this FOMO that strikes. We end up being very irrational in these cases, Like okay, something very interesting and cool is going on in Roblox, it doesn't mean that it's automatically going to happen everywhere else and that's going to happen overnight. You can name any number of trends, like social shopping is starting to mature a bit, maybe with TikTok shop and so on. Also, there's other problems that TikTok faces, but there's a lot of assumptions that things are just very transferable or very scalable in ways that might not be the case. I mean, you mentioned now build or buy. And you mentioned now build or buy.

Speaker 1:

I'm just curious because there are these massive tools or new platforms coming out, like let's just look at OpenAI and ChatGPT and then there's a feeling that a lot of MarTech that's being built on that are just UI wrappers for that. Maybe they've done some kind of like you know, they've made their own GPT or something like this, but they're not really training new models or anything. They're just doing some, you know, feeding it some PDFs or I don't know what. Like it's not clear to me right now when we're looking at that kind of technology, like you mentioned, in agents on the website, for example. Is that something that we should build or buy? Website, for example is that something that we should build or buy? Because I find it right now really foggy, like how much of that is just a UI layered on GPT or how much of it is really something genuinely being built.

Speaker 2:

That's totally fair and I would say you have two opposing trends. Yeah, from one side you have platforms ingesting as much within themselves. So I would say, if you're in the world of media and you're looking at assets outside of media, probably the space for building becomes smaller and smaller because arguably the platforms have an interest to do it themselves and then effectively there's very few that have to differentiate. So the argument when you're looking in that space, you're less likely to find simple wrapper just around one of the APIs, whether it's for the creation of assets and so on. If I look on the experience instead, that's where things become a bit more interesting, because platforms which have eaten a lot of the marketing budgets of the years they have not had really a play on the experience side. So what happens on your own website and apps and internal systems? Here I would argue that it really does depend on one factor how that is embedded within your other systems that are using for the experience, because, for instance, there's fantastic customer service software.

Speaker 2:

I really like what Zendesk are doing and the work, the AI functionality they bring into that. I look at how even Salesforce is doing, but if I think of the Shopify Magento of this world, I see them adding more value and bringing those into the ecosystem, rather than again you building it Because you're using already most of the rest of the software on the side. So to me, the build, even if I look at the virtual agent example which, again, it shouldn't need to be everywhere, right For instance a bit of a digression I don't, actually I'm not a fan of some of the functionality that are GNI-driven that you see popping out everywhere, because they have no value. Like you see on LinkedIn, after every post that little star is on. I'm not sure it adds really much value. If I look even at virtual agent, it probably adds more value in an app where I am solving a problem maybe of the customer that says, hey, take a photo, show me what are some of the products, what can I see that I can buy in that manner. So it really has to be solved around a particular problem, particular entry point. You have.

Speaker 2:

To me, the builder is about just providing. It really has to be something meaningfully different and I suspect for everything that matches digital marketing it's highly unlikely that the benefit is worth it. So again, I'll give you an example. I see now companies that they are giving you Gen EI-based chatbot functionalities. We've seen the media space. On the experience side, you are most likely better off sharing your business data with a SaaS course under the agreement and the way they operate to train and augment the functionality of the SaaS, rather than building it yourself, as long as integrated with the rest of your systems. So if I look at again in the virtual agent or the chatbots of this world, there are great providers that all you have to do is share the structure in a particular manner.

Speaker 2:

To me, the key is always how differentiated it can be. Is there really incremental value that we can provide when we build it that the tool cannot? What is the landscape? Because if there's more than one service that can provide that, the price will probably over time come down or at least not get to a point where it is unaffordable for you. What is the maintenance cost you have? How integrated is with the rest of the stack. There's very few cases where I see where it makes sense to build entirely from scratch. More often than not it's about providing the data and systems, so the tool is tailored to your setup In a way. I give a comparison Look at value-based bidding in search or rebuilding the logic for the building algorithm. You just feed in the data and your business logic. That's kind of the way to summarize and how I see the industry thriving.

