Introducing Compass: The AI-Powered System for Smarter Ads

[00:00:00] Richard Gaffin: Hey folks, welcome to the Ecommerce Playbook Podcast. I'm your host, Richard Gaffin, Director of Digital Product Strategy here at Common Thread Collective. And I'm joined as I always am by Taylor Holiday, our CEO here at CTC. Taylor, what's going on, man?

[00:00:14] Taylor Holiday: We both have our logo wear on from our team's big weekend. So congratulations to you and the Philadelphia Eagles.

I know they're big in Portland, so it's good. I'm glad. I'm glad that you feel that deep connection.

[00:00:27] Richard Gaffin: That's right. I sure do.

[00:00:29] Taylor Holiday: But for me, I also suffered a big loss this weekend. It wasn't the Kansas City Chiefs.

I'm not a Chiefs fan, although I was sad to see the Peshawar Homes not get the three peat, but my Roboducks, my son's robotics team finished 30 out of 30 second teams this weekend in our Robo showdown against the other schools of the area. So tough, tough weekend for the Roboducks, but

Congrats on the Eagles victory.

[00:00:51] Richard Gaffin: Yeah, no, thanks. So on the you know, robo showdown, what are these battle bots or what 

[00:00:57] Taylor Holiday: Yeah, well, so there's a, it's really quite interesting. There's a A stand with four holes in it and you have like two minutes to get your robot to pick up these balls and make as many of them into these little containers as possible. And so robotics is new to my kid's school this year. So it's their first season.

And so your, your robot is like an asset of the school. Like you can keep

building it over time. And so some of these kids at schools have been building these robots for years. And these things are incredible the way

that they pick up these balls and dunk them into these things. And We'll just say, Mike, the RoboDucks have some, they have some development to do.

They have some work to do, 

but it was it was still fascinating to see. It's a whole subculture. This thing to bounce between like a basketball game into the robotics showdown is like quite a culture shock in

a different, different, different world.

[00:01:49] Richard Gaffin: Yeah. I can imagine. Well, a rebuilding season for the Roboducks maybe this year. I don't know. Cool. Well, speaking of tech, hey, there's a segue for you. We're today what we want to talk about is something that we've kind of been teasing on the pod a little bit over the last few weeks, which is what we're calling Compass, which is essentially our new creative operating system here at Comethread.

And what we want to walk you through is exactly what it is and kind of how it works. And maybe specifically some of the. Tactical or strategic underpinnings that kind of led us to create this thing. So let's kind of start at the beginning, just describing for the folks, what this is and why we built it.

[00:02:26] Taylor Holiday: Man, this, this goes back so, so many years creative systems have been probably the bane of my existence as a agency CEO for, for over a decade. It's been trying to solve for designing a consistent. repeatable appropriately resourced and effective creative system for advertising. It's a really challenging task to do. And in addition to that layer on the evolution of our industry towards a emphasis on creative volume as the lever that meta is pushing further and further into demanding for Driving efficacy and the research and data that we've come to understand that that is really the lever makes the most impact. And it's become a really big problem to actually be able to design a system to produce that much volume consistently, and even trying to answer the question is how much, right? So there's all these different pieces of the evolution of our space. That's led to. Trying to think about how to do this from a data driven thoughtful way that allows us to understand the task at hand and deliver against it in the time constraints that we have at a financially viable way.

Like, that's like at the simplest level, that's what a brand has to do, is every month determine how many things to make. What things to focus on and then to build a system for producing it consistently and effectively that delivers the results that it cares about. And that is a monumental task to accomplish.

[00:03:51] Richard Gaffin: Yeah. So let's, let's backtrack then to talk a little bit about, because I, remember this personally, when I was sort of in the creative department, creative strategist, all the type, all that type of thing, like the extent to which it was the bane of your existence. And so maybe, maybe highlight some of the ways that, that this is a uniquely difficult problem.

[00:04:08] Taylor Holiday: Yeah. So CTC, I don't know how many people know our journey of creative, but we have sort of existed at every portion of the creative process at different times. We did full scale video production for TV commercials and things like that to having a. 10, 000 square foot studio inside of our office, where we would be doing photo shoots every single day, photo and video production to doing branding.

As at the very beginning stages of a business's creative journey to to UGC creators, to an influencer marketing team inside of CTC for product seeding and content development. Like any way that you could imagine trying to produce creative, we've participated in it at various points. And it's always been.

To operationalize a consistent set of ideology. And Richard, this probably goes all the way back to you and I building our advertising philosophy class together as an initial starting point of saying, okay, hold on. We have all these ideas about how to do this thing we're trying to do. And our job is to take. Build a coherent ideology that's consistent and then to figure out how to operationalize it across everybody in our organization so that we're all repeatedly producing high quality work. And that's just, that's a, that's a hard task in every discipline, but in creative in particular, it's incredibly challenging to be able to do that.

[00:05:28] Richard Gaffin: And I recall like a lot of the issue there being that there's something about producing It's difficult to actually operationalize the creative process itself. Maybe like ultimately you have to have some sort of idea that then translates and speaks to another person. And as a creative endeavor, it's difficult to actually kind of break that into its component parts.

So the issue was always, how do you include something that's so kind of soft into an actual more mechanical system that can actually produce something in an expected timeframe, which I think was like the big. Okay. The big issue there,

particularly like something along the lines of like taking a gamble on a piece of creative, ultimately like is, is a gamble.

That doesn't make sense if you spend too much time and resources on it and it inevitably fails.

So how does this kind of begin to evolve us towards solving for that kind of 

[00:06:17] Taylor Holiday: So, I have sort of really tried to. Think about the parts of the creative process that can be defined, that can be bound into specific requirements and outputs and estimations and which parts of it cannot and where you need to create flexibility and where you need to create rigidity. And Richard, you've been a sort of big inspiration in thinking through that.

I think that ultimately, What I have found is that the, the key to great creative process is actually constraints is actually to appropriately define the constraints that the process is trying to work within open ended creative workflows, where you just sort of say, we're trying to create an unknown number of ads to create a really open set of purpose, like be better is almost impossible to actually systematize.

So the first thing I think I really wanted to try and understand is. Is there a way to define what we're actually trying to accomplish? So one of the things that I felt that was sort of really weird is that when we would do these creative scopes for customers one, there was no connection to like the financial system.

So like when you would define how many ads we're going to make from you for you, it always sort of felt arbitrary. We're going to make 20, we're going to make 40, we're going to make 60. And it was just sort of this like random number relative to your budget. And then what that translated to in terms of. Add costs on my side in production. And there was no actual question of like, what's the task to solve for, for the month? How many ads does that require? When do they need to go live? And then how do you back into a system from there? And so the, the first thing that felt very odd to me is that defining the scope was ambiguous and lacked sophisticated data driven thinking.

And I felt like

we could forget deciding that what is the ad going to be yet. I felt like where data could really help us is to define the scope of the job to be done. And that's where I think the first step of our creative system leaned in was to say, we have all the pieces to actually answer this.

And there is a data driven way to answer this. So at the very least, we know what is the job that we're trying to do.

[00:08:17] Richard Gaffin: So, so the first step then to systematizing it is well, like you're saying, understanding the constraints, but then I think like the first, it felt to me anyway, like the first step towards that was, was actually setting some definitions and one of those being, for instance, and we've talked about this probably for a year or so defining what an ad is versus let's say an iteration or an asset or.

A campaign or whatever the case may be, because even like in contract language, you would throw in words like we'll deliver 15 ads, but that could mean a number of different things to a number of different people. So once the definitions are set, then all of a sudden the restraint exists to work with them, it's kind 

[00:08:55] Taylor Holiday: Yeah. So the definition is certainly positive, but I even see like, there's this big debate going on in Twitter right now about agencies and SaaS providers, like using brands. Using brands, logos and various points and whatnot. But and so Cody put out this tweet of like, these things are all borderline fraud.

And one of them was like calling 50 iterations, 50 ads. And I just go like, Oh man, this is like, what's so frustrating is that nobody's actually trying that. Like the idea that that's like a. Attempt at fraud is just, it's just a lack of language clarity. And so this is the thing that came up all the time is that what is it an ad?

