00:00.031 –> 00:10.628
[SPEAKER_02]: In 2025, with all stuff happening with the AI, with the latest legislations, but in general, the speed at which companies move nowadays is nowhere close to what it used to be.

00:11.128 –> 00:15.315
[SPEAKER_02]: So it means that some of the processes change faster that we can change the reporting for that.

00:15.876 –> 00:26.933
[SPEAKER_02]: That means somebody might have, and usually a lot of cases, myself, have to be in a meeting that is not related to me yet, but in two months from now, I was going to break something and I need to be aware of that way in advance.

00:29.545 –> 00:32.673
[SPEAKER_01]: It’s the e-commerce master plan podcast.

00:32.693 –> 00:37.185
[SPEAKER_01]: It has to help you solve your marketing problems and grow your e-commerce business.

00:37.205 –> 00:45.327
[SPEAKER_01]: Cutting through the hinder to bring you inspiration and advice from the e-commerce sector and beyond, here’s your host, Chloe Thomas.

00:48.293 –> 00:50.476
[SPEAKER_00]: Hello, and welcome, it’s great to have you here.

00:50.516 –> 00:54.441
[SPEAKER_00]: Thank you for hitting play and choosing to listen to one of our inspiring guests.

00:55.042 –> 01:00.288
[SPEAKER_00]: Now I love to give a little shout out to the listeners who are kind enough to take the time to post a review of the show.

01:00.308 –> 01:02.631
[SPEAKER_00]: So we’ve got one of those to do in this episode.

01:03.312 –> 01:12.464
[SPEAKER_00]: Huge, thank you to Babs MT from the US for your five star rating on Apple podcast.

01:12.484 –> 01:13.525
[SPEAKER_00]: Thank you very much.

01:13.505 –> 01:22.998
[SPEAKER_00]: Babs empty rights, great host, blushing, a great podcast, and says, you know what makes for a great podcast host, intellectual curiosity.

01:23.098 –> 01:34.154
[SPEAKER_00]: Chloe has it in spades and shows up in the depth and breadth of topics covered and guests invited, where the inspirational or practical and inspirational e-commerce master plan always offers a lot of knowledge.

01:34.554 –> 01:35.455
[SPEAKER_00]: Definitely listen.

01:36.016 –> 01:36.797
[SPEAKER_00]: Thank you, Babs.

01:36.978 –> 01:40.963
[SPEAKER_00]: Blushing slightly over here, but thank you very much for that lovely review.

01:40.943 –> 01:49.971
[SPEAKER_00]: If you would like one of these shout outs at the start of an episode, go to ECMP.info for such a review to find out how to leave one.

01:50.752 –> 01:51.012
[SPEAKER_00]: Right.

01:51.092 –> 01:52.554
[SPEAKER_00]: What are we got coming up in this episode?

01:52.634 –> 02:01.922
[SPEAKER_00]: Well, we are not going to be talking very much about the e-commerce store that’s on, which is cat car wow.

02:02.323 –> 02:10.410
[SPEAKER_00]: Instead, we’re going to be talking about brand side analytics and tracking.

02:10.390 –> 02:15.376
[SPEAKER_00]: in the world of tracking analytics, cookies, data, all that kind of good stuff.

02:15.897 –> 02:38.625
[SPEAKER_00]: And he’s going to be talking about how we, how he at CarWow manages his team, how they make the decisions on what to do, how they work with the rest of the company, some cool tool tips, AI tips, lots of practical advice, lots of sane advice to help you make the right decisions with how deep you go on all these things, because I know you’re all wrestling with it at the moment.

02:38.605 –> 02:47.081
[SPEAKER_00]: So if data analytics and tracking has you all buffuzzled, prepare to get some great tips, tools, and advice to help you manage it all without pulling all your hair out.

02:47.863 –> 02:54.195
[SPEAKER_00]: Huge thanks as well to Raven for getting in touch to offer to talk about this crucial topic because he’s ace as you’re about to find out.

02:55.077 –> 02:59.365
[SPEAKER_00]: And make sure you listen right to the end because um, his top tips are super hot too.

03:04.998 –> 03:07.001
[SPEAKER_00]: and now to introduce our special guest.

03:07.461 –> 03:16.494
[SPEAKER_00]: Roman Petra-Chankov is the lead marketing analyst at CarWow, so yes, we’re going to be getting deep into tracking attribution and all that GCGC stuff.

03:17.075 –> 03:28.170
[SPEAKER_00]: CarWow has grown from a review site to be a one-stop shop for changing your car, offering customers the ability to buy a new or used car or to sell their existing vehicle.

03:28.150 –> 03:32.516
[SPEAKER_00]: Founded in 2013, they now operate in the UK, Germany and Spain.

03:33.017 –> 03:45.133
[SPEAKER_00]: In 2024, they grew 50% with nearly £3 billion worth of car’s bought on the site, and £1.8 billion of cars listed for sale through the Selmei car service.

03:45.474 –> 03:46.395
[SPEAKER_00]: Hello Roman!

03:46.949 –> 03:48.894
[SPEAKER_02]: Hello, thank you very much for having me here.

03:49.155 –> 03:52.243
[SPEAKER_02]: So nice to be not just a listener, but also a participated.

03:52.363 –> 03:52.945
[SPEAKER_00]: Brilliant.

03:53.366 –> 03:57.035
[SPEAKER_00]: Brilliant to have you here and to talk about such a juicy topic.

03:57.095 –> 04:01.988
[SPEAKER_00]: First off, how did you end up in this world of e-commerce and attribution and all these good things?

04:02.542 –> 04:05.125
[SPEAKER_02]: Like a lot of people by accident, I would probably say.

04:05.145 –> 04:10.932
[SPEAKER_02]: I have been a computer science student and I just have ended up in the marketing agency.

04:10.972 –> 04:25.630
[SPEAKER_02]: I spent about five years on my early Korean agencies doing everything from PTC, Facebook ads, a little bit of email, serial optimization, talking to clients, and then a very, very wild time as a lot of agencies are.

04:25.610 –> 04:37.865
[SPEAKER_02]: And then after about five years on the agency side, I decided that I want to move to the client side, where you have a little bit more time to figure things out, and to build a bit more long-term sustainable solutions.

04:38.566 –> 04:47.497
[SPEAKER_02]: And essentially, throughout all this time, I have been doing most of the performance in e-commerce and lead gen, which are very, very close and not good a lot.