Speaker 1:

Yeah, I agree with that. I agree with that totally. I think it's just a matter of like. These days there's a lot like data activation is a bit of a buzzword itself by now. But I think one thing I encourage people to do is remember that, like, first party is not just audiences, first party is all kinds of data that we have.

Speaker 2:

Love it.

Speaker 1:

Yeah.

Speaker 2:

That's what we say all the time. I try to say business data as much as possible rather than first-party data these days.

Speaker 1:

Yeah, because for people it just has become just totally synonymous or exclusively with audiences and so on. But yeah, we've got all kinds of data. That's first-party to us. I saw something interesting that you shared online recently. You had a matrix or a quadrant analysis or a four box model. I love these four box models, you know, they're just like candy for the brain, but you had. So you had, you know, a horizontal axis and a vertical axis. Don't remember for sure which was which, but on one axis you've got going from augmentation to automation and then, on the other hand, you've got from native to custom, so that in one example you could end up with something like custom automation or combine that to, let's say, native augmentation. You could just come up with these different combinations. Could you walk us through that framework while you were talking about there?

Speaker 2:

Yeah, and you see, I always think of digital marketing as, at the end of the day, it comes to two types of decisions the decisions you take or the decisions you let the algorithm take. And if you think of anything we do is always either, let's say, when you look at the results of an MMM, we take the readouts and then we make a budgeting decision. That's fairly manual, then it's fed into the different platforms. But if I look at automation, it goes back to, for instance, serving in shopping certain assets rather than others because they're automatically tagged, we may have lower, greater cost of goods, availability and so on. Whatever logic we may come up with. Imagine these are the two ends from having data to take something by fairly manual action and then that automated luck, and then whether it's done natively within a platform, as constrained by the platform, or something fully built yourself. So again, these days no one would probably try to come up with a bidding algorithm outside of a Google meta or something like that. So again, it simply wouldn't really work and most likely you'd really do it native within the platform where the inventory sits, and it's why you're often giving that data to the app.

Speaker 2:

Now the challenge we have as an industry, we get excited about a solution and I know I sound like I'm pooping on the Maltech party. I love Maltech, by the way. I love it. It's my job. I guess a lot of my refrains and constraints come from seeing many projects failing because the expectations were inflated. I just want everyone to come to the start and say, hey, this project, everyone is aligned on the expectations and there are no surprises six to 12 months later when the tool is used. So that's where most of my work comes from make sure that those expectations are fair from the start and delivered in the end.

Speaker 2:

So within that, I find that many starts with the idea of that custom, fully automated art. The truth is even and MMM is a fantastic example of that so many times people just take it but don't trust the outcome or still have a lot of conversations internally on the constraints there. Maybe we don't want to really move the budget from here to there, or we have a relationship with this agency on this part of the inventory and so on. So what happens in the real world? You don't really trust in that many instances to have something fully automated and very often you don't even want it fully custom. It just doesn't make sense.

Speaker 2:

So most companies should start with maximizing what I call platform perfect, meaning customize and set up the platforms to the best that may be with the native functionalities. Make sure you've exhausted that. Imagine on a yield curve is the very start. Then usually you have two decisions. Either at this moment in time I'm moving towards a custom route because the platform didn't have some functionalities I wanted, so I tend to move up in the box, but I'm still augmenting some of the decision.

Speaker 2:

That example I started with MMM and saying I may start with the Google's MMM or the methods of this word and then customize it and really making my own setup. I take the library but I actually go and take and build my own logic afterwards. Or then the other way around, to the right of that box where I said, okay, not only have I augmented my decision, but now I'm going to fully trust an entirely automated so you could in theory have some changes. I'm going to take the MMM example again, where actually the channel mix and the badges could be even changed automatically. There's nothing to prevent you from building that it's just very often isn't practical.

