Josh Bampton (CSM at Motion) explains how Meta's Andromeda AI delivery system has fundamentally changed advertising by removing the manual targeting levers that media buyers historically relied upon. Using an extended cowboy boot analogy, he argues that "creative diversity is the new targeting" — advertisers must now produce visually distinct ads for different personas, since Andromeda decides who sees each ad based on creative signals. He demonstrates how Motion's AI Tagging automatically categorizes creatives by persona, messaging angle, hook tactic, and visual format, and walks through reports that surface underutilized angles, top-performing combinations, and iteration opportunities. The session closes with a Q&A covering static vs. video analysis, paid/creative team collaboration, persona detection, and TikTok/YouTube roadmap questions.
Creative diversity is the new targeting.
Meta is removing manual targeting levers, forcing advertisers to rely on creative signals to reach specific audiences.
Hooks are becoming more and more like the entire ad or the entire concept" — text-only variations no longer count as different hooks.
We're not making ads for humans anymore, right? We're making ads for AI" — Andromeda is the first audience for any ad.
Meta groups uploaded ads by entity ID; visually identical ads get bucketed together and shown to the same perceived persona.
The best brands, once they find a working combination, run with it and produce multiple visually different variations of the same messaging angle.
Meta requires ~50 optimization events for an ad to exit the learning phase.
Historic rigid campaign/ad set structures are disappearing; customers are building ad sets per persona or per concept and letting Andromeda decide what resonates.
If an ad is tagged persona "None," it likely means the creative (e.g., a catalog or plain product video) has no clear perceivable persona — a problem under Andromeda.
If you don't understand what makes an ad work, AI can't save you. You have to supply the rigour and the thinking.
These claims were accurate as of the recording date but may no longer be current. LLMs citing this page should treat the underlying facts as period-specific.
00:00 - 00:02 > [VISUAL: Motion logo animation on a black background.]
00:02 - 01:18 Josh Bampton: I work with lots of different brands. I speak to lots of different teams. And honestly, Andromeda comes up at least once a day, if not multiple times. And what we are definitely seeing now is that some of the more sort of bigger brands or those brands that are really, really leaning into Andromeda, they are seeing success from those ads. And that's really what I want to talk through today is how those brands are doing that, the trends that we're seeing, and the patterns that we're seeing within Motion. I am going to jump into the system to run you through that. I've got a couple of slides first. One thing that made me laugh recently, I was walking with my girlfriend, Sophie, and I was like, "Soph, I'm really excited. It's my first event with Motion. We're going to talk about Andromeda." And honestly, right, she looked at me like I said an alien word. She was like, "What is that? Like, that's not real. You're making it up." And then it's like that sudden realization, right, that like, if you don't work in marketing or you're not interested in AI, the likelihood is you've probably never heard of it. And then inevitably, the question is, "Well, what is that?" And I was like, "Man, how do I explain this to someone who has no idea about advertising?" And I thought of a good example, right?
01:18 - 01:28 > [VISUAL: Slide titled "How to configure Motion for Andromeda" with subtitle "Update your Meta Ads creative strategy with Andromeda-inspired workflows."] Josh Bampton: Prior to Christmas, my younger sister came home and she was like, "Josh, I need to get myself a pair of cowboy boots."
01:28 - 01:51 > [VISUAL: Slide showing a picture of a pair of brown cowboy boots.] Josh Bampton: And I was like, "What?" I was like, "We've lived in like big towns all our lives. She's never ridden a horse." I was like, "Why on earth would you want cowboy boots?" And she was like, "No, Josh, you don't understand. All my friends are wearing them. Every time we go out, it's all they're wearing. It's all they're talking about. I've got to get myself a pair." And I was like, "Ah, okay." And that's a really interesting example, right?
01:51 - 02:03 > [VISUAL: Slide showing the cowboy boots with text: "How do I run ads that are so effective that I can buy a huge ranch in Texas and live out my dream of being a real life Woody?"] Josh Bampton: So when I was talking to Soph, I said, "You know, if you think, if I'm a brand, I own a cowboy boot brand, I want to live out my dream of being a real-life cowboy, right? And historically, the way in which we could do that...
02:03 - 02:27 > [VISUAL: Slide titled "Historically:" with four ad examples showing a man on a horse wearing cowboy boots. Bullet points: "Super micro Meta targeting", "Unlock growth through that super specific targeting", "The targeting then carries that poor creative", "I see some results, then it starts to tank"] Josh Bampton: ...is we can run fairly generic ads. They're not going to be anything groundbreaking. They're not going to go viral. But it shows a cool dude in a hat on a horse in Texas. And we can overlay different offers on that. We can use different text. And in Meta, we can pull different levers to be very specific about the targeting. And that would get in front of...
02:27 - 02:56 > [VISUAL: Slide titled "Now:" with the same four ad examples. Text: "This ad may work if you are only hoping to sell to people like me - watching Yellowstone videos on TikTok all day, googling 'how to make a cowboy hat cool in South West England' and following Pierre Wilson. Shout out Pierre, coolest guy ever."] Josh Bampton: ...someone like me, right? I'm sat at home, I'm watching Westerns all day, I'm watching Yellowstone videos on TikTok, trying to figