Scaling Without Pain: The New Rules Of Ecommerce Acquisition — Tiago Costa | Why Meta Ads Fail Scaling, Why Ad Spend Hits Limits, How AI Finds Ideal Buyers, How Data Drives Better Sales, How Shopify Data Fuels Growth, How One-Click Campaigns Work (#460)

In this episode, we explore how first-party data and AI are fixing the problem of rising ad costs. 

Tiago Costa Rocha, CEO of Full Venue and the creator of Clustie.ai, explains why traditional Meta ads are failing many Shopify brands today. 

He shares how brands can stop guessing and use their own customer data to find high-value buyers. Tiago also breaks down how to scale campaigns faster using "one-click" AI tools to lower costs and increase sales.

Topics discussed in this episode:  

• How rising ad costs affect Shopify growth. 
• What causes the "feedback loop" in ad spend. 
• Why brands must identify their true ICP. 
• How Clusty connects directly to Shopify data. 
• What the one-click campaign builder automates. 
• Why first-party data beats platform algorithms. 
• How AI identifies high-propensity buyer groups. 
• How to scale budgets without killing results. 
  
  

Links & Resources 

Website: https://www.fullvenue.ai/
Shopify App Store: https://apps.shopify.com/clustie-ai-marketing-segments
LinkedIn: https://www.linkedin.com/company/fullvenueai/

Get access to more free resources by visiting the show notes at https://tinyurl.com/yd2uxe9p

______________________________________________________

LOVE THE SHOW? HERE ARE THE NEXT STEPS!

Follow the podcast to get every bonus episode. Tap follow now and don’t miss out!

Rate & Review: Help others discover the show by rating the show on Apple Podcasts at https://tinyurl.com/ecb-apple-podcasts

Join our Free Newsletter: https://newsletter.ecommercecoffeebreak.com/

Support The Show On Patreon: https://www.patreon.com/EcommerceCoffeeBreak

Partner with us: https://ecommercecoffeebreak.com/partner-with-us/

Transcript

00;00;00;01 - 00;00;22;18
Unknown
In the end, the behavior of your customers, they tell you a story. The data storytelling is a thing. And when you connect all these data from your Shopify, everything related with purchases, related with abandoned cards, related with customers, related with a value that the customer spends when they spend, all this information is going to be analyzed.

00;00;22;20 - 00;00;48;05
Unknown
Hello. Welcome to another episode of the E-commerce Coffee Break podcast. A lot has changed in e-commerce in 2025 at cost kept rising, tracking kept getting weaker, and for many Shopify brands, better ads just don't work like they used to. It's a real growth problem. So today we're tackling a bigger question how our first party data and rewriting the rules of e-commerce acquisition and more importantly, how can brands take control instead of just guessing inside at platforms?

00;00;48;07 - 00;01;07;08
Unknown
To break this down, I'm joined by Tiago Costa. He is the CEO of full venue and the force behind the Clustie.ai platform, helping e-commerce brands use their own customer data and AI to find high value buyers and to run smarter campaigns. Tiago, great to have you on the show. Thank you very much for this class. It's a pleasure to be here.

00;01;07;11 - 00;01;34;04
Unknown
And as just saying that, you you briefly resume everything that's on our mind and the reason why we are here and the reason why we are doing this, it's exactly to tackle that kind of problems and help brands to use intelligence with their brands. Really simple. So, yeah. Happy to be here. Thank you. Let's dive into it. So many Shopify brands feel like they're just burning cash on better ads these days.

00;01;34;06 - 00;01;54;14
Unknown
What's the problem? Why is that so difficult nowadays was met. It's it's it's a it's like a feedback loop, you know, because let's focus on the kind of brand that that sells. Okay. Let's let's create this assumption. We will talk about a brand that is selling. They have a marketing department or even to the management and they decide to.

00;01;54;15 - 00;02;24;07
Unknown
Okay, let's will. So to grow, we need to go for different channels. We need to to act on different marketing pieces. And one of the pieces are the advertising platforms. And the advertising platforms. They need budget to show what you have your product to a specific audience. And the thing is, every time that you invest more to get more sales, you are targeting to a broader audience.

