The Feedback Loop Is the Missing Layer in Performance Marketing

Machine-generated transcript from the episode video. Speaker labels are not included and some names and product terms may be transcribed phonetically.

[0:00] The ad shipped, the creative looked great, the campaign went live, and performance still stalled. Not because targeting failed, not because budget was wrong, but because teams scaled ad volume faster than their ability to learn. In this episode of GTM Vault, George House explains why most performance marketing failures are not media problems. They are feedback loop failures between creativity, data, and iteration. Welcome to GTM Vault. Trusted by over 25,000 founders and operators building modern revenue systems. Today's episode is about what actually breaks his company scale performance marketing. Not spend not channels, not learning velocity. My guest is George House, head of Canva Grow. Before Canva, George founded Magic Brief, a creative intelligence platform acquired by Canva. Before that, he built his career inside top creative agencies, watching the same failure repeat. Teams launch ads, they collect results, and then they move on. Learning never compounds. This is not a conversation about ad formats or tools. It is a conversation about GTM feedback

[1:07] loops. Today's core question, what breaks when teams scale ad volume faster than their ability to learn? Let's get into it. George, you have seen this from agencies from a founder seat and now from inside Canva. What pattern kept repeating that convinced you that this was a system failure, not a creative talent problem? Good question, Rick, and thanks for having me on the podcast. I'd say the main thing that I saw break was really, to your point, that lack of a feedback loop. So the teams or the creative roles that were responsible for creating the assets, whether they were designers, video editors, copywriters, they weren't always the same people that were looking at the data and the ads manager and the spreadsheets that were showing what was actually working with inside the ad accounts. I think a lot of that comes from creative people in their nature aren't normally the same set of people that are looking at those large tables of data and inferring patterns about what is working and why. So without having that kind of direct way of learning from ads that were running in the account that was the biggest I guess lack of a bridge between the actual context and the data and what was going live with inside the ad accounts.

[2:14] What pattern um okay got it. When teams say creative stop working what is actually breaking underneath? Yeah, I'd say the thing that breaks underneath is the inability to take the learnings from your ad account and have those passed forwards to the teams that are making the advertising. So, what is breaking is essentially the ads that you will make tomorrow aren't based on the learnings that came from your ad account yesterday. And there's various ways that that happens. It's either teams are using agencies that don't necessarily have the same context. So they're going off highlevel learnings or outdated learnings or looking at things that are kind of I guess broader takeaways from a category rather understanding rather than understanding the nuance of a particular business and what they should really be selling and the unique selling points that they should be pushing in the advertising. The other things that are breaking is the tooling stack. So teams are having to jump between different tools and the context is being lost within those. So not having kind of a well set up and minimal stack in terms

[3:20] of the way that your advertising uh is scaled and structured and created and then a lot of it just comes down to the ways that the team is actually working. So a lack of having kind of uh structured and repeatable creative feedback sessions or whip sessions where someone in the team actually takes that responsibility for sharing back the learnings of what is happening in an ad account and the whole team can sit there look at the work that's performed and everyone can come to a conclusion on what they should make next based on what they're seeing perform in the ad accounts. And that is on all levels dissecting every part of the ad as part of that creative fatigue or that that creative um drill down. And what is the earliest signal that creative learning has stalled? Well, the first one would be lack of spend in the ad in the first place. So, it's really not kind of getting out of the adset and there's not enough in it in terms of creative diversity or it's just not meeting the target market that that ad was designed for and therefore it's not getting out the gate. But other signs that learning has stalled or the ad is starting to fatigue is the primary metrics um and things like your CPM just start rapidly rising or the frequency is

[4:32] rising on that ad itself. Creative used to be abundant now it is the constraint. Why has creative become the bottleneck instead of spend or distribution? Yeah, I I mean I think AI is a big part of that. the cost of of creating assets is getting closer and closer to zero. And now anyone on the team with a series of prompts can get to an output, whether that's a static or a video ad. So creative is no longer the bottleneck from a production standpoint. It's more so deciding what to make um and where to take the direction of your messaging and your ad account. And then the other bottlenecks that are starting to kind of form around that is being able to get that creative that you've generated um or put together into your ad account in a a low friction way and then have a single place where those learnings come back to that the team can check. So in a world where it's easier than ever to make that advertising, knowing exactly what to make is now the primary driver that is defining the the greatest teams out there. What happens inside teams when iteration slows but pressure to scale increases?