Is it a unique ad ID? Is it a completely different visual? Is it a change to the headline? Does that change? Like, so you actually have to work hard to create shared meaning on these words. Cause there is no objectively true definition of the word ad, right? So you have to create a system for defining the terms within the system.

That's hard. That requires effort. And then from there, you can begin to model the relationships between The number of ads and the outcomes that you care about. So this is like the number one problem that I see in everybody's creative system, every one of you cannot say for sure how many ads you need to make next month to accomplish your financial outcome. And that is like, that is at the heart of the underlying problem is how many things am I trying to make? And what I see people do is that they, they sort of invert the process where they go like, well, I have two designers, they have 10 available hours, therefore they can make 15 ads. So we make 15 ads and they work from like. Resource to available capacity versus requirement to resourcing. And this is like at the heart of the creative process problem that I see with every client we have is that they come to us with some creative system that is insufficient for the actual job to be done with how many ads they actually need to produce, and they have no viable mechanism for getting there. Above all else, what I felt really confident in. Is that this is math? This part of it is math. We can model the relationship between your desired budget and the number of ads that are likely to accomplish that spend volume. This is like you can solve this problem with math as a starting point for defining the boundary of the system. And so that was the first thing I really wanted to do was just like, get clear on one helping brands to educate them towards this research that we found that was basically like most ads don't work. This is a hits business. The vast majority of your ads will fail. You'll have a very small hit rate and you can model the expected spend per ad. Like these are things that mathematically can be done. And that allows us to set up then in partnership with the rest of the CDC statless system, where we have clear financial planning connected to the marketing calendar that produces a monthly media budget. You pair that with the modeled relationship between spend an ad, and you can actually get to an answer of how many ads do I need to make next month to with a 95 percent confidence, be able to spend this budget. That is a starting point for now a project that you can execute and resource appropriately.

[00:12:00] Richard Gaffin: So that's kind of like, let's say that maybe the first part of what Compass does is it's, it's a sort of a, what would you call it a program or whatever built into statless that. Pairs essentially like your budget for the upcoming month or whatever with, and then kind of spits out the number of ads that you're going to need.

So that's kind of like the basic restraint that it provides. But of course, with the advent of AI, there's also an element to which Compass sort of blurs the lines between math and art by providing other sorts of insights. So let's, let's talk a little bit about the other kind of things that Compass provides beyond just 

[00:12:32] Taylor Holiday: So step one, Compass combines the financial modeling that we do in our planning process to get you to a media budget per month. It models the relationship between spend and add over time to get to an expected number of ads needed every month. So it sets up the framework. Then once we know how many ads we need to make, we now have to ask a series of other questions, which is one is what ads or what are we selling? Okay. This is sort of step two. So if we think about this offer audience angle framework, this next step is offer. Okay. And what are you selling? And this question I think has like four steps to answering it. Number one begins with the marketing calendar. So if you think about in our planning process, we're going to define the marketing calendar, which includes really important things like go live dates for product releases. So if I'm going into February and on February 22nd, I'm releasing a new product, there is some amount of the remaining creative work to be done. So let's say I, my model tells me I need to make 500 ads. There are currently 200 live in my ad account. I've got a Delta of 300 ads that are the February job to be done. Well, when I begin to break apart that 300 into what ads am I going to make? I begin with the things I have to do. So the product release comes first. And the product release, we can actually get to a question of how many ads do we need for this specific product release. And this is another area where I see people do all the most random crap in the world. They'll have a tiny like inventory run of a color skew and send me 74 ads. And then they'll have a big giant release with a massive inventory position and send me seven ads. And it's like, we've got to bring these things into relationship. So what, what I, what, what this allows us to do is we can sort of tier rank the, the product release relative to the inventory position. And then based on the number of ads, determine some budget that determines a certain number of ads, again, using that spend per ad relationship to get me to, okay, this product release is a 24 ad product release. Okay, cool. So now I know there's 24 of this offer. Okay, cool. That's step one. Get your product releases scheduled, the things that have to be done. Okay. Step two then is to look at the best sellers that are underrepresented in your ad account current. Okay. So of the products that you anticipate, you, you're going to do the highest volume on. And we, we use an analysis that we call the product sensitivity analysis, where we look at the, the volume of spend by product. Relative to the revenue by skew and, and to try to identify, Ooh, we're underspending on, I'll give you a specific example right now for one of our clients, their women's product line is doing really well. And so it's in the ad account. So it's taken up 80 percent of the spend in the ad account, but women's is only supposed to be 40 percent of the revenue. So that means that men's and in particular the men's t shirts is underserved in the ad account. So that would be phase two then is now we want to make sure that we're We're getting net new creative to support men's products and lift up that portion of the ad account. So now you're going to go through and you're going to sign again of these 300 ads.