00;02;24;07 - 00;02;45;12
Unknown
You are targeting to people that you don't know at all. And there's a problem because, okay, you invest ten and you sell 20. If you invest 20, you won't sell 40. And and this is this is the problem because in the end, the brands start to increase the budget that they have allocated on advertising and they don't see results.

00;02;45;15 - 00;03;14;23
Unknown
And what they do, they change the marketing team, they change the marketing agencies, they change their mindsets. Probably they change their products, but they didn't look into their data. They didn't look into what were their are their customers, what they can do with their customers to identify exactly what's their ICP and as a startup. And, you know, it's we need to define our recipe and that's the word that's when a startup is going from from 0 to 1.

00;03;14;25 - 00;03;37;23
Unknown
If you don't know exactly the ICP, you'll be just running in circles without finding the people that watch our product or even to understand if your product has a fit. So for all the brands, there are spending money on ads. They need to find their recipe. And we have a tool that is finding their ICP. We they on top of their data with a couple of weeks.

00;03;37;26 - 00;04;03;04
Unknown
Mm hmm. I think right now a lot of people have the feeling specifically was they better at platform because everything is a I know that they're just basically have a black void and hope for the best. What better figures out back in the good old days was better was Facebook. You had a very granular targeting and you really could work out and try to target your exact ICP.

00;04;03;04 - 00;04;23;27
Unknown
That's not longer the case now as matter is not providing you this information, I think it was cluster you came up with a way to to help with that. Let's dive right into it. So clarity is obviously your brainchild and you growing it quickly. How does it work? Yeah, let's see. It's a simple tool that connects with the Shopify stores.

00;04;24;00 - 00;04;46;17
Unknown
Okay. That's for all the Shopify brands. So brand selling on the Shopify platform, they can go to Shopify App Store, they can download clips, they can install clips in this case. And when installing, let's see what they will do. The process that they will pass through is connect to their Shopify store to cluster. We will bring all the data.

00;04;46;18 - 00;05;24;25
Unknown
Okay, good. Seamless products and are there So first party data, pure data, no tricks, no secret sources. So only the data that's the brands have on their Shopify stores. Then inside of our tool, we also ask the brands to connect their matter account in order that we can have this bridge between you and the advertising accounts. Okay. And inside our black box question, we have a set of models that are identifying the best customers inside the databases from the brands.

00;05;24;28 - 00;05;54;29
Unknown
The thing is, we have this powerhouse of segmentation, and this power of segmentation is powered by a and is talking to you, so is showing you which kind of segments you should activate. It's allowing you not only to integrate the segments to matter. You have like a button where you can press the segment to matter. But if you are a marketer that you don't want to do the thing that you did all the time before.

00;05;55;00 - 00;06;32;14
Unknown
So imagine that you are a performance marketing engineer or a performance marketing manager, and you are creating a campaign and you always create the Met the campaigns accordingly to what you know, what to what the brand told you to do. If you want to challenge that mindset, we created a feature that is called the one click Campaign Builder that not only inserts the audience in matter, but creates entire structure of a matter campaign with campaign ad set and ad based on all the information that the brand as on their account.

00;06;32;14 - 00;07;07;10
Unknown
So we are running a set of business rules, putting some on top of it and creating this campaign with one click. And the only thing that the customer needs to do is to go to my account and click like. So yeah, so really simple, really straightforward. And this resulted from, I would say, three years of exploration, three years working as a marketing agency, doing this kind of thing, but connecting with tech teams to the different APIs of the different data sources, connecting to the different advertising platforms and after, of course, raising capital.

00;07;07;10 - 00;07;26;03
Unknown
We tried to productize what we were doing and we arrived too closely. MM hmm. Now, it's actually quite interesting that you said you're working. Basically, you come from the problems that you have. You produced a tool that helps with sort of solving this problem and with the help of I now we're talking about high intent audiences and that's what everyone is after.