04:47.517 –> 04:53.644
[SPEAKER_02]: A huge part of it has been Google Ads, or Google related solutions, like analytics, tag manager, you name it.

04:53.744 –> 04:55.026
[SPEAKER_02]: I have tried most of them.

04:55.310 –> 05:14.905
[SPEAKER_00]: Excellent and have you found the now you’re specialising in the tracking attribution data side of it all have you found that early grounding in understanding how all the other channels work has made things easier for you because you get the affiliate principles the Google ads principles the meta ads principles and so on and so forth.

05:15.340 –> 05:24.898
[SPEAKER_02]: I’m in that unique position because at the same time I understand how back-end works, tech works, tracking works, how JavaScript is actually getting executed.

05:25.299 –> 05:30.970
[SPEAKER_02]: But I can go to marketing team and talk marketing language and then go and go to finance team and talk finance language.

05:30.950 –> 05:38.423
[SPEAKER_02]: It makes life so much easier to understand, but it does it mean to do something, but it also was going to show in the end of it.

05:39.004 –> 05:44.053
[SPEAKER_02]: Besides that, when you go into all the new tech that comes up, it’s Fisherman comes to Kuki.

05:44.073 –> 05:46.157
[SPEAKER_02]: I’ll favor it consent changes.

05:46.137 –> 05:51.843
[SPEAKER_02]: It helps a lot to understand how it impacts the exact setup and the data flow.

05:52.403 –> 05:55.206
[SPEAKER_02]: So yeah, it definitely makes my life so much easier.

05:55.266 –> 06:02.754
[SPEAKER_02]: But at the same time, it’s just as far as I say, that I haven’t done my own campaign in Facebook for like many, many years.

06:03.114 –> 06:04.936
[SPEAKER_02]: And the tech is changing.

06:05.016 –> 06:08.179
[SPEAKER_02]: But I always feel like, oh, it used to be like that.

06:08.640 –> 06:09.621
[SPEAKER_02]: It’s no longer like that.

06:09.781 –> 06:13.264
[SPEAKER_02]: And I have to remind myself that there are people that know it’s so much better.

06:13.445 –> 06:14.946
[SPEAKER_02]: And I need to learn from them as well.

06:15.365 –> 06:21.954
[SPEAKER_00]: I know what you mean, having run an ad agency for 10 years, I would not trust myself in a Google Ads account or a Facebook Ads account anymore.

06:22.195 –> 06:24.077
[SPEAKER_00]: I know enough to be really dangerous.

06:25.099 –> 06:26.581
[SPEAKER_00]: So don’t let me near it, anybody.

06:26.661 –> 06:27.843
[SPEAKER_00]: Don’t do it.

06:27.863 –> 06:39.880
[SPEAKER_00]: So clearly, Rayman, you’re working for a very large company with a huge amount of data across different sites, different selling methods, and huge volumes of traffic coming through as well.

06:39.980 –> 06:44.747
[SPEAKER_00]: And then I know because you’ve already told me

06:44.727 –> 06:47.470
[SPEAKER_00]: So how do you go about managing it?

06:48.191 –> 06:49.312
[SPEAKER_00]: You know what I mean?

06:49.333 –> 06:57.162
[SPEAKER_00]: I suspect I’ll audience a more familiar with managing an ad campaign, a managing an email campaign, where you know what you’ve got to deliver, you know what you’ve got to do.

06:57.202 –> 07:05.672
[SPEAKER_00]: So how on the analytics and the data side of it, do you work out where to focus you in your team’s effort and how to deliver for the business?

07:06.445 –> 07:31.322
[SPEAKER_02]: It’s a very good question and you know, inside we have a phrase saying that if the prioritization doesn’t hurt, you know, prioritizing well and I think that describes really well how we just like what we try to do the thing that I try to do a lot is different people in my team have a different background and it’s very important to have enough time to fix the technical back backlog and things that you have to be prepared for.

07:31.302 –> 07:43.945
[SPEAKER_02]: Like a good example of that, we’re all thinking that G clearly is going to be disappearing at some time, so you have to put it all a research upfront and you find the person who can do all that work before it all happens.

07:44.466 –> 07:51.859
[SPEAKER_02]: But that person isn’t not very likely to be the person who’s going to be very good at the BI and talking to millions stakeholders.

07:51.839 –> 08:14.067
[SPEAKER_02]: So, it’s never easy, finding the right person, finding the right team, I try to be always in the process, understanding how things are done, and some things are doing myself, but because my background allows me to, for example, it just so happens that I mostly do all GTM changes myself, instead of delegating to somebody, because GTM is such a rare skill when you talk about analytics.

08:14.047 –> 08:24.921
[SPEAKER_02]: Usually a lot of companies big enough will have like a tracking person for example We decided to go a little bit different route because we don’t do a lot of changes, but That’s essentially how we divide work internally.

08:24.941 –> 08:25.382
[SPEAKER_00]: Got you.

08:25.402 –> 08:31.009
[SPEAKER_00]: And for everyone listing GTM Google Tag Manager just in case you’re wondering And BI, I’m guessing is business intelligence?

08:31.549 –> 08:33.412
[SPEAKER_02]: A dashboarding, you can think about it this way.

08:33.592 –> 08:36.155
[SPEAKER_00]: Yeah, it’s one of those utterly meaningless phrases, isn’t it?

08:37.357 –> 08:38.558
[SPEAKER_00]: It’s like business intelligence.

08:38.639 –> 08:39.059
[SPEAKER_00]: What?

08:39.319 –> 08:39.880
[SPEAKER_00]: Dashboards?

08:40.020 –> 08:40.901
[SPEAKER_00]: Dashboards people?

08:41.222 –> 08:42.163
[SPEAKER_02]: Dashboards here.

08:42.548 –> 08:53.241
[SPEAKER_00]: So your team, you know, the lumping together of analytics and data science, data engineering, Google Tag Manager or tracking, all this stuff, stuff together.

08:53.261 –> 08:56.766
[SPEAKER_00]: How do you, where do you spend most of your time, actually?

08:56.946 –> 09:00.570
[SPEAKER_00]: Is it on the maintenance and they’re getting ready for new legislation?

09:00.590 –> 09:07.459
[SPEAKER_00]: Is it keeping the sea sweet, happy with their dashboards or is it something else that that occupies you most of the time day to day?