[5:38] I'd say it mostly looks like more money being spent on bad ads in the sense that you have an ad account that's not hugely efficient. You have a lack of velocity or creative diversity with inside the ad account. And a lot of the time that is driven by teams wanting to make advertising that looks on brand. But that lack of diversity in the formats or the layouts or messaging that is one of the things that can really start to kind of slow down or drive an ad account to be inefficient. So that is I guess some of the things that you see when a team or an account themsel an ad account itself starts to kind of degrade. How does this bottleneck quietly cap growth long before CAC spikes? Yeah. So I guess the knock-on effect of having a lack of creative diversity, not having messaging that's tested and taking from your previous learnings, not having a larger roster of creative talent or different messaging angles to test. This leads to having a team that has a smaller number of creative in the account where there's no diversity. It's not really tested on anything. That creative is becoming stale. So you'll

[6:46] start to see those signs of fatigue or the inefficient creative and that will obviously compound into an increase in CAC. Other things that you see with smaller teams is like uh incorrect setup of their tracking. So pixels not installed correctly. They're not using the conversions API set up well. So they're actually sending bad data back to the ad platforms. And then that is also a big component of of what is leading to an increased cost of acquisition is they're just not only working with maybe poor advertising but poor data as well that's coming back. So an inability to create that feedback loop on multiple fronts. I want to talk a little bit about Canva grow and not as a tool but more so as a system. Why does splitting inspiration, creation, publishing and insight across tools destroy learning?

[7:31] Yeah, good question. So every time you have to send data back um through platforms or through people, you're going to lose a bit of the context in terms of whether that is the assets that you have on hand like the footage and B-roll and the underlying creative that you're using within your advertising. The more you move that around, you're losing the kind of structure organization in terms of what assets have worked and which pieces you should use. Again, you're also starting to lose that understanding or that kind of single messaging matrix of the ways that you're portraying your brands and what is actually working there. And then the performance data itself is also kind of getting lost between those tools as well as the inspiration. So, if you've saved kind of competitor ads or ads that you found interesting or you want to take from, that all starts to become a real mess and it's a mixture of different spreadsheets or Google Drive links. Um, and it really is all over the place. So it overwhelms a small team and then even for larger teams it's just an unnecessary point of friction. So one of the big things that we're doing with inside Canva grow is bringing in all of those different parts of the creative workflow into a single tool and Canvas in such a strong position to do that

[8:41] because we already hold all of the brand context with things like brand kits and then we also have the context of someone's products that they're selling, previous designs they've created. So we really do have a source of truth and now we're bringing in the performance data and we currently support Meta and Tik Tok and Canva grow but we're adding more platforms in the coming months. So all of this data sits with inside one place and that feedback loop can actually thrive inside Canva Grow cuz you've got everything there right from um create which is our pillar for generating assets. So you give the context of the product or service you're looking to create an ad for. We generate you an ad on the fly based on all of those learnings and all of that context we have about your brand. We've then got the launch pillar which can deploy an ad to any platform via Canva Grow. That whole experience is handled with inside the product. And then we've got the learn tab which is ingesting the data back from the ad platforms and giving the Canva user great context on on what is working for that particular ad but also giving them a view of their entire ad account or marketing function. So you can see there between those three

[9:50] pillars having them all sit together where that handover is seamless. Once I generate an ad I can quickly publish it once it's published I can come back to see the data. Then when I create another ad the following week, that performance data is actually sent back to the create flow. So you actually have this spinning flywheel, but it's all existing with inside the one product and the one ecosystem. So how do how does this closed loop shift team behavior and not just output? Yeah. So one of the main things is is being able to have a team think more about the the wider more holistic marketing strategy and the messaging and the way that they want to portray their brand. And it gives them more time to refine things like the particular brand identity or the craft in the ads because we're getting them from zero to one so much faster out of that creation flow.