00;07;26;05 - 00;08;01;09
Unknown
What does that mean in simple terms? In the end, the behavior of your customers, they tell you a story. So the data storytelling is a thing and not from today, but from a couple of years before data as something to say. And when you connect all these data from your Shopify, everything related with purchases, related with abandoned cards, related with customers, related with the value that the customer spends when they spend the the sales peaks that you have during the year, depending on the seasonality of what we are talking about.

00;08;01;15 - 00;08;22;09
Unknown
So all these information is going to be analyzed not by our team, but by our models. Okay. So in the models, we are talking about models, really simple models that RFM models, and then we can put on top of RFM, we can put clustering on top of it. So we have like a set of models and generating different audiences.

00;08;22;12 - 00;08;51;25
Unknown
Right now we are fully focusing to different types of audiences. One of them, we call it the propensity bias. So if the brand is pushing for a conversion campaign and usually all of them have some conversion campaigns that they push, are they they really burn budget There. They should use the propensity buying audiences. But we also developed a model that is performing really well, the LTV model that is focused on always going to buy on the next 30, 60, 90 days.

00;08;52;02 - 00;09;18;03
Unknown
And we are also allowing brands to try and stress sports to to match to this kind of segment from our product. So when we say I propensity buyers are people that look like they will buy, this is all based in data. There is no guesswork, there is no manual work. So the work was building The models. We the tools that we have was creating this platform that is running these models really slightly, really fast.

00;09;18;03 - 00;09;43;14
Unknown
And the thing is, is to be really fast and simple and integrating them and then challenging everything that matter and Google or what they are doing in terms of AI because of course we are talking about huge players with the infinite capacity to develop, AI to develop, to develop these different strategies in terms of of conversion campaigns, for example.

00;09;43;16 - 00;10;09;24
Unknown
But at the same time, everything that they launch gets really crowded really fast. And the ones that think slightly different and what we are doing is not more than just giving intelligence on the brands data and there is any platform, any platform, neither open air, neither matter, neither Google that can have this, that the brand has the brands as this information and it's on their ends.

00;10;10;00 - 00;10;36;29
Unknown
And we are allowing the brand just to connect their platforms, their infrastructure, and using this kind of audience brain to leverage the results on the marketing side. So it's all about efficiency, what the past more or less manual that goes into the AI. And then obviously you can scale quite easily from there. Now, was every a model training is an issue with your system, was you how long is the training before it really can get started?

00;10;37;02 - 00;10;57;23
Unknown
Yeah. So usually it depends on the platform that you are delivering your ads because the segmentation itself, we can produce it in seconds once we have. So when you are installing cluster, you have like a loop. When we are bringing the data from your Shopify and depending on the size of your Shopify, it can have some minutes too to load all the data.

00;10;57;23 - 00;11;27;02
Unknown
And this is not a this is not something that we can control because Shopify obliges you, obliges you to make several requests depending on the amount of data that you have. So we need to go through all the process, but once we have the data, the models are calculated really fast. So in seconds and the segments are generated and integrated in method, but after being in matter, after being launched in a campaign, you have the learning curve, you have the learning phase.

00;11;27;08 - 00;11;58;17
Unknown
And what we saw during all these tests that we've been doing since we launched the class is, is that our segmentation takes around two weeks to perform properly, 2 to 3 weeks to perform properly. And that's why we are also giving this 30 days trial to all the brands that wants to test our product because we want them to really make a test, because more than just trying to listen and see what happens, they need to wait 30 days to see the real value, but they won't be paying for it.

00;11;58;19 - 00;12;20;16
Unknown
And at the same time, we are getting a lot of information, a lot of good feedback from different customers. Some of the feedback are gaining because we are following everything that they are doing, some of the feedback given from the customer itself. So so yes, really, really powerful. I think it's a great offer and patience is something that a lot of digital marketers, me included, do usually not have and was better.

00;12;20;16 - 00;12;43;04
Unknown
Obviously you need to wait until you have the number of conversions or the time scale to really see what happens. You mentioned something earlier, which I'm interested to find out more about, and a lot of our listeners have been in a situation when do you scale? You throw more money on your well-performing campaign, it kills the campaign. How does your system help in preventing this?