I've got 24 of the new product launch. I want 15 to this new t shirts, 30 to the pants, whatever. And you're beginning to build out. Okay. Of those things, how many are going to each offer? And then So the second is like that sort of deficit on my best sellers. That's like the second thing we go after. The third category is like sort of the opposite.

It's like excess inventory. This is, this is things where we have, our inventory has aged out beyond the place we're comfortable. And we're actually probably looking to turn this back into cash more. More frequently. So there's this tier of things where it's like, Ooh, this might be a liquidation funnel.

This might be a lower Ross. This might be some way in which we're moving things at a discount, but we've got way too many of product X and it's sitting there and not moving and there's nothing in the ecosystem that's going to make it move. We need to move it. And then the last category is just iterations on your evergreen concepts of the things that are working.

It's not to say because women's is working. We stop. We look at each of those campaigns and go, okay, do any of them need refreshers of creative over the coming weeks? Okay. Okay, cool. And ultimately you're going to allocate all 300 of those remaining ads to some product category or offer. This could also include things like a site wide promotion or sale.

That's sort of in that first category of marketing calendar moments that you're going to support. And now we've, now we've designed the offers, everything that we're focused on selling for this next month. And that brings us now to the actual job of the creative strategist. So everything I've set up to this point is done by first. A data scientist step two by a growth strategist to actually design the offer work. And now the creative strategist can pick up a very clear set of constraints. I need to make 15 ads for this 32 ads for this 57 ads for this. And they all have a bounded expectation of the offer. And the timing of the category and how many, now you can ask the actual creative strategy question, which is related to who and why.

Right? So now we get into angle and audience design, and there's some work that the art tool does to help them speed up that process because 300 is a lot of ideas that come up.

[00:17:37] Richard Gaffin: That's right. So let's, let's kind of dive into that, right? So all of the kind of quantitative work has been done. And now there's some sort of qualitative work to be done as well. But this is something that the Compass tool also. Helps you with specifically through like AI learning models. So

let's talk about when the creative strategist comes to this, they understand exactly what they need.

They're just trying to answer these questions of, of why and to whom. Obviously, like what, what, sorry, what I should say is like what the tool helps them do is generate those ideas, right? They don't have to do it in their own heads. They can 

[00:18:09] Taylor Holiday: That's right. 

[00:18:09] Richard Gaffin: talk to me a little bit about how that works.

[00:18:11] Taylor Holiday: So what we do at the audience level is we analyze. We upload all of the available qualitative data from reviews from the business into an LLM and it categorizes those reviews into personas that we would use as the audience options for each customer. So I'm looking at bambooers audiences right now, the number one. 24 percent of the reviews or personas fall into a category of what we call the hydration devotee, devotee. These are like sort of very rich personas It actually generates an image of who the person is, and it has this description.

It says, a skincare enthusiast who prioritizes intense hydration for dry skin, particularly favoring rich moisturizing products for mature skin needs. They value quality and effectiveness, even at a higher cost, and deeply appreciates the natural earthy sense and textures that accompany their skincare ritual. She's a female between 50 and 70. likely living in a dry or cold climate region. She's retired or semi professional has interest in natural beauty products and self care. Her personality is dedicated, conscientious, luxury seeker, and sensorily appreciative. Her favorite products are the intense cactus concentrate gift, the petite grain moisturizer. And now, okay, I have an idea of who I'm talking to. Right. So think about some of the things that come to life that are live in a cold environment, bamboo, or moisturizers really help with cold, dry or dry skin created in cold, windy environments. Okay. So now as a creative strategist, and if I know, all right, I'm selling intense cactus concentrate to the the hydration devotee. Now, when I go to think about the angle or the benefit that that person cares about, I can be more specific. And in fact, we actually offer them specific angles related to that persona. So the angles we would define we, we break them into 30 different components. This is based on an HBR. Harvard business review study of like the 30 psychological reasons people purchase and we break them.