09:08.198 –> 09:29.295
[SPEAKER_02]: I think a huge part of it is education internally in the best sense of that, talking to people and helping them to understand the data is a huge part of my job, because as you said, like we have a lot of different data points coming in.

09:29.275 –> 09:33.523
[SPEAKER_02]: And it’s a quiet, quick journey from the user perspective, especially if you know what you want.

09:33.984 –> 09:41.878
[SPEAKER_02]: If you want to sell your car, you need to take pictures of that, you need to take the service history, you need to understand the condition and the scratches.

09:42.319 –> 09:47.368
[SPEAKER_02]: The journey for those two users is a very different and the time it takes for those users is also different.

09:47.348 –> 10:09.154
[SPEAKER_02]: So if you try to put those numbers together and say, well, there’s a person buying car lead and there’s a person selling a car lead, those are not same intent leads and you need to go through different, but in the GC it’s going to look like two columns just one by one and you need to go through different teams and departments to explain, hey, we work with them differently, we need to invest them differently.

10:09.134 –> 10:12.478
[SPEAKER_02]: A huge part of it is understanding also how the business changes.

10:13.078 –> 10:24.190
[SPEAKER_02]: I think in 2025 with all stuff happening with the I, with the latest legislations, but in general the speed at which companies move nowadays is nowhere close to what it used to be.

10:24.691 –> 10:28.896
[SPEAKER_02]: So it means that some of the processes change faster that we can change the reporting for that.

10:29.436 –> 10:37.605
[SPEAKER_02]: That means somebody might even, usually a lot of cases myself have to be an emitting that is not related to me yet, but in two months from now it’s going to break something.

10:37.585 –> 10:40.848
[SPEAKER_02]: And I need to be aware of that way when advanced.

10:41.509 –> 10:48.737
[SPEAKER_02]: But I try to, like about 50% of my time is probably going to actually doing delivering, like coding or reviewing something.

10:49.537 –> 10:52.380
[SPEAKER_02]: And the team composition also constantly changing.

10:52.520 –> 11:04.673
[SPEAKER_02]: For example, last year, we have been having a date engineers a part of the team, because we were basically Google vehicle ads, which is the PMAX for vehicle ads or a feed-based,

11:04.653 –> 11:08.018
[SPEAKER_02]: advertising that’s specifically for cars and the UK has been launched.

11:08.598 –> 11:11.602
[SPEAKER_02]: A lot of other e-commerce advertisers will have been very familiar with the product.

11:11.663 –> 11:16.229
[SPEAKER_02]: It’s a Google Merchant Center, it’s just for like over the decade, hasn’t been available for cars.

11:16.950 –> 11:23.679
[SPEAKER_02]: And we had a special person temporarily in the team, just to make sure we have the feed right, we sent it in the right way.

11:23.999 –> 11:28.065
[SPEAKER_02]: And that, for example, completely changed my workflow because I have to be in it.

11:28.626 –> 11:34.153
[SPEAKER_02]: Because the person is going to, at some point, move forward and somebody internal

11:34.133 –> 11:36.197
[SPEAKER_02]: and basically everything around it.

11:36.858 –> 11:39.222
[SPEAKER_00]: There must be huge opportunity as well.

11:39.463 –> 11:44.912
[SPEAKER_00]: You don’t suddenly being able to use P-Max and Google shopping.

11:44.972 –> 11:46.876
[SPEAKER_00]: You don’t want to drop the ball on that one, do you?

11:46.896 –> 11:49.380
[SPEAKER_00]: So it’s like huge focus for the year.

11:49.420 –> 11:53.728
[SPEAKER_00]: And you’ve mentioned quite a few of the Google tools.

11:53.708 –> 11:56.094
[SPEAKER_00]: Are you, to manage all of this?

11:56.354 –> 12:02.830
[SPEAKER_00]: Are you focused on good deployment of GTM, good deployment of GA4, and the Google suite?

12:02.870 –> 12:08.865
[SPEAKER_00]: Are you using additional analytics and tracking tools to manage all that side of things?

12:09.183 –> 12:11.727
[SPEAKER_02]: Actually, internally, we don’t use Google Analytics much.

12:12.008 –> 12:15.073
[SPEAKER_02]: We use Amplitude to track most of the events.

12:15.193 –> 12:17.457
[SPEAKER_02]: It just so happened that we use Amplitude.

12:17.477 –> 12:20.582
[SPEAKER_02]: They have benefits, ups and downs for both of those products.

12:20.602 –> 12:28.035
[SPEAKER_02]: I think Amplitude is really good when it comes to understanding how the product works and how product journeys is essentially have been

12:28.015 –> 12:31.541
[SPEAKER_02]: product, focus tool that has evolved into supporting a lot of marketing.

12:31.922 –> 12:36.309
[SPEAKER_02]: But what it means for us is most of our analytics is not happening in unspecific tool.

12:36.349 –> 12:45.805
[SPEAKER_02]: We put it in snowflake, which is a database system, and we do a lot of heavy database calculation, sequeling around to get data surfaced.

12:45.785 –> 12:57.297
[SPEAKER_02]: Again, the good example of that, when you sell your car, the time it takes to sell the car and for a car be actually collected and data to actually come to us is more than one or two days obviously.

12:57.878 –> 13:03.764
[SPEAKER_02]: If you want to sell your car tomorrow, you probably want to wait for a new car to be delivered to you before the car is picked up.

13:04.305 –> 13:15.657
[SPEAKER_02]: That creates delays and Google Analytics only allows for example for seven days of delay if it’s past seven days you can’t upload anything.

13:15.637 –> 13:18.222
[SPEAKER_02]: not Google tools for a lot of different cases.

13:18.438 –> 13:26.188
[SPEAKER_00]: So for you, because of the size and the complexity of the business, it’s, let’s gather the raw data in your snowflake database.

13:26.729 –> 13:35.881
[SPEAKER_00]: So then you can pull it into additional data manipulation platforms, as you need to read it and use it for different services.

13:35.981 –> 13:39.185
[SPEAKER_00]: But you’re not going to trust anything else with the raw data.

13:39.205 –> 13:48.377
[SPEAKER_00]: You won’t control of that raw data, centrally, so then whatever the business needs, you have the ability to go and then analyze it.

13:48.610 –> 13:49.331
[SPEAKER_02]: Yeah, I’m sent.