00;12;43;06 - 00;13;10;20
Unknown
Yes. So what we are doing is we are if you if you look into impressions and the reach of a normal or broad event advantage plus marketing campaign, the numbers are huge. So you are targeting a lot of people. By using our kind of segmentation immediately, you will have lower impressions, a lower number of impressions and a lower number of reach.

00;13;10;23 - 00;13;34;06
Unknown
So you are reaching less people. It allows you to multiply the budgets that you put on that campaign, because imagine if you have a conversion campaign reaching 1 million people that is performing well and you invest in that campaign, probably the results will start to slow down because you have 1 million and probably there is now much more people.

00;13;34;06 - 00;14;00;19
Unknown
So engage with our brands. But if you are looking to specific people that came from your database and even that you built lookalikes on it, that's what we do to find new customers. We build to look alike on site audiences. The even though the the impressions in the region are not so high so you can put more money on these campaigns because you are focusing on niche, on a niche, the less people, less money.

00;14;00;21 - 00;14;25;16
Unknown
Customer acquisition costs lower the average order value usually is higher. So you have a few metrics that you can play with and you can put more money on those campaigns at the same time, you need always to have in mind that you should look into all the campaign details. And the campaign is not only the budgets, is not only the audience, is not only the copies, the creative.

00;14;25;16 - 00;14;53;13
Unknown
So it's it's everything. It's all it's working together. And that's why we we also have this future coming in and we called it the coolest insights that is talking with a customer every day in our platform saying what's happening mainly related with audiences and with budget optimization, because that's where we can play, but also giving some other advices to the customer, things that you could do to increase the results of their campaigns.

00;14;53;18 - 00;15;17;28
Unknown
Mm hmm. It's actually something I wanted to ask you. So if you have connected closely with Shopify, with your platform, how does the typical day to day work from a marketing manager look like when they're working with was closed? Yes, super simple. So they setup every time that they are planning a campaign launch. They include the classic campaign or the glossy segments inside of their campaigns.

00;15;18;00 - 00;15;50;10
Unknown
So thinking about close to yet as the I propensity buyers. So higher is lower customer acquisition costs. So this kind of of ad said let's say these ideas should be in the campaign and launch the campaign and following the numbers. So I would if if I was the marketing manager of a brands and if I was using to, I would go there every day to see the return of the classic campaign, to see if I have some information coming, relevant information of something that I should change in matter.

00;15;50;13 - 00;16;15;00
Unknown
And at the same time, every time that I need to refresh audiences, every time that I need to create new campaigns, go again to close to do the process again. And yes, just wait for the results to come. But in the end, this is like a tool that should be used during campaign launch, during campaign optimization, and super simple 5 minutes inside close to and you can do your work after.

00;16;15;06 - 00;16;46;18
Unknown
Okay, excellent. Could you share some success stories or case studies? You don't need to name a brand of businesses that have effectively leveraged this technology. Yes, of course. Something that is getting really interesting right now is the reach that we are having. So it's fully different to run an agency or to run a product. So fully difference. And right now, even even we the agency with the agency style, we were doing projects all over the world in five, seven geographies, something like that.

00;16;46;21 - 00;17;18;02
Unknown
But today we are having installations from all around the world because even even focusing on advertising for specific parts of the globe, we are getting we are getting people from all other parts of the world. So inside these brands, we have some really interesting cases. And I have a case of a brand that is coming from Colombia, if I'm not mistaken, that they have a lot of campaigns when we are talking about the vertical of beauty and cosmetics.

00;17;18;04 - 00;17;41;00
Unknown
Okay. And they have several campaigns for several countries all over the world. And what they decided to do was to create a second campaign like a Navy test. And that's something that we we like to say to to everyone that is installed in closely that you don't need to do a Navy tests to see that list as a name that on inside the campaigns.