Speaker 2:

And I give you the MMM example, but I could give you the same around bidding or creative personalized experience and so on. So it's around first make sure you maximize the possibility of the platform that everyone else has and then see where you want to go and build, build further automation so that you don't actually have to take any decisions because you trust the output of the platform and actually it's working very well. So it's about speed of adaptation, or maybe the challenge is there are some features you didn't want, you didn't get, and that's where you go and build them.

Speaker 1:

Yeah, I like that. I like that framework. Thanks for sharing it with us. And yeah, I like that. I like that framework. Thanks for sharing it with us. And yeah, I mean also I'm laughing at myself about this thing about, but I guess at this point, they've been big stakeholders for a while now in the whole rollout of GDPR and everything that has followed since. So when you talk about marketing teams collaborating with legal, with IT, in terms of AI, do you see this as A change in direction or just kind of an acceleration, a change in degree of what's already been occurring?

Speaker 2:

I'd say the latter, but with one important change. Over the last five or so years I worked a lot more with IT and legal than I would have expected 10 years ago. But what happened is that it never came from a shared roadmap, it was marketing's roadmap and then tried to go to IT and legal after when they had already made a decision and then had to try to align to the roadmap of ways of working, which meant that it took a long time. We have clients that for years they've been talking about the activation of audiences by customer match or trust EVP, which again is customer match. But we've been talking for years because, again, there are a lot of things that weren't in place and the challenge I see now that speed really cannot work anymore.

Speaker 2:

Now, if I look again at the industry, I see a really high penetration rate of your enterprise tech Right, so everyone has it. The reason you don't really go anymore into a client and see that they don't have a CRM with CDP-like functionalities or STP. It's very rare that you don't go and see a good analytics platform and most have DV360 or another great DSP. Again, it becomes really hard to have a differentiator on tech. Tech alone will not be the differentiator.

Speaker 2:

Where I see the differentiator is the use of the tech, the ingestion, both sides of business data, as we call it, rather than first party just being audience identified. And with that in mind, you can't really do without IT and legal. So, while before you had a pointed, time-sensitive, reactive question, now it's an integral part of your strategy and the ones that will thrive will be the one that will have something slightly different from the competition. And, as I mentioned, once you've done your platform perfect, ticked a box and say I got all the features I should have. And, as I mentioned, once you've done your platform perfect, tick the box and say I got all the features I should have. Now it's about bringing my business data, which, by definition, will be different than the competitors and will be better for our own business. So that's why, to me now the shared roadmap is key. Unfortunately, I don't see enough of that. I think I can count again on one hand the amount of times in which I've seen CMOs and CTOs having a shared roadmap for the technology.

Speaker 2:

It's really, really rare to see so it is a continuation of the journey, but it's also a very different frame of mind Again if I think of 10 years ago you remember when Google Tag Manager was launched, the tag management solution I remember the value prop.

Speaker 2:

I wish I went. Maybe I should go on a web archive. The value prop of something was you're not going to need IT. And now it's all you, you marketers you're not going to need IT, you can do all on your own. And we've also seen first how that went in terms of how, eventually, the quality in some ways that led to how bad it was as a result, but also when you start adding server-side measurement and all the other part, it just it wasn't a good value prop anymore. And that's the challenge you have now. It is we got to think less about legal IT and marketing. But what is our shared technology roadmap? What do we need to make sure that we do as a business? So yeah again, long answer for a short question, but you as a business. So yeah again, long answer for a short question, but it really has to accelerate, but it has to start from a schedule.

Speaker 1:

Yeah, I, I think that's a that's a great way of looking at. No long answer is always appreciated on this podcast, by the way. So, yeah, now we're we're we're transitioning bit by bit here from kind of this ai topic toward into the realm of like, privacy and so on. You, recently you were highlighting kind of the I don't know a nice way of calling this, the acronym vomit of privacy that's out there. You know, gdpr, itp, ccpa, scan or S-K-A-N that's one from Apple, right, I don't even know that one that well, to be honest, dma 3p cookie deprecation.