So there's like emotional benefit, wellness, functional benefit, quality, sensory benefit, smell, sensory benefit, touch, comparison of price, emotional benefit, fun or entertainment, emotional benefit, reduction of anxiety. So those are all angles that this specific persona cares about. Right? So now if I, let's just take the smell angle.

So if I go, okay, I am delivering intense cactus concentrate, To a woman between the ages of 50 and 70 living in a cold environment who like, who responds to smell that gets you pretty far down the road to writing an idea for an ad brief and a visual set to apply against that offer audience and angle. And so the, the, the, the language analysis, giving these categories again, more boundaries, more constraints, less. Open ended thinking and saying, okay, you have to sell this product to this person at this angle. Now write a brief to do that. And in fact, if you want a prompt to help you write a brief, here you go.

Here's an automated LLM output that now can get translated into a concept that goes into the concept log with a go live date and a plan publication. And so that, that, again, this is all about taking a big, giant problem, breaking it down into building blocks and Legos that creative strategists can use to accelerate the pace of their own work.

[00:21:27] Richard Gaffin: So, yeah, so basically what it does and you kind of, you laid it out. Well, just there's like, it reduces the amount of that sort of like the inspiration part of the creative process. It reduces it almost as much as it possibly can at the end of the day. Like the creative strategist is still going to come up with, this is how we're going to say how we're going to communicate emotional benefit wellness to a holistic beauty devotee using this product or whatever.

But they're not just thinking across, like, how do I sell, you know, Any old product from this brand to somebody, they have a bunch of really, really specific guardrails to kind of direct them towards what they have to create. And then once it's created, It automatically kind of goes back into the system

and it's logged into the concept blog and all that kind of thing.

So, let's talk a little bit about 

[00:22:09] Taylor Holiday: Well, can I just, you said some important things there. The other thing that what this does is it enables actually better reporting and analysis, right? Because if you can sort of make an add to anyone with any offer at any time, like the ability to then organize that into a system of analyzing, okay, how does this audience perform is very complicated and depends on like this like really Challenging ability to create syntax continuity over time for

all customers. But when we're assigning these sort of like meta attributes to each campaign that gets published, yes, it creates a syntax that's consistent, but it doesn't actually depend on it.

And so it allows us also to start to organize the data in a way that we can begin to create better analysis of how these audiences perform and how these different offers and concepts perform as well.

[00:22:57] Richard Gaffin: Yeah. Well, it creates a framework for like creative testing for lack of a better word that goes beyond changing the background color to blue or whatever, or like tweak the headline, whatever that means. It's like, there's, there's much more clarity around Kind of the actual, what's actually important about why an ad did or didn't work.

It's like, obviously the angle off for an audience. We have like a little bit more data on what that is. So we can sort of. Iterate where we need to iterate in ways that are more complicated than just, I don't know, switching 

[00:23:24] Taylor Holiday: That's right. But, and I also think that brands should do more work to change their collective understanding at the root level versus the ad level. What do I mean by that? Which is that like, let's define the audiences that we're going to talk to. And let's stay bound within that until someone agrees to edit the audience pool. So like, if we create five audiences, what I don't want is all of a sudden you introducing a seventh, eighth, ninth, and 10th, like we're, we're, we're going to work within these. And now if someone wants to make a case that there's a sixth or to edit one of the five, then cool. We're editing the option pool for all participants. Not sort of constantly creating new end nodes to the system. And this is just another thing where like brands don't do enough work to stop and ask themselves, okay, who are we trying to build knowledge and understanding of within some sort of audience persona with continuity, right? It's like this constant tiny sampling all over the place versus like this deeper thought around it.

Who is this person? How do I make them care about this? How do I try and try and try until I found the right thing that they really care about? And then how am I constantly analyzing what they're telling me about what they care about? Right? And so I think both of those things, they allow for like a little bit deeper work, a little bit richer work.