13:49.592 –> 13:59.368
[SPEAKER_02]: You can even simplify it sometimes, I say, they’re always going to be tools that do 99% of the things, but they’re not going to do exactly the thing that you need the most for the cases like this.

14:00.129 –> 14:06.079
[SPEAKER_00]: So I would have seen a lot of our audience don’t need that level of complexity.

14:06.953 –> 14:10.578
[SPEAKER_00]: And actually, let’s expand this out before I ask you this question.

14:12.180 –> 14:26.700
[SPEAKER_00]: There is a huge amount that all e-commerce businesses are trying to get to grips with at the moment, around tracking integration, making your ads work better, complying with new legal stuff coming down the lines about cookies and God knows whatever else.

14:26.680 –> 14:30.887
[SPEAKER_00]: And there’s a lot of new tech around that promises the world.

14:31.708 –> 14:37.877
[SPEAKER_00]: And how would you advise them to go about working out what they need or what they need to care about?

14:37.937 –> 14:44.568
[SPEAKER_00]: Because it will be very easy to kind of get lost in the attribution tracking data box at the moment.

14:44.688 –> 14:49.996
[SPEAKER_00]: Even if you’re running a tiny e-commerce business and get completely overwhelmed with it, also how do you filter all of that?

14:50.036 –> 14:52.099
[SPEAKER_00]: How do you make sure you’re making the right decisions?

14:52.480 –> 14:53.802
[SPEAKER_00]: Bit massive questions, sorry.

14:54.052 –> 15:05.846
[SPEAKER_02]: It’s a massive question and I’m going to start very very boring and I’m going to say regardless of where you’re in the journey, you need to go and read what’s the actual requirements for having a consent banner and quickest what you can and cannot do.

15:05.866 –> 15:07.428
[SPEAKER_02]: It’s very very important.

15:07.528 –> 15:13.115
[SPEAKER_02]: It’s different across the UK and you and I have been on this journey myself for the course of the last two years.

15:13.215 –> 15:15.898
[SPEAKER_02]: I’ve learned a lot of things that I didn’t expect.

15:16.279 –> 15:20.464
[SPEAKER_02]: How GDPR is completely unrelated, for example, how the quickest work in the UK.

15:20.444 –> 15:24.576
[SPEAKER_02]: and a lot of people need to definitely put some time and effort into understanding it.

15:24.796 –> 15:25.719
[SPEAKER_02]: You don’t need to go deep.

15:25.959 –> 15:35.466
[SPEAKER_02]: We have an amazing Gemini, Chudgipiti, all of this amazing tools will help you to learn it, but it’s going to be very, very educated for most of the people.

15:35.446 –> 15:39.734
[SPEAKER_02]: I think where it comes to picking the right choice, it comes to the size of the business.

15:40.175 –> 15:55.923
[SPEAKER_02]: If you small e-commerce shop that mostly drives around the BBC ads, lead gen ads, or it comes to a few people in the team in the company, and you have a relative, a small budget, you’re not going to get any benefits from doing a advanced attribution.

15:56.023 –> 15:56.705
[SPEAKER_02]: It’s just not good.

15:56.725 –> 15:59.209
[SPEAKER_02]: You don’t have enough data to do that.

15:59.189 –> 16:23.779
[SPEAKER_02]: When you have brand marketing, and you have brand, as an allocated spend in your budget, and you start thinking about a lot of people having a returning transaction, when you come to the things like LTV, which life got lifetime customer value, and the cost of acquisition, our favorite kind of LTV, that’s where you start asking question, what is a new customer, and what is the returning customer, and who I shouldn’t

16:23.759 –> 16:24.760
[SPEAKER_02]: actually activate.

16:25.101 –> 16:30.448
[SPEAKER_02]: And this is where it becomes way more profitable to invest in the tooling and understanding it.

16:30.488 –> 16:34.032
[SPEAKER_02]: Exact tooling will probably come down to exact e-commerce shop.

16:34.072 –> 16:46.588
[SPEAKER_02]: It’s very hard for me to say Google Analytics for example provides a ton of insights, but in a lot of cases like this, if your tracking is just not quite right, it’s going to come with amazing but wrong insights.

16:47.029 –> 16:49.412
[SPEAKER_02]: And that’s probably a very dangerous

16:49.780 –> 16:55.229
[SPEAKER_00]: So number one, make sure we’ve got our consent banners working properly for the audiences.

16:55.610 –> 17:05.406
[SPEAKER_00]: We’re targeting which basically go and speak to Gemini or chat GPT to get started and then make sure you’ve got the right cookie banner set up correctly.

17:05.586 –> 17:09.493
[SPEAKER_00]: Is it fairly straightforward but we do have to get into it with that be fair to say?

17:09.980 –> 17:11.622
[SPEAKER_02]: It’s very first, very straightforward.

17:11.642 –> 17:13.245
[SPEAKER_02]: If you’re a big brand, let’s let’s put it there.

17:13.265 –> 17:14.907
[SPEAKER_02]: I’ll take heavily good advice.

17:15.448 –> 17:20.075
[SPEAKER_02]: If you’re a small brand, there are plenty of solutions out of the box that will help you to do it right.

17:20.676 –> 17:20.776
[SPEAKER_00]: Cool.

17:20.796 –> 17:30.250
[SPEAKER_00]: And then second thing we need to make sure we’re getting right is that we’re feeding the right conversion data and tracking back into our app platforms.

17:30.938 –> 17:32.823
[SPEAKER_02]: and a huge board.

17:33.044 –> 17:45.741
[SPEAKER_00]: And would you recommend getting third party software into do that or is that something you start of just by using what you’ve already got access to with the ad platforms technology or do we need to find something to bolt on to make that easier?

17:46.430 –> 17:52.980
[SPEAKER_02]: I’m a person who thinks if you have a chance to solve the problem straight forward, you should take that chance and try that.

17:53.781 –> 17:55.403
[SPEAKER_02]: But I’m a lazy person.

17:55.523 –> 17:58.828
[SPEAKER_02]: I really love lazy solutions that don’t break over time.

17:58.868 –> 18:03.595
[SPEAKER_02]: But the lot of companies cannot solve the problem by that.

18:03.575 –> 18:12.726
[SPEAKER_02]: Like one of the previous guests, for example, on the podcast show that they have 10, 15, 20 different platforms when they advertise and not all of them integratable together.

18:12.766 –> 18:14.669
[SPEAKER_02]: It’s very hard then goal.