[10:40] And then we're cutting down the amount of time that they're having to spend on laborious tasks like uploading ads to Meta's ads manager or even having to navigate a series of spreadsheets to find out what has worked or try and infer what they should do next week. So creative teams can be much more focused on how they want to sell or shape their brand versus a series of kind of small tasks that are messy and unfulfilling. AI can act as a multiplier but not a savior. Tell me a little bit about how AI outperforms humans when it comes to creative workflows today. Yeah, I'd say what we're seeing in terms of where AI is really performing, its ability to do things like research, look at a wider category and pull learnings in terms of what is being spoken about or customer reviews and getting down to kind of granular insights about the types of things that are making your customers buy or have some kind of emotional reaction to your brand or product and pulling those out and summarizing them as a place to work from. It also does a really good job at generating messaging or things like ad copy where it can take

[11:47] a set of context. Take an ad as an example and then generate all the subsequent copy that will sit alongside the ad. It can also do a really good job of writing things like articles or more kind of text content focused outputs. Where we're seeing it, I guess have rapid improvement is on the generation of of static ads. We're coming a long way there. Canvas got some awesome technology on our side as well when it comes to being able to work with AI to create static designs. There's a new Canva product that came out a couple of weeks ago called Magic Layers that can actually break down a static output into a a Canva design. So you can edit all of the layers and that gives teams an amazing amount of flexibility to take the best-in-class when it comes to generating a static ad but then take that into Canva and actually refine it and change the copy the messaging or make tweaks and modifications to it. So where we're really seeing kind of teams excel or AI excel is where teams can use certain tools that will take the best of what AI has created but still let the user have the kind of last mile in terms

[12:56] of their ability to make tweaks and really kind of polish something or make it the best fit that it possibly can for that product or business. And why does AI accelerate bad positioning faster than it fixes it? Yeah, good question. I would say it's almost that feedback loop that you would get into. For one, if you're sending bad data to the LLM or the ad platform and the way that kind of targeting or that generation is being done, if you're giving it poor context on what's working or where you are in the landscape or where you want to be, then you'd see it kind of double down on that messaging or that angle versus having kind of the ability to step back and look at kind of the the larger ad account or marketing function or or brand positioning and actually make kind of larger decisions. about where you're actually sitting and what are the things that you can do to improve your marketing on business. And I guess you get to more of that kind of echo chamber if you're going back and forth between an LLM and you're just doubling down further and further on something that's not necessarily the best path forwards. Without without modernday infrastructure and systems into identifying what colors

[14:06] or captions or copy correlate to an increase in click-through or a decrease in click-through. uh teams aren't able to identify what is working best across audiences. Can you tell us more about how this is effective? I think I answered the question for you. Yeah, I I I think I get I think I get the um shape of the question. It's what one of the things that we brought over from Magic Brief were the concept of creative recommendations. So at first we were giving teams really clear like visualization of what is working within their ad account and we would actually score an ad on multiple fronts to tell them is the hook working, is the call to action working, giving them kind of a complete view on what they could do to improve that ad. And that worked really well for some teams and some roles like the creative strategy role for example or the media buyers would get a lot of value out of that. But we wanted to take it a step further and this is something that we've carried through to Canva Grow is we will actually generate plain text recommendations. So, it's almost like

[15:13] you've had a marketer sit with you and actually write out what you could do to improve a particular ad all the way through to that actually becoming a brief that gives the uh the marketer the best possible context or instruction of how they could make that ad better or how they could create a net new variant of that ad that takes on all the learnings of the ad that's already had the spend and gone through the learning period. And that was a huge success at Magic Brief and it continues to be today in Canva Grow of actually giving um our users and marketers or small business owners that complete context on what they could do to make an ad or their entire marketing function better, but written out in a way that's super easy to understand. So, you're not having to go and look at a table of data. You're seeing a sentence that gives you kind of complete context on where to go next.