Versus this just such fragmented random ideation, because when, when, when often what happens is when there's a pressure for volume, the ideas tend to become more scattered and fragmented because there's no system for an up volume enablement keeps those boundaries in place.

[00:24:48] Richard Gaffin: Right. Yeah. It allows you to refine, refine. Yeah. It allows you to refine the thought work on the front end, I guess. Like, the more that you do this type of work, the more that you will be able to generate better work in the future in a way that like a B testing, like little kind of tweaks doesn't really.

Provide you with any actionable information of any kind, I think. So let's talk about then what, what the sort of workflow. Okay. So we talked about like the, what happens before that kind of moment of creative ideation, then during, and then afterwards it kind of goes off, let's say back into the, back into the system, goes off for analysis for being run, all that type of things.

So talk to me about like what the workflow kind of looks like afterwards.

[00:25:26] Taylor Holiday: Well, so what happens is the end result of building this sequence of offer audience angle is a, is a, what we call a concept, which in CTC language is ultimately the design of a campaign. That's going to go live and meta. And so what, what the creative strategist is doing is ultimately building a campaign that the media buyer will then pick up and publish with an expected go live date and budget. And so that concept plan actually shows up in the media plan, Which shows up into the monthly budget plan so that you can see, okay, on February 22nd, these two campaigns are going to go live with this budget. That's going to represent this portion of the spend plan for the month. And before the month starts, the expectation is that every campaign that's going to go live is planned with the go live date and budget such that we can look out and go, okay. We are going to spend 822, 000 and we're going to spend it exactly in this way across these campaigns and that part of like starting the month with a clear plan of where every dollar is going to go. And again, not because you're going to be right. Some campaigns inevitably going to perform better or worse, but it allows you every day to see where you're importantly wrong or right and whether or not you have enough creative plan to get to where you're going. And again, the number one problem I see for everybody is that they have under planned the creative output for the job at hand. And then they scramble and they're on this, like the hamster wheel of creative that exists in the world of people running around, trying to make more ads because they under serve the task is like, it's exhausting and it is endless.

And it's because most internal brand design systems are not built like this, especially the larger, the business you get, there's usually like a design department. With a ticketing system and there's tickets that come in for point of sale displays and emails and every other design task in the world. And the ad is just like another ticket in the system, right? And that will not work for meta creative volume scale. You have to be, have dedicated resources, understanding where you're going. And the other big thing that this does is it helps to, when you model out the number of ads that you need every month, What you can do is you can understand what portion of this work should be dedicated to FT ease and what should be variable.

So if

you think about, like, let's say that the bound of expectation of ads per month runs from 200 to 1000, and it is that different because e commerce businesses are that seasonal. And so the lowest common denominator every month is you need at least 200 ads every single month. Well, there's the job of FTE resources.

You need to resource to 200, but that flex work up to a thousand that only exists for one month or two months, that variable work. Is where you need to bring in flex staffing, whether that's agency, whether that's freelance, whether that's designers that show up, but you, you can't, what I see teams do systems do is they either resource to the 200 and they're constantly stretching and overwhelmed in the other periods or they're under delivering, or they staff to the thousand and then half the year they have people who were doing one 50 amount of work that they needed to elsewhere and they're completely over resource.

[00:28:23] Richard Gaffin: So it sounds like essentially what this does from, from beginning to end. It sounds like this is really like a system for reducing. I don't know if we're reducing thought work is the right word, but a lot of the times, like, I think when we approached the creative problem the issue was that basically like everybody from like the people who were like planning the operations, the actual like strategists themselves to the graphic designers, we're all working with the kind of a blank page, basically like I'm sitting here in front of, of nothing and I have to just come up with something and that has to result in 200 ads or 1000 ads, whatever the case may be,

but with what this does is reduces.

It gives you an answer, basically, to all of the different restrictions that need to be in place before any creative work actually gets done.

[00:29:04] Taylor Holiday: want 

[00:29:05] Richard Gaffin: And then kind of helps you understand maybe after the fact, what to do about it next time.

And, and that, that's kind of like, to me feels like the thing that makes creative so difficult when it comes to slotting into the e commerce world is it requires all of this sort of weird kind of abstract thought work.

But if that can be reduced to the single point and then everything around it, all the thought work is already done. Feels like that's ultimately what solves the problem.