18:14.709 –> 18:17.753
[SPEAKER_02]: They say, well, lyrics and only have Google ads cost and being cost are there.

18:18.233 –> 18:20.416
[SPEAKER_02]: That is just not going to solve the problem.

18:20.456 –> 18:23.280
[SPEAKER_02]: This is where you need to look into a better solution.

18:23.340 –> 18:26.984
[SPEAKER_02]: It’s going to solve at least cost allocation, basic attribution for you.

18:27.525 –> 18:33.312
[SPEAKER_00]: I think it’s what’s coming out of this for me is that it’s all about

18:33.612 –> 18:36.137
[SPEAKER_00]: What are your objectives, what’s the cost, what’s the risks?

18:36.899 –> 18:46.137
[SPEAKER_00]: You know, so what are you actually trying to improve and how much could that cost you, how much potential benefit is there and then find the solution that’s cost effective to get you there?

18:46.758 –> 18:51.344
[SPEAKER_02]: Every platform we get takes some time to manage that platform internally.

18:51.604 –> 18:52.986
[SPEAKER_02]: It’s the same with SQLs, right?

18:53.066 –> 18:59.274
[SPEAKER_02]: If you have 20, 30, 50 SQLs, you probably can manage them in G-shades in Excel, whatever is a preference, right?

18:59.875 –> 19:10.929
[SPEAKER_02]: If you have 500, 600, 5000, you will need to have a solution that will help you to automatically manage them, track them, and the same comes to attribution.

19:11.389 –> 19:16.596
[SPEAKER_02]: I think it’s really, in the context of V-commerce, I think,

19:16.576 –> 19:24.168
[SPEAKER_02]: A lot of sales coming through Amazon, eBay and a lot of companies that doesn’t actually, don’t they don’t really give you much insights?

19:24.228 –> 19:25.350
[SPEAKER_02]: What’s happening inside of it?

19:26.011 –> 19:29.676
[SPEAKER_02]: And then you will have to adjust for significantly more complex solutions.

19:30.237 –> 19:43.157
[SPEAKER_02]: But if you are in that lucky situation when you have quite a lot of traffic coming to your website and you can track that traffic, having at least some basic attribution that you understand that is not last click is going to be really, really good.

19:43.137 –> 19:47.523
[SPEAKER_02]: For example, you can have a beta-driven attribution to Google Analytics provides.

19:47.583 –> 19:50.107
[SPEAKER_02]: So Google Ads provides, you just need to understand what they do.

19:50.147 –> 19:56.495
[SPEAKER_02]: For example, a good example there is that Google Ads attribution does not equal to analytics attribution.

19:56.515 –> 19:57.276
[SPEAKER_02]: They do different things.

19:57.957 –> 20:04.526
[SPEAKER_02]: A lot of people get lost in there, and a lot of things and details there that are very, very important to understand.

20:04.566 –> 20:10.815
[SPEAKER_02]: At the same time, it doesn’t mean that one of them is bad, they do different things.

20:11.251 –> 20:25.311
[SPEAKER_00]: Do you think it’s ever possible to get to a perfect attribution, or is it about understanding what each model is telling you, and either the business or you individually deciding which one you’re most comfortable with?

20:25.882 –> 20:29.367
[SPEAKER_02]: I wish there was a way to have a perfect attribution.

20:30.128 –> 20:33.614
[SPEAKER_02]: It doesn’t exist and it hurts me to say and acknowledge that.

20:34.375 –> 20:46.253
[SPEAKER_02]: We have this, which is a very common way of looking at the triangle, which is not really it’s applicable to every business, but it’s more applicable to the businesses that have way more things that they can measure directly.

20:46.233 –> 20:55.807
[SPEAKER_02]: which is an attribution in your day-to-day foundation or thing that feeds into Google, feeds into Facebook and everything else and tell you roughly how much you need to spend.

20:56.348 –> 21:05.561
[SPEAKER_02]: Then you have incremental testing, it’s essentially when you try to measure a specific channel and understand how the channel incrementally impact your revenue.

21:05.621 –> 21:12.932
[SPEAKER_02]: So for example, you take one region that is enabled, one region is disabled and they’re being

21:12.912 –> 21:16.617
[SPEAKER_02]: If you haven’t tried that, the Facebook provides a solution for that.

21:17.137 –> 21:18.139
[SPEAKER_02]: Things called Bred Lift.

21:18.840 –> 21:22.564
[SPEAKER_02]: Google lets provide the solution for that YouTube as well and etc etc etc.

21:23.145 –> 21:26.409
[SPEAKER_02]: And that allows you to have a general understanding how different channels do.

21:26.449 –> 21:32.977
[SPEAKER_02]: And then you have a huge mountain peak, which is a marketing mix modeling.

21:33.398 –> 21:35.781
[SPEAKER_02]: When you plug in all of this incremental statistics,

21:35.761 –> 21:43.889
[SPEAKER_02]: and it goes into last 3, 4, 5 years of your spend and tries to understand how different channels impact your spend.

21:43.929 –> 21:45.711
[SPEAKER_02]: And it’s a basically a correlation model.

21:46.252 –> 21:49.916
[SPEAKER_02]: It’s not very good to understand how the campaign has been doing or keyword.

21:49.956 –> 21:51.197
[SPEAKER_02]: It just cannot do that.

21:51.617 –> 21:59.946
[SPEAKER_02]: But it can give your calibration on saying like Facebook actually is a 30% more than you usually would think it is based on the data for the last 3, 4 years.

21:59.926 –> 22:11.810
[SPEAKER_02]: That’s how in a deal world it should look like, but in reality how many companies can actually build a system and test it and invest, it’s an expensive thing to have and it’s an expensive thing to invest in.

22:12.652 –> 22:16.399
[SPEAKER_02]: So depending on the size it will basically depends on your outcome.

22:16.767 –> 22:21.880
[SPEAKER_00]: And of course, if you’re running it once a year, the ad platforms change things all the times.

22:21.920 –> 22:23.484
[SPEAKER_00]: There’s a little question.

22:23.524 –> 22:27.374
[SPEAKER_00]: Let’s not go into here about the validity of it in the first place.

22:27.394 –> 22:29.920
[SPEAKER_00]: But yeah, it can be hugely, hugely useful in the right places.

22:29.980 –> 22:33.088
[SPEAKER_00]: Now, Roman, you’ve mentioned AI.