[16:03] The super exciting thing about the way that we're building inside Canva Grow is we can actually take those learnings and take those recommendations and you'll be able to hit apply. So we'll take the instruction that we've given a user and apply that into the next uh creation flow. So, if we've said, "Hey, you should try a headline that has different messaging or a different content creator or a different editing style," we'll actually be able to carry that context and that recommendations right through um to create the next ad on your behalf. The other thing that we're looking at at the moment is how to do more proactive generation. So you could come into your Canva grow dashboard and see the ads that would be the best possible ads that you could create at that moment in time based on all of the context of your ad account uh or your business that's connected to Canva. I love that. And is does that dynamically update and make suggestions based on um I don't know weekly cadence or even a daily kind of update around what's going Yeah, it updates based on the cadence that's coming from the ad account. So as

[17:10] new ads go live and that data comes back from Meta or Tik Tok or various other ad platforms, we then ingest that, understand the asset, come to a conclusion on what could be done to improve it and that's when we generate those recommendations and show them in the UI. We're also working on ways of giving a owner of a small business or a marketer at a large organization context about their ad account so that they can come into Canva grow and we will have kind of a single summary a single pane of glass that gives them complete visibility on these are the metrics that are underperforming these are the ads that are winning these are the ads that you should double down on and this is what we think you should do next week and we really want to get to a point where it's essentially like an AI I CMO that can cover all channels. So giving context on a new subject line that a user could use based on what is working with inside their ad account, recommending new channels that they could explore or scale and essentially working alongside um a marketer or someone who's relatively new to marketing and is just looking to scale their business. And that's when we want

[18:18] to move from uh I guess prompts or a way of working that is maybe fixated on a particular end output and work more on the the actual intent of the thing that the person wants to do. So, if you own a beauty brand and you've launched a new product, helping you scale efficiently by going to your existing customers that we know would be in market for the new product that's about to go live and then working on a retargeting campaign for those particular customers versus spending money to acquire customers that you already have on the mailing list. And those things, while they might seem more obvious for a more sophisticated marketer or an agency, for an SMB, there's work that needs to go into making that accessible and making it easy to understand and set up things like that retargeting campaign. And we think everyone should have the access to do that. And we want to build the tools with inside Canva and Canva grow to make that possible so every team as empowered as they possibly can be to scale their marketing and their business. Yeah, it sounds like you're commoditizing the performance marketing function and not

[19:26] just one specific specialization, but also um you know the design component as much as the execution component and then the data analytics component and using that to iterate and and create a better better version of the ads. That's um that's quite an offering and making this accessible not only to the marketing function now but to any small mid to medium-siz business any founder operator CEO that's looking to kind of put their performance advertising on autopilot or push have minimal effort and be able to continuously optimize their ads which will increase the likelihood of success. Tell me more about what kind of a price point does this come at?

[20:19] Yeah, good question. In terms of pricing, the current Canva Grow product is available to almost every Canva user and then credits are consumed, AI credits are consumed when you generate ads. So, that's the kind of primary means of how Grow is monetized. But that's still something that um we're working through because Grow is a relatively new product with inside the Canva ecosystem and has only been live for about four to five months now. So we're still focusing mostly on how we get product market fit with Grow as it continues to expand with inside the Canva ecosystem. Got it. And then um with that said, great creative still requires judgment. I can imagine you can tag the images and identify what is correlating to an increase in clickth through and what is correlating to a decrease? That could be the colors used. That could be the different images used. What performance metrics lie most often in these creative decisions? Yeah, I'd say well firstly a lot of

[21:25] performance metrics can lie if you've got your tracking set up wrong. So, that's kind of a key one to watch out for. But one of the metrics that I would say trips teams up most is the click-through rate. Because click-through rate can be great for seeing that there's like people that are curious and there's kind of inbound curiosity for the the product that you're selling. Or it can mean that the ad has done an amazing job of capturing someone's attention and obviously driving them to click, but it doesn't always mean that they're the right buyer or right persona for the thing that you're selling. And unless the downstream metrics from the click-through rate are also performing well, that can mean that you're sending good traffic to the wrong product page or landing page, or the offer isn't actually compelling enough to drive them through to buy. So, I think some teams maybe get carried away with click-through rate, viewing that to be the kind of driver of creative success, unless the downstream metrics are kind of all adding up um behind that, that can be a dangerous path to track down in the sense that you're um intriguing people but not necessarily getting all the way through to the sale if you're

[22:34] just looking at click-through rate or focusing on it. And what insights actually correlate with repeatable wins? Yeah, that's a good question and that's one of the things that we've been trying to pull out um with the business that was built with Magic Brief and now what we're building at Canva grow is helping teams like get to those correlations of what are the things that I spoke about, what was the creator that I showed, what was the messaging angle or format, what was the hook that worked when it came to driving success of this outlier piece of creative. I mean, part of the issue is it's relatively hard now to come to exact terms about kind of the pieces that came together to make an ad win. And that is also on the platform side as well. There's things that are in no way related to the underlying creative asset. Say for example, one of the ads that you run that's running in your ad account gets positive engagement from a user um on the Instagram timeline and then that ad will actually get pushed to the front of the ad set because the algorithm will start to see early positive engagement

[23:44] from someone who looks like a customer. That will push the ad all the way through to the point where it's the largest spender. That's the one that starts driving the creative learnings and starts achieving outcomes for the brand. So, not necessarily was that anything that went into the creative itself, but something that happened when it hit the timeline, the Tro Ford forward. So, I guess that's one reason why it's not super predictable in terms of being able to guess the next winner. And a lot of great marketers, if you were to show them 10 Facebook ads and try and have them pick which one was going to be the big spender, it's a relatively hard thing to do. So we want to build the toolings and help teams of all sizes get to a place where they have a system that acts more like a product function in the way that they can work. So they can test and learn in these almost structured sprints of getting more work than ever that resembles their brand, speaks to the true selling points that they know already work and do the best possible job of portraying their product to the audience that they know they have. and then running that that flywheel on repeat so that they can get to a point where they achieve those creative outliers that really do move the brand or the business forwards. So

[24:56] setting up I guess the scaffolding to make that test and learn possible so that they're not having to try and reverse engineer the science of kind of how they how they are getting to that great outcome. And how do teams let data lead without turning marketing into spreadsheet theater? Yeah, I'd say ways of doing that are still focusing your creative audits or your creative crit sessions on the underlying creative. So, not getting too carried away zooming out to a point where you are only looking at spreadsheets or ads manager consoles and you're still looking at the underlying video assets or statics that are sitting with inside your ad account. looking at those and trying to draw patterns between what was it in these three ads that really stand out that have driven performance. Still holding on to things like creative diversity as I mentioned before to be such kind of a powerful way of being able to um create a thriving set of things like creative graduation really help there. So for example at eucalyptus where I was creative director we had a fairly sophisticated way of handling creative graduation. So if we

[26:04] had a static ad that we had launched and tested and it did well, we would take that static ad and turn it into a video. If the video continued to perform well, we would take that video and turn it into a YouTube ad. If that YouTube ad did well, we would take that ad and then turn it into a TV. So a method of being able to look at what is working in an account and graduate it all the way through. So that's a way of being able to um use performance data to build on the underlying creative that you have without getting kind of too carried away at purely looking at the creative metrics. And I would say it's also thinking about how you structure your team so that you can always have people who are across the context of what is working but it's more of their skill set or it's more of their remmit to focus on how that comes together in terms of creative concepts. So being able to have a delicate but powerful balance of your media buyers but then your creatives or your copywriters or your video editors on the other side of that and having them work in harmony. So there is a um role or capability that's focused on the

[27:11] numbers but then you balance that by having someone who is purely focused on the creative craft and production side. Now how does Canva grow enable speed without eroding brand consistency? One of the major ways that we do that is we plug into the Canva brand kits. So there's tens of millions of brand kits that exist today for businesses all over the world that they've created. And that carries through context like the tone of voice, the hex code, colors, typography, and everything that kind of makes a brand I guess the brand that it is and how it shows up. And so when you generate an ad with inside the the Canva grow function, it's taking and pulling that context from your brand kit. So, it's on brand every time and it's the right colors, the right typography, the right branding, and the right product images have all been pulled through. And that I guess is an issue that many teams have today if they're using more standard creative generation tools is the outputs that they're getting are more all over the show and there's a lack of consistency. It doesn't look like their own brand or their own

[28:18] assets. Um, and every time they go to tweak that ad or generate another version, it looks like a completely different asset, which is one of the strong points as well with Magic Grow, oh, sorry, Magic Layers, as I mentioned before, where you can actually take an asset that's yours, even if it's a static, have it broken down into layers and rework that into another variant. So, you're still having that underlying control of your own asset that's already on brand, and just being able to tweak it or nudge it forwards in a way that's going to be performant. I love that. Yeah. um that holistic approach is going to really optimize the workflow and increase the likelihood of a of a of an ad that resonates with an audience specific to the the brand um for that campaign. So, where do teams overoptimize and lose narrative coherence? Yeah, I'd say teams overoptimize by maybe getting too into the weeds of granular performance metrics. um around like a series of static ads versus being able to zoom out and look at like what is working from a messaging standpoint.

[29:25] What is working across channels? What are the the audiences that we're really resonating with and being able to build on that versus getting too into the weeds of a single channel or a single ad set or a single piece of creative where that work is being done on maybe too much of a granular level and they're losing that ability to step back and look at their entire marketing function as a whole is where I would say teams go wrong or they get too into the weeds um with just purely one tra source. and they're not balancing that out to say, well, what is working from a messaging standpoint on the the Google search ads that we're running and then how does that relate back to like the subject lines that people are clicking in emails and having that more kind of widespread view of how their brand is showing up and where and how it's performing as well. And I think that is one of the powerful roles that you get from more sophisticated marketers, being able to have that cross channel visibility and sharing those learnings, but not something that necessarily a small business owner would be able to carry through. Um, that's another big

[30:32] thing that we want to figure out how that we can best enable with inside growth so that the metrics that we give someone are relevant but not to the degree where we're starting to distract them or confuse them or getting them to silo in on a particular campaign or a particular ad and focusing more on an ad account or a marketing function in its entirety in terms of how that can be improved. And what would you say fundamentally changes in GTM execution as teams grow? Say your ability to build systems becomes more and more of necessity. If you're starting out as a solo entrepreneur or the single person working in marketing for a startup, you don't have to rely on systems in the same way because you're the one creating the video ads. You're the one uploading them to the ad platform. You're the one looking at the ad platform data and deciding what to make next. But as soon as you start to scale that function, you need that consolidated view on like the messaging matrix that's covering off each of your brand or products, you need that kind of learning hub to store everything as I mentioned before around storing all of your content and assets. It's like as you start to scale your