00;17;41;07 - 00;18;13;18
Unknown
It's another thing that you have another tool that you have to generate sales that are more qualified, that are better sales, that will reflect in better first party data. But some customers, of course, they want to test. They want to be sure that they should pay for the product and they duplicate the campaign. So in this case they were targeting Deutschland, they were targeting Spain, Italy, France, so a lot of countries in Europe and they just duplicated the campaign and create a new campaign with the same name and the score.

00;18;13;20 - 00;18;55;11
Unknown
The results are super impressive even for me, to be honest, because there there is no any country where close to is not over passing the results of the original campaign. And we are talking about return on ad spend, 50% higher customer acquisition costs like 15 to 20% lower, and that your average order values. Then we have cases. I can tell you that usually the worst cases that you have that we have right now with the brands that installed close to our cases, where we are keeping the same number, the same performance of the best campaigns that they have.

00;18;55;14 - 00;19;19;29
Unknown
Okay, So imagine a company that is working on a 3.0 host and we launch the classic campaign and we got that 3.0 ROAS. But with 10 minutes of installation and one click of campaign launch. And this is crazy because probably the best campaign that the customer have took years to generate those results. And we we have it in seconds.

00;19;20;01 - 00;19;39;14
Unknown
So we have this range of the stores that are more or less with the same results. Super challenging to go to convert a star like that from our free to brave. So when to try a lens and you want to convert it to a paid customer, it's super challenging to convince them to stay when the results are the same.

00;19;39;19 - 00;20;01;08
Unknown
But we what we try to explain is that the product, it's much more than increasing the rules directly. So you have new audiences that will have a compounding facts effects that will over over the the numbers that you have, but not in the first three weeks, but probably in the first three months that the same time we will have a full awareness on the data that you have.

00;20;01;08 - 00;20;26;00
Unknown
You will have segment segmentation being generated with one click, but then we have all the other customers where we are surpassing the numbers that are a lot of them. So right now we are with a we the free to play with tax of around 50%. That is crazy. And I'm when I'm talking with the team, I'm always telling that probably we won't be able to keep this kind of numbers when we increase the funnel.

00;20;26;05 - 00;20;52;00
Unknown
So when we are having like 50 installations of they probably we will not get 25 of those 50 transforming into a paid customer in the second month. But we are driving this figure. So right now we are talking about an average increase of 50% on the return that spends lowering the customer acquisition cost from 15 to 30%. And the average are they're valuing the same proportion 15 to 20% either.

00;20;52;08 - 00;21;10;02
Unknown
So these are the numbers that we commonly have, and that's the power of AI. And even if you're not increasing massively, you're saving a ton of time and this is time as money. So basically you only can win. Now, you touched a little bit on the onboarding. That is very quickly, but walk me through the typical onboarding process for a new user.

00;21;10;02 - 00;21;31;26
Unknown
What steps, how involved, how long does it take to get up and running? Yeah, so usually in two days you have your campaign running already, but we have two different onboarding types that we are testing them. One of them is fully organic. So imagine clause that you have your Shopify store and after this this interview we will to install to you will do it all by yourself.

00;21;31;28 - 00;22;08;27
Unknown
So I would say that probably in one or 2 hours you can have your campaigns running in in matter. And tomorrow we will be able probably to have a conversion or something like that. And then you have the second level where we are taking the customer by the hand. Okay, So our customer success team sees that someone is interested because they feel the form in one of our landing pages or someone installed close to on Shopify App Store and installed a couple of minutes later and we are on top of it.

00;22;08;27 - 00;22;41;14
Unknown
We are calling, we are sending an email and we are helping the customers. So we schedule a meeting probably for tomorrow or the day after tomorrow. We have like a 15 minutes onboarding call where we leave everything ready on the the client side and then with all the customers, right now we have time allocated to them and we have like 15 to 20 minutes weekly with all of them just for feedback because everything that we need right now, it's feedback from those using cluster and of course that we have our insights.

00;22;41;16 - 00;23;03;07
Unknown
We, we try to pass the best way to leverage the products and what we are doing in the advertising platform. But as important as explaining to the customer how to use cluster is to getting feedback from the customer, because when you are launching a product, the customer knows what you need to have. So we need to bring it.