22:33.068 –> 22:48.597
[SPEAKER_00]: You’ve also mentioned that you love a lazy solution, so any, any cool lazy solutions, any cool ways of utilizing AI in the area in which you work that it would be called to share with the audience, any tips.

22:49.066 –> 22:54.016
[SPEAKER_02]: They’re very close to what I’m working on, but they’re also very close to e-commerce.

22:54.537 –> 22:57.444
[SPEAKER_02]: So there has been two cases recently.

22:57.564 –> 23:04.358
[SPEAKER_02]: One of these Google ad scripts, we even as a big company as we are, we still do use Google ad scripts, they’re great for a lot of small things.

23:04.338 –> 23:07.241
[SPEAKER_02]: And I cannot write it natively.

23:07.261 –> 23:09.102
[SPEAKER_02]: Like I don’t know JavaScript at that level.

23:09.603 –> 23:18.170
[SPEAKER_02]: What I did is I was Gemini to create me all the documentation about it, put it in a file, and then I uploaded a file to Chaggipeteen, said, that is what you can use.

23:18.631 –> 23:22.975
[SPEAKER_02]: And it completely solved all the problem of it using in existing functions.

23:23.555 –> 23:26.478
[SPEAKER_02]: And now I have a system that writes really good Google ad scripts.

23:26.958 –> 23:29.901
[SPEAKER_02]: But I also can get it instructions, how to do things.

23:30.502 –> 23:34.345
[SPEAKER_02]: And then I thought, look, what if I do the same with a CRR

23:34.325 –> 23:41.986
[SPEAKER_02]: I think it was Gemini deep thinking to research the internet, that provide me all the advice is how to make a good landing page.

23:42.668 –> 23:50.089
[SPEAKER_02]: Put that in the paper and then come to JPT and say, by the way, I’m changing the brands that actually doesn’t matter which one you use.

23:50.069 –> 24:18.878
[SPEAKER_02]: Can you please analyze this page against this instruction provided in this structure thing and then provide it with a prioritization point for the project manager so I can make my pitch inside the company without writing it and then finalize it with a lovely message to my let’s say colleague Fred that I will try to pitch this idea in the first place and then you have this amazing system that comes up with a top five greatest things you should test on your landing page without you investing much time it’s good enough to get the ball rolling

24:18.858 –> 24:28.953
[SPEAKER_02]: It’s never good enough to create something absolutely amazing, linear in grade, but it saved so much time and effort into just getting, okay, is that learning page actually good or not?

24:29.594 –> 24:30.155
[SPEAKER_00]: I love that.

24:30.636 –> 24:44.497
[SPEAKER_00]: Would does it only work if you use one AI LLM tool to do the research, and then a separate one to be the active day-to-day helper?

24:44.517 –> 24:45.979
[SPEAKER_00]: Or could you do it all on the one?

24:46.415 –> 24:47.557
[SPEAKER_02]: You can do it all in the one.

24:47.857 –> 24:51.702
[SPEAKER_02]: I think it comes down to talking restrictions and how much time you have.

24:52.363 –> 24:56.048
[SPEAKER_02]: I would assume that majority of people listening are not very technical.

24:56.088 –> 25:04.499
[SPEAKER_02]: So for example, going through things called Gemini CLI, which is amazing to just scraping the data and that you can ask to it again and again in this free.

25:04.980 –> 25:09.947
[SPEAKER_02]: But it takes by some time to set up and tropic, what’s their AI do you remember?

25:10.367 –> 25:12.570
[SPEAKER_00]: Oh, no, I’m…

25:13.031 –> 25:13.131
[UNKNOWN]: Cool.

25:13.533 –> 25:15.538
[SPEAKER_02]: The law allows you to do it from the browser.

25:15.658 –> 25:16.480
[SPEAKER_02]: It’s amazing.

25:16.660 –> 25:17.783
[SPEAKER_02]: Like, I love Cloud.

25:17.863 –> 25:19.387
[SPEAKER_02]: I was doing renovation recently.

25:19.407 –> 25:21.753
[SPEAKER_02]: And like, can you go and research UK laws and everything?

25:21.793 –> 25:24.519
[SPEAKER_02]: Because they should do aware of when them doing renovation.

25:24.559 –> 25:25.762
[SPEAKER_02]: Because they’re doing it for the first time.

25:25.782 –> 25:27.546
[SPEAKER_02]: It does exactly the same thing.

25:27.647 –> 25:30.293
[SPEAKER_02]: It just has a small limit.

25:30.273 –> 25:38.305
[SPEAKER_02]: The only reason why I was using in the challenge of GPT is that at the time there is thing called custom GPT so you can upload the commands and share it internally in the company.

25:38.866 –> 25:45.977
[SPEAKER_02]: Gemina has gems but you can’t share it with other people, at least you couldn’t at the point when I was doing that, same with untrumpic.

25:46.297 –> 25:50.003
[SPEAKER_02]: So it’s essentially just me using

25:50.303 –> 25:52.446
[SPEAKER_00]: Got you, so then you can share it with the rest of your team.

25:52.787 –> 25:58.155
[SPEAKER_00]: So they can also do the Google scripts and they can also review landing pages without having to rebuild the whole thing.

25:58.756 –> 25:59.176
[SPEAKER_02]: Exactly.

25:59.217 –> 26:01.220
[SPEAKER_00]: Yeah, it evolved so fast.

26:01.260 –> 26:02.201
[SPEAKER_00]: We have to keep changing.

26:02.361 –> 26:09.793
[SPEAKER_00]: And you mentioned that about advising on landing pages, et cetera, I would imagine.

26:09.813 –> 26:18.606
[SPEAKER_00]: I think it’s very easy for people to assume that an analytics data team are purely reactive

26:18.586 –> 26:23.914
[SPEAKER_00]: you know, people send you work request and you do the work request or you notice something’s broken and you fix it.

26:24.454 –> 26:32.927
[SPEAKER_00]: But it sounds like you’re quite proactive as well in terms of analyzing data, making suggestions to teams and so on and so forth.

26:32.967 –> 26:34.789
[SPEAKER_00]: Is that a big part of what you’re doing as well?

26:34.809 –> 26:39.556
[SPEAKER_00]: And you mentioned earlier about educating the rest of the company on this.