[31:41] team or your go to market function, if you don't get a structured way of being able to produce creative, ship it to platform and store the learnings, you really start to end up in a bit of a mess. And that either costs you in terms of team time or resource or it costs you because you're just not putting out performant creative in the way that you could be if you had a better approach in the way that you set up or build a system. Yeah, I'd say for one, a data standpoint, they set up maybe heavy tooling too soon in terms of the way that they're pulling data back from the ad accounts and they jump straight from something like using ads manager to sending data to something like Snowflake or just go kind of too overboard. And then you get to a point where those creative learnings aren't democratized in the same way because you went so far into kind of setting up um tooling or a complex workflow that you need like a data scientist or a a media buyer to be across that data in order to pull out those learnings and send it back to anyone in the team. That was one of the

[32:47] big things that we focused on when building magic brief is building a creative tool and a creative analytics product that anyone in the team could access and anyone in the team could understand. So making it really easy to query what is working and break that down by different campaigns or products that you're selling or different seasons that you you ran that creative asset that series of creative assets and being able to drill down with things like AI tags that exist today with inside Canva Grow. So, we tag all of your ads with different um labels like the type of content, whether it was UGC, whether it was shot in a studio, was a seasonal campaign or an evergreen campaign. And that makes it really easy, as an example, for anyone in the team to be able to drill down into that creative and find out what works. And you see teams on the other side of that that maybe go too far in terms of the the tooling in their stack of building a workflow or putting together tools that are like enterprise level. But that means that so many people that are actually working in their marketing

[33:53] function can't actually access the data um in a in a reasonable way without getting someone else to kind of run that query on their behalf. which goes back to that wider issue of building unnecessary friction which slows down creative velocity because they no longer have that flywheel loop running because they went too heavy in terms of setting up tooling. The other thing that I've seen um in terms of maybe over complicating the creative workflow or your tooling stack is going overboard too early in terms of the content management system. So setting up um digital asset managers that are like enterprise versions when you could have had something much more streamlined and just set up a better system with folders and tagging. So getting that balance right so that you can find your creative learnings or your creative assets in a way that is easy and anyone in the team can do it alongside you versus going to the full extreme. That might be the best set of tools if you're a 200 person

[34:59] marketing team, but not necessarily the best set of tools for a 5 to 10 to 20 person team. What can enterprise teams learn from how small teams ship and learn? Sorry, can you repeat the question? I just cut out. What can enterprise teams learn from how small teams ship and learn? I'd say one of the things that enterprises can look at as a way of learning from the smaller teams is being more agile in terms of how teams are put together and the way that like back to that example if you are a solo entrepreneur or founder running an e-commerce business or if you're the single marketer at a startup the ability to have that single single person who's working on the creative uploading it and drawing from the feedback and being able to have everything contained within the same set context in the same person. Although it's not necessarily scalable, the ability to have all of those learnings even subconsciously coming through in the next set of creative that is being developed for the brand that is working because all of the context is obviously sitting there with the same person. So the thing if you're running a marketing function for an enterprise would be how

[36:07] can we get closer to that but still have a scaled solution. So whether that's breaking your team down more so into squads where you've got dedicated teams for each of maybe the brands that you're looking after if you're a house of brands or you're a large agency or you've got particular squads that you've put together for different objectives um with inside the business but not overwhelming the underlying marketing function or creative team in terms of them having to hold too much context about a wealth of different objectives or products that are all running and being live at the same time. So being able to have smaller, more focused teams that all have all of the context in one place and I would say that's becoming only more important um with AI in terms of those smaller teams can get more done and they can automate work or put together workflows where that small set of team that holds the context can go and have other pieces of that creative workflow done on their behalf. So that's an even more advantageous place to be versus having a wider team that's kind of more distributed across different objectives in the business and different