Speaker 1:

So, yeah, I mean, how, how are businesses adapting to these changes? Like it, and and is there a way? I mean, maybe you've outlined in for us a bit there when you're talking about shared road maps and stuff, but I think people just feel overwhelmed by the pace of change and we feel like we're reacting, putting out fires, coping, and sure, there are probably some larger enterprise advertisers who have been working this a while and they're more prepared than others, but I don't want to judge anyone. That's probably not the way the average business feels. Is there a way we can get from this reactive mode to a more active mode Is there?

Speaker 2:

a way we can get from this reactive mode to a more active mode. Two parts answer and I'll say first I actually think they cope surprisingly well when you add all the changes plus COVID, plus a recession that maybe wasn't really a recession, a vibe solution.

Speaker 1:

Yeah.

Speaker 2:

And then imagine a lot of the inflationary pressures, layoffs it's been really hard for marketers in the last five years. So I would actually say I think the industry has reacted pretty well all things considered, especially since they weren't really used to that way of working. Now, could it have been better? Could it have been faster? Certainly so, but I think it was a tall order on the industry as a whole. Now to your point.

Speaker 2:

The second part is how do we move from coping to actually driving? And I think we have to take a bit of a step back, because too often we follow features right. And I'm in my mid-30s I'm not sure how old you are, but probably similar age range and I think the a lot of my clients now they have pnls are in my same age range as well. But we've learned a lot from platforms. That's how we learn digital marketing, which led us to this way of working, which is always chase the feature, new feature gets announced by the platform. You try to understand it, you see, you implement it. We don't necessarily have the mindset to have a roadmap in our brains, which is, at the end of the day, I got five levers as a digital marketer, which is targeting, bidding, creative experience and measurement, and arguably the latter is to help the other four.

Speaker 2:

If you have this mental roadmap in your mind, it's always how you're doing well against them that any new features should be assessed and introduced.

Speaker 2:

But you have a roadmap around the use case and then you can evaluate a number of solutions and actually be proactive about that.

Speaker 2:

And you'll often find that around third-party cookie deprecation, we just keep talking about it because of Chrome, but there's so many environments in which actually really is. It's already been there for a long, long time and you're potentially ignoring them because you're just chasing the news on one area there. So I guess my point there is just taking a bit of a step back, constantly keeping an eye on the changes in the industry, but your mental roadmap should be around the use case and the levers, not the features, and then you can evaluate on a three six-month window every new feature that support or impact any of your abilities that you have described, because, back to the start, your problems in each of those four five areas haven't changed yeah, and I I mean I like this idea of evaluating on a, you know, quarterly or biannual basis these features, rather than, you know, freaking out on day one as soon as there's some kind of a press release or however.

Speaker 1:

I mean, I think there's often again FOMO, feeling like people wanting to be a first mover or see if there's some kind of tactical arbitrage that they can get by being by acting quick. But when you view it from that higher level of like what's actually moving the needle against these areas, I think it is kind of a more just a measured way of approaching that. So I think that's great advice.

Speaker 2:

And put numbers against them. Right, maybe because I'm a catalyst, but to me, once again, it's just put numbers. Put numbers on the impact. What is the likely impact? What is it now? When? Is it about revenue? Is it an opportunity cost yeah, that your revenue otherwise missed or is it actually impacting your efficiency, your bottom line, at this moment in time? One way or another, it's just put some numbers against each of them and I understand.

Speaker 2:

Some of them are hard to describe and see, I remember we work a lot with Google Analytics. So around the deployment of Google Analytics 4, I understand often it's hard to make a case for speeding up one feature or another, because an analytics platform is a cost center by definition rather than a value-driving system, unless you take actions that result from that platform in itself. So I understand it's quite hard, but it doesn't have to be perfect from that platform in itself. So I understand it's quite hard, but it doesn't have to be perfect, and you always have a benchmark of the current status quo. So the challenge we often have is try to look at that in isolation rather than the current state, and then it becomes a bit easier because you know what you're losing or what you're adding to. It is actually quite straightforward when you look at it that way.