26:39.576 –> 26:47.648
[SPEAKER_00]: So is there a big amount of liaison and advice flowing in both directions between you and other teams in the business?

26:48.151 –> 26:54.482
[SPEAKER_02]: Yeah, and we have a lot of heated discussions, and sometimes we shout at each other but we still love that interactions.

26:55.143 –> 26:59.811
[SPEAKER_02]: That’s all with the limits, but what I mean by that is we try to challenge.

27:00.072 –> 27:04.500
[SPEAKER_02]: I have a lot of experience to bring to the table, but there’s a lot I need to learn and understand.

27:04.540 –> 27:06.463
[SPEAKER_02]: For example, when it comes to…

27:06.443 –> 27:14.411
[SPEAKER_02]: some lending phase optimization, I know some basic things that’s going to be good to very good for Google ads and how to connect things together.

27:14.731 –> 27:17.814
[SPEAKER_02]: But then the product owner would know so much that about the user journey.

27:18.075 –> 27:26.863
[SPEAKER_02]: And then product analysts who can bring you something like a cohort analysis, or you actually people who are hitting that page are mostly returning users rather than being new users.

27:27.284 –> 27:33.430
[SPEAKER_02]: So having right people at the table is very important,

27:33.410 –> 27:59.569
[SPEAKER_01]: not having 25 stake holders in a page trying to define what’s the best headline, which happens quite a lot in a company nowadays.

28:01.810 –> 28:07.019
[SPEAKER_00]: Okay, I love this section because me and our listeners some really quick ideas for taking our businesses to the next level.

28:07.419 –> 28:09.242
[SPEAKER_00]: Roman, are you ready for the top tips?

28:09.863 –> 28:11.366
[SPEAKER_02]: As far as I can be ready.

28:12.928 –> 28:13.389
[SPEAKER_00]: Excellent.

28:13.529 –> 28:14.812
[SPEAKER_00]: Okay, the book top tip.

28:15.212 –> 28:22.384
[SPEAKER_00]: If everyone listening to this podcast agreed to take Friday off and read a book to make their business better, which book would you recommend?

28:22.769 –> 28:24.211
[SPEAKER_02]: I’ve read it quite some time ago.

28:24.251 –> 28:26.994
[SPEAKER_02]: It’s called decoded the size behind why we buy.

28:27.014 –> 28:37.948
[SPEAKER_02]: I find it’s very easy to read because I have a lot of basically very simple examples to explain to people why people pick this price versus that price, which is very good for a commerce to know.

28:38.789 –> 28:45.157
[SPEAKER_02]: But it also way easier than to read the thing fast and slow, which is very hard to get through.

28:45.197 –> 28:47.139
[SPEAKER_02]: So I would probably start there.

28:47.179 –> 28:51.004
[SPEAKER_02]: If you haven’t read the thing fast and slow, read decoded,

28:51.473 –> 28:54.776
[SPEAKER_00]: I often wonder about Thinkfast and slow in the paper back edition.

28:54.896 –> 28:58.479
[SPEAKER_00]: If someone at Penguin went, how can we make this even harder to read?

28:58.620 –> 29:00.561
[SPEAKER_00]: Let’s make it a teeny, tiny font.

29:01.883 –> 29:02.203
[SPEAKER_00]: It sucks.

29:02.223 –> 29:10.190
[SPEAKER_00]: So when you pick it up and just go, oh, Daniel Kanderman, I would love to read this, but I just can’t, it’s gonna be something else today.

29:10.490 –> 29:11.832
[SPEAKER_00]: So love that recommendation.

29:12.552 –> 29:19.759
[SPEAKER_00]: The traffic top tap, tip even, which marketing method do you either prize above all others or think doesn’t get the press it deserves.

29:20.043 –> 29:22.286
[SPEAKER_02]: I think it’s influenced the marketing done right.

29:22.666 –> 29:29.835
[SPEAKER_02]: I think a lot of companies go for influence from marketing and they either do go big or try a little bit, just tiny little bit.

29:30.336 –> 29:42.270
[SPEAKER_02]: And I have seen the multiple companies who just picked the right influencers and the right time with the right offer, having it very done very nicely and it comes on a basically lead-gen, lead-gen.

29:42.250 –> 29:47.759
[SPEAKER_02]: or the driving business very well, but making it right is difficult.

29:48.280 –> 29:49.882
[SPEAKER_02]: The same thing happens with Google Ads.

29:50.083 –> 29:55.011
[SPEAKER_02]: We just get used to the idea that we need to make everything they’re perfect to everybody knows how to use it.

29:55.351 –> 29:57.054
[SPEAKER_02]: But you come to a new platform.

29:57.074 –> 29:59.378
[SPEAKER_02]: You’re like, it just just click a couple of things.

29:59.959 –> 30:04.145
[SPEAKER_02]: I think that’s where content marketing on first marketing get really misunderstood.

30:04.346 –> 30:08.933
[SPEAKER_02]: You also need to have a very thorough discussion and don’t

30:09.267 –> 30:16.241
[SPEAKER_00]: It is such a clever science, even though it doesn’t appear like a science in the way that Google Ads does.

30:16.441 –> 30:17.704
[SPEAKER_00]: Yeah, totally agree with you.

30:18.245 –> 30:18.906
[SPEAKER_00]: Tooltop tip.

30:19.027 –> 30:28.305
[SPEAKER_00]: Maybe a collaboration tool, a social media plug in, a phone app or just a way of working, is the record little tool you use that makes you and your team more efficient day to day.

30:28.589 –> 30:32.794
[SPEAKER_02]: It has been one tool that I’ve been using for the last two years, which has been high touch.

30:33.415 –> 30:41.344
[SPEAKER_02]: High touch is essentially a platform that allows it to take data from your data warehouse, which is a set of Google Amazon, Facebook, anywhere else you want.

30:42.045 –> 30:49.534
[SPEAKER_02]: And that platform has literally transformed what we do internally in the data side in Carauh.

30:49.514 –> 31:02.907
[SPEAKER_02]: One of the greatest things we have been one of the first month to implement a user data matching in Google Ads, just because HiTage allowed us to have a latest API connection and everything very, very quickly.

31:03.428 –> 31:06.651
[SPEAKER_02]: They have amazing customer support and I just love working with it.

31:07.112 –> 31:17.682
[SPEAKER_02]: Just one thing to mention that is for more like an advanced shops and the users, if you have

31:18.134 –> 31:19.017
[SPEAKER_00]: very cool.