[00:29:28] Taylor Holiday: That's exactly right. It defines the job to be done and then gives you the tooling and resourcing to be specific about how to solve it. So, so I just think about something like inventory, like it can't be that it's your strategist job to also have like immense clarity of the age of inventory that exists and be

thoughtful about making like that.

Like if you are asking that of someone, like you're just, you're creating a unicorn that doesn't really exist. And instead what you need to be delivering to that person is this is how many ads you need to make. We need to make them about these things. And here's the source material for good data analysis about who to talk to and why, and like in that space, people can be highly effective.

Now, is there also an opportunity to like, even take the system a step further where it also generates ideas within the bounds of that, for sure. Like, I think we're going to get to the AI doing most of this creative production work, but at the very least right now, I think we're Like if I think about like if a creative strategist were to start with a brand brand new, and that you were to ask them, like, what is all the research that you do to understand this brand? Like, what we're trying to do is answer that question automatically for them is to not make them go spend 50 hours dragging out all of the historical ad performance and reading a bunch of reviews and looking at the best selling products and understanding the margin and inventory position and marketing calendar.

And like, you know, The hours and hours of work that would go into understanding that we want to give it to them instantaneously at their fingertips, in the most thoughtful way possible that aligns with the intentions of the brand and is connected to the financial objectives of the organization. Like that's the job. And so this creative strategy product, that's why we're calling it a creative strategy product. It's not a production product yet. This is all about accelerating and improving the quality of creative strategy work that exists. I think is is a monumental step for the industry to really think about how to do this job better.

[00:31:08] Richard Gaffin: Well, as we often like to say, great information creates great inspiration. And I think ultimately what Compass does is it gives you all of the information. So you can just have sort of have the expectation for your creative strategists that what they have in front of them is all of the best information.

And therefore their shots Every shot that they take creating an ad is going to be the best it can possibly be because they have the research right in front of them.

[00:31:31] Taylor Holiday: That's right, we're, we're, and the Compass, like the whole thing about these systems is they're partners in improving the quality of your work. Right? And so that's what we want Compass to be for creative strategists is we want them to feel more enabled than ever before to do better work faster.

Right? And to understand how much work is actually being asked of me. And have I done my job? Well, because you know, what's constantly frustrating is like, Okay. Make five ads, get told they don't work to make five ads, get told they don't work and feel frustrated that like, I'm failing at this instead, we want to design a system with the acknowledgement that most things don't work to in light of that, figure out how many things should we make to ensure that we have a high chance of succeeding. And then what should we say along the way? And like that is freedom that actually gives people a sense of clarity of what they're trying to do. And I think it's, it's going to be a big step for us in improving that process.

[00:32:15] Richard Gaffin: All right. Well, so if you'd like to check it out, you know, where to find this common thread code. com, hit the hire us button. Also, I believe our brand new URL adcompass.ai. Go check it out. Drop us a line. We would love to chat about building this for you, solving this most difficult problem for you right now.

Taylor, anything else you want to hit on this?

[00:32:32] Taylor Holiday: I think we're going to do a product demo here in the next couple of weeks. I probably won't be on the audio version of this. Maybe you'll see it on the socials, but you're going to see more and more. And you guys know DTC, Jacob, he's been leading this charge. Big thanks to him. Jacob was L our director of AI for this project.

This is really the first sort of output of that. We're going to continue to evolve this. We're going to provide more data sourcing that informs the persona development. So thinking beyond just reviews to either what other sort of semantic database of information can we provide to enrich the information.

We're going to get this all the way through to automated, to add production where we're bringing in some generative AI and other things as well. So lots more to come and how the system evolves, but it has to be on a foundation of connection to the financial plan. Enrichment with the marketing calendar and an understanding of the of the job to be done for the business.

And in that place, and this is why I think CTC is uniquely positioned to do this is because that's the start, right? And without the financial plan and the clarity of the budget, creative strategy cannot work. It will be disassociated. So it's really key that that foundation is in place and where I think we're able to offer something really novel into the world.

[00:33:35] Richard Gaffin: Sorry, folks. All right. Well, thank you all for listening. Thanks for joining us. And we will talk to you again next week. See ya.