Speaker 1:

You threw in there as kind of like a little side comment earlier that cookies have been dead in a lot of environments for a while Now. Last week, at time of recording, google announced that they'll be delaying third-party cookie deprecation, again on the basis of feedback from these competition authorities. The concern is can they implement the privacy sandbox and can they implement these workarounds in a way that doesn't kind of self-privilege themselves? I mean, I'm fascinated by this kind of antagonism between privacy and competition. It's a really interesting dynamic, because the things that you do to support privacy are also the things you would do to build your garden walls a little higher. So it's an interesting little dilemma there. But I mean, yeah, what's going on here from your perspective? How anticipated was this and how, like you know, explain it to a CMO, or explain it to even a CEO who's not into the weeds as much?

Speaker 2:

I guess two sides One. I'm going to raise my hand and say I did not expect a delay anymore.

Speaker 2:

But, it was a bit smelly in the air in the last month or two because, as you know, we are already in Q2. And when you talk about Q3, q4, and you don't have really clear deadlines, we all know and you've been a product manager so you know that when you get closer to the date and it's still a generic date it smells of delay. So that that one, I would say probably the last 30 to 45 days I started having some doubts, but before I was, I did not think that anymore, just because I'm mindful that google has a lot of stake in terms of credibility as well when they set their blinds. If every time they have to push them back, there is a thin line to play between giving the industry more time and then people just waiting till the last minute. So it's always hard. So here I really empathize with Google and I understand that part of that was outside of the control. So again, I actually do see it as a good news because, at the end of the day, the control. So again, I actually do see as a good news because, at the end of the day, the message has been ingrained in everyone's brain for a long time. I don't actually think anyone would pause or revert anymore at this stage, because whatever they had in motion was already going to happen. So I don't actually see that having any meaningful negative impact on what the industry is doing and I think I welcome it and give everyone a bit more time.

Speaker 2:

Now I haven't actually seen the C-suite caring that much anymore about it, especially the delay, because I think we've been ingraining it for so long that it's become pretty much something that you've seen in every QBR deck for probably two, three years. It's been boring. The challenge, I see, is that one thing is saying they would want the business to adopt these new mechanisms and tools. One thing is really being ingrained in the business. So I think that that's really the challenge.

Speaker 2:

I think we will continue for probably another six to nine months, and effectively only when everything is switched off, because, once again I go back to the point, there's never really a question about what I would say of Chrome, but that challenge has been there for a long, long time on Safari and if you're looking for some business, chrome is not even their main traffic driver. So I'll tell you there is a challenge of the topic becoming a distraction and we should go back to how solid is our measurement stack? So I actively try to make my conversation being again about how good is your target in building creative experience and measurement, rather than, hey, have you implemented what an alternative to third-party cookies? How much of your targeting is third-party cookie-based? Or, if I look back in scan, did you do the new feature and such? So I try as much as possible to be from that angle, even though at the end of the day, you may end up doing the very same feature or solution as a result of that.

Speaker 1:

Yeah, that's interesting. I mean I feel that I have empathy here for Google as well, and I was also. I'll raise my hand and say I was a bit surprised. But also, just when we roll out, I can just think of things that we've rolled out. And this rollout gradient, so to say, didn't really make sense. It was like, you know, 1% for the first six months of the year and then accelerating to 100% in the second half of the year and culminating in Q4. It's all like just as a gradient. That's not a gradient, that's like smashing into a brick wall, and that part of the plan never quite made sense to me, but still, it's a tricky situation.

Speaker 1:

Yeah, and also, as you mentioned, imagine it depends on your demographics. You could have, you know it, it depends on your demographics. It could be that mobile is just incredibly important to you and that your demographic is mostly owning iPhones or, whatever the case might be, using Safari by default and every other kind of protection that's on an iPhone, and this has been a longer topic for a while. So I think it's worth calling out that this is just part of the internet. It is not the internet, so to say.

Speaker 2:

And I think one thing to add probably is I would talk a bit less about the features you're losing and I think this is maybe my message to the industry but talk a bit more about the feature you have now and the new ones.