31:19.218 –> 31:27.267
[SPEAKER_00]: So is it a feed management platform or is it more of a strip back to the bones you can make it do what you actually want to happen?

31:27.331 –> 31:34.419
[SPEAKER_02]: It’s the second, but I’m also need to be fair that they have launched recently CDP, which is a customer data platform.

31:34.439 –> 31:42.929
[SPEAKER_02]: We can go and say, create new audience of people who have made an order over the last two weeks and send it to Facebook, DV360, to Google Ads and everything.

31:42.949 –> 31:46.734
[SPEAKER_02]: And it’s just going to do it all by itself very quickly.

31:46.894 –> 31:50.878
[SPEAKER_02]: And it’s going to show you all the statistics, what you’re missing, what emails are wrong.

31:51.399 –> 31:55.684
[SPEAKER_02]: It’s just such a click and drag kind of experience that works so well.

31:56.137 –> 31:56.719
[SPEAKER_00]: nice.

31:56.740 –> 31:57.744
[SPEAKER_00]: Love that recommendation.

31:57.844 –> 31:59.772
[SPEAKER_00]: Okay, finally the carbon top tip.

31:59.853 –> 32:03.970
[SPEAKER_00]: What’s your favorite way to reduce the carbon footprint of an e-commerce store?

32:04.203 –> 32:17.996
[SPEAKER_02]: The hard one for me, but I would probably be very boring and say, learning how to work remotely efficiently, because what you see in a lot of companies nowadays is that everybody goes remote, so you have 12 meetings a day, and then like, every multi-work doesn’t work.

32:18.937 –> 32:21.119
[SPEAKER_02]: It does, you need to find the right balance.

32:21.139 –> 32:32.370
[SPEAKER_02]: I’m a personal love, and it’s some time and office, sometimes remote, but do you hire people remotely, just need to know how to work with that, and how to set up correct targets and work with it efficiently.

32:32.873 –> 32:35.015
[SPEAKER_00]: Yes, totally get what you’re coming from.

32:35.035 –> 32:40.621
[SPEAKER_00]: I have say my heart sunk during the pandemic when there were all these stories of people putting spyware.

32:40.961 –> 32:47.708
[SPEAKER_00]: You know, everyone’s now working from home, so suddenly there was a boom in spyware to track if your team were actually working or not.

32:47.728 –> 32:52.152
[SPEAKER_00]: It’s like, if you can’t trust them to do their job, why did you hire them in the first place?

32:52.533 –> 32:57.818
[SPEAKER_00]: Although it says a lot more about the owner than it does about the staff, I think.

32:57.798 –> 33:03.298
[SPEAKER_00]: Anyway, before I go off on one about that, Roman love that tip, thank you and thank you so much for coming on the show.

33:03.338 –> 33:08.095
[SPEAKER_00]: Before we say goodbye, could you let people know we’re thinking in contact with you and Karl, please?

33:08.193 –> 33:20.999
[SPEAKER_02]: Yeah, thank you very much for having me here, connect to me on link it in, which is going to be Roman PetroChankov, or you can also go to caro.co.uk and try to find a new car or sell your existing car.

33:21.379 –> 33:28.193
[SPEAKER_02]: We’ll also exist in Germany and Spain, so if you form those countries also, caro.de in caro.es.

33:28.763 –> 33:30.627
[SPEAKER_00]: Awesome, Roman, thank you again for coming on.

33:30.667 –> 33:36.018
[SPEAKER_00]: It’s been very cool hearing a brand-side perspective on all these challenges we’re doing.

33:36.058 –> 33:40.767
[SPEAKER_00]: And you make it seem a lot less stressful than it does a lot of the time.

33:40.827 –> 33:42.852
[SPEAKER_00]: So thank you so much, Roman, really appreciate it.

33:43.353 –> 33:44.595
[SPEAKER_02]: Thank you very much for having me.

33:50.296 –> 34:10.431
[SPEAKER_00]: What brilliant pragmatic practical advice and loads of cool little tool type tips in there as well really cool to chat to Raymond he’s just suggested he should come back on every year and give us a little what’s going on in the world of analytics data etc update I think it’s a cracking idea if you think so to getting content with me on LinkedIn and let me know what you think.

34:10.411 –> 34:19.140
[SPEAKER_00]: So lots in there to unpack around how to go about making sane decisions about sorting out your data.

34:19.741 –> 34:27.329
[SPEAKER_00]: Loved the tips around consent banners and the tracking side of things and the be pragmatic about it all.

34:27.669 –> 34:36.098
[SPEAKER_00]: Get that balance between what you have to do and what the risks are and then make it work for your business and don’t get too stressed out about it.

34:36.078 –> 34:43.754
[SPEAKER_00]: you can get your hands on our notes from this episode including those top tips and links to what we mentioned by heading over to ecommercemasterplanned.com.

34:44.295 –> 34:57.782
[SPEAKER_00]: You can also use a direct episode short link that’s ECMP.info forward slash episode number and once you get to the website go on add yourself to my email list so you don’t miss any of the other things I share to help you improve your business.

34:57.762 –> 35:13.620
[SPEAKER_00]: If you liked this episode, then make sure you check out Episode 570, where Jamie Hill is one of our WWYD guests, that’s one of our what would you do guests and she’s getting into how you can massively improve operational efficiency.

35:14.521 –> 35:24.512
[SPEAKER_00]: And if you want to hear more episodes with massive retailers, then head to the website ECMP.info forward slash big to get your hands on all those episodes.

35:24.948 –> 35:29.012
[SPEAKER_00]: Thank you for tuning in to this and every episode of the e-commerce master plan podcast.

35:29.453 –> 35:39.943
[SPEAKER_00]: I bring you a new interview every week to inspire and help e-commerce business owners like you to succeed and thrive with your businesses, including progressing along the path to carbon neutral.

35:40.904 –> 35:46.290
[SPEAKER_00]: So, you know someone this show can help please tell them to listen to the e-commerce master plan podcast.

35:46.750 –> 35:49.453
[SPEAKER_00]: I hope you have a great week and don’t forget to keep optimizing.

35:52.505 –> 35:55.981
[SPEAKER_01]: Thank you for listening to the e-commerce Master Plan podcast.

35:56.383 –> 36:01.688
[SPEAKER_01]: Find out more at e-commercemasterplanned.com slash podcast.