00;23;03;14 - 00;23;25;08
Unknown
Yeah, yeah. It's a, it's a win win situation, I think from both ends there. Now obviously you can save a ton of money with clarity, but how does your pricing structure work? How do you charge for it? Yeah, we are, we are just iterating around it. So working a lot Sterling a lot. We wanted something that was like a no brainer for the brands.

00;23;25;15 - 00;24;08;04
Unknown
Okay, So right now if you go to Shopify and they sell close to, you will see that we have a range of we have three different prices right now on the on the pricing model, depending on the amount of customers that you have on your Shopify store. Okay. So right now it's the way that we are working and you have like $80, $200 and $400 and $400 is for a huge star with more than 150,000 customers in the database in between 10,000 and 150,000 of the the mid-tier and the $79, you have 44 for shops with less than 10,000 customers.

00;24;08;07 - 00;24;35;13
Unknown
So the idea of this pricing is to be really simple, really easy, no brainer, no brainer to pay this because everything that we are doing, time saving all this intelligence advertising insights, we that actionable actions. When you are putting a price on a tool, you need to look into into your competition because because you need to you need to fit with our competition and we need it also.

00;24;35;16 - 00;24;55;16
Unknown
So we believe that right now we are ready to to go for the numbers that we need as a startup. We need to reach more than 1 million a year by the end of 2026 because we have investors, we have goals. We need to go to the series a and in the end of the year. So we believe that we are ready to to go there.

00;24;55;18 - 00;25;14;14
Unknown
And on top of that, you have a 30 day test drive. So I think that's a no brainer to try it out. And then talking about 300 or $400 a month for a bigger brand, that's actually peanuts. There's not really much money for what actually you can save on time and get a better conversion step before our coffee break comes to an end today.

00;25;14;14 - 00;25;33;27
Unknown
Is there anything you want to share with our listeners that we haven't covered yet? No. The message is, is the one that we are trying to pass since the beginning. There is we have this tool. We have an incredible team doing this here in Portugal. That's that's crazy. Portugal is is growing a lot in terms of technology, in terms of the startup startup landscape.

00;25;33;29 - 00;25;58;28
Unknown
It's doing incredible things. And we have this dream of transforming this company on a unicorn and more than that, to democratize the way brands use all this intelligence on their on their databases. So we want to be a really simple tool that is being efficient, generating more sales, finding high value customers with a couple of clicks, and everybody's is happy in the end.

00;25;58;28 - 00;26;28;12
Unknown
So yes, and the most important message is this is work in progress. So what we are today, we won't be tomorrow, we will be better and we will be better if you have if we have more feedback from customers and this loop that we are creating here and even internally, the message that I try to pass to everyone in the company is that we are continuingly learning everything that the customer needs to tell us in order to be better tomorrow.

00;26;28;15 - 00;26;49;13
Unknown
And allies are a huge chance for new customers as well. So for our listeners, if you have feedback, please share it with Tiago. It will help you at the end of a day. Now where can our listeners go and find out more about you? If they want to know about Calista, they should go to CAA. They can go also to our company website, full venue.

00;26;49;14 - 00;27;10;08
Unknown
So full venue is the mother of Calista so close to resides under the umbrella of venue. So it's a product that we launch this year. You will find everything on an also on our social networks, on LinkedIn, on Instagram, our full venue. Yes. And of course, you can try to reach me, send me a message if you have any question, any feedback.

00;27;10;08 - 00;27;31;01
Unknown
If you are interested in StarLink list, you feel free to do it all by yourself because all the team built a process for you. The brand, to to be able to do this without any action on our side. So yeah, but feel free to reach out to reach. I am always super interesting learning more about the customer feeling on what we are doing.

00;27;31;03 - 00;27;48;16
Unknown
I will put the links in the show notes and you're just one click away. To our listeners, reach out to Tiago, try clarity out. I think it's a great solution to make your life easier when you're on metalheads, and thank you for giving us an insight also into a startup. It's always exciting times. Thanks so much for your time today.

00;27;48;22 - 00;27;50;01
Unknown
Thanks a lot. It was a pleasure.