Speaker 2:

I feel that every time, the conversation is about what we're losing rather than what we're getting. So people struggle to be two messages. Give them one message. For instance, the age-old question is how do I reach incremental audiences and how do I make sure I'm not exposing someone that I've already exposed 17 times before to the same impression? Let's talk about the features that are being deployed in the platforms to support that, and I would say we should talk so much less about the deprecation of third-party cookies. They're interrelated, they're just two sides of the very same metal, but people are incredibly struggling in making the connection between the two. And I and I go to again how good is your targeting? And then you start talking about the different decision. Are you able to discern between audiences you've reached before or not? Are you able to reach in-market or out-of-market? How are you doing that? Again, those should be the questions, and then the features against that, and I think it's taking us too long as an industry to talk about that.

Speaker 1:

Yeah, I think well actually. So there was a article. Do you know Eric Seufert, if I'm saying his name right, I don't know how he pronounces it, I don't know he's got some really good articles and he was describing the broken brain of marketers. And I think that we do have kind of a broken brain from these short-term measurement loops that we've been in and because it's actually you could think of it as like a dopamine loop.

Speaker 1:

If we're making a metaphor here or like this, it's sort of a cycle of addiction that we've gotten into for years. We've gotten a little addicted to this user level measurements and the question is how far did it really bring us compared to if we switch into an econometric model? It'll be very interesting to see how companies kind of detox themselves from this and get used to different ways of thinking about these. And I think the way you're describing it is a very mature way of describing it just looking at these fundamental areas and checking your futures against that and focusing not just on what we lose but on what we gain. I like all of that a lot.

Speaker 2:

I think it's the key for me is there's so much going on and new stuff every day, but as we go back from the very start, the things we want to do are still the same. There's one interesting point on I I. What I love of marketing these days is there's so much more research in what works and doesn't work that I'm pretty sure not only I didn't see three years ago. Maybe you know you may be in your bubble and not see, not look for it, but I'm actually generally a believer that the last three to five years are generally the golden age of marketing effectiveness and you see so many studies and we find that often we over index in areas we think we have control of advertiser, like targeting, and we potentially forgot about everything else that matters within the marketing spectrum. And that's why I go back to that illusion of control made everyone so comfortable.

Speaker 2:

I think it was an illusion. We just have to realize that it never was. Never will I actually hate the expression. Right message, right person, right time. I just can't stand it. It never was.

Speaker 2:

Even with all the information you could have, it never was perfect. It never will. So to me, is the point around really thinking, looking at the research, which is a bit of a relief. I'm a new parent and you know when you see those standards I say you know what? At the end of the day, whatever you do, you can't screw up or make them the best in the world. At the end of the day, there's so many other factors outside of your control. So take you with a bit of a relief.

Speaker 1:

I think there's like this weather system around us and we are, you know, focused on sometimes tiny little things in that weather system that's the best analogy I may use again about the weather system.

Speaker 2:

yes, that's a great analogy to say how a litter we have control of, and I think that that's the key to do more.

Speaker 1:

Well, I mean, we're sadly on our time box, daniel, so I'm going to I'll have to let you go, but any final words, or where can we find you online too?

Speaker 2:

I'm pretty active on LinkedIn. I don't know if I have words. Seems a bit can I say cocky from my side? So yeah, no real final words. It seems a bit can I say cocky on the phone from my side? So yeah, no real final words. It's been a generally good conversation On LinkedIn. I try to post my ramblings so if anyone is interested they can find me there.

Speaker 1:

No, they're not ramblings. They're very well-structured, interesting ideas. I love reading your stuff. That's why I wanted to bring you on the show. So thanks for everything that you do and thanks for joining the show. Thank you, mike, a lovely day. Thanks for listening to Growing E-Commerce and if you enjoyed this podcast, please share it with your friends. This is a production of Smarter E-Commerce, also known as SMEC. To learn more about us, visit at smarter-ecommercecom.