[37:14] campaigns that are being worked on and therefore losing that ability to just focus in on one objective um and fueling the marketing and the creative assets that need to go into making that perform. Are GTM tools converging into fewer, more opinionated systems? Yeah, I'd say that's what I've seen in the market is people are are consolidating different parts of that creative workflow. As I've mentioned, Canva Grow being a good example of that, bringing all those pillars that would have existed um as separate parts of the workflow or completely separate tools and bringing that all under the same um and you're seeing that happen more and more across the board where people are realizing that the benefit of being able to have that complete feedback loop. So giving off context or assets and having those live in different systems or different vendors or providers is is definitely less advantageous than consolidating into a kind of single stack or a single tool where it can be made possible. I guess the other thing with SMBs as well, not only do you benefit from having all of that context in one place, you can also save massively on costs if you can consolidate things that would have been different tools, different

[38:24] subscriptions, and have them all sit under the same hood. And what does the performance marketing stack look like 3 years from now? Yeah, I'd say it will continue to consolidate. It will look more and more like a system of record where the winner really will be the place where your assets live, all of your um audience data and your CDP, all of your performance data, everything that you have that signals what makes your marketing work, all lives with inside one platform. and then uh the use of LLMs, the use of um generative models all benefit from the context of where that lives. But all of the orchestration piece as well can be done by that same platform because it doesn't matter if you have an amazing ability to create thousands of assets. If those aren't going live or they're going live to the wrong audience with the wrong targeting, um that's a failure as well. So, a system that can understand and have the context of what works for your brand, but can equally help you put that work live and the channels where it perform best and have that overarching context

[39:32] about the channels that are performing and where to allocate your spend or which channels to try and activate next. But that will be done increasingly so by a fewer number of tools that have more access and more data feeding into them. And with that said, uh, where will founders still need to design custom GTM loops? Where will founders still need to design custom GTM loops? I would say the custom loops will make sense where there is particular nuance around someone's go to market or the audience that they have or the intrinsic knowledge that someone has about the way that their product functions and where they can create those viral growth loops where they can find alpha in the way that their product works in terms of the onboarding experience or being able to drive retention by finding a novel way to kind of interact with with a particular user and doing things based on the context of their own product and knowing the ins and outs of how it works and how to bring users into it and help them understand it and find value. And then

[40:38] also looking for things or looking for things with inside the industry that are like more novel and less explored that still represent some kind of alpha that aren't the kind of classic go-tos that would have come from like an LLM recommendation or kind of an industry course where they can go out there and try something novel and interesting and things that rely more on the founder themsself in terms of whether that's content that's being put together, whether it's like building on like human relationships that have been made and how you can kind of build um a go to market function around your ability to scale that. But things that are maybe less obvious and rely more on an understanding of your particular domain, your industry, your product or the the real people that you sell to.

[41:27] Moving on to the rapid fire section of the pod. In one sentence, first instinct, George, what metric lies most often in performance marketing? I'd say click-through rate. Clickthrough rate is a dangerous one in the sense that it a lot of the time can fool someone into thinking that's a high performing ad, but it can just be an interesting ad where someone or many people have been hooked by the piece of content and it's driven a level of curiosity that has pushed them to the landing page or the product page. But if you are not seeing that click-through rate accompanied with the downstream metrics like whether it's ad to card or booking leads, then it can't tell the complete story. So I'd say clickthrough rates definitely uh the dangerous run that potentially lies in that in that sense. All right, moving on to the rapid fire section of the pod. George, in one sentence first instinct, what metric lies most often in performance marketing? Clickthrough, right? What is the earliest sign creative learning has stalled? Uh, a lack of spend going to those ads.

[42:30] Static generation with unedit uneditable outputs. What mistake early teams make when scaling ads? Not enough creative diversity. And what belief about performance marketing is simply outdated? A belief that creative and media um are separate disciplines. It's the same system just measured at different time scales. For years, GTM rewarded volume. More ads, more tests, more spend. The next era rewards learning speed. Growth does not stall because teams stop creating. It stalls because insights stop compounding. George, thank you for showing what modern GTM looks like when creativity, data, and systems finally operate as one. This is GTM vault. Build loops. Non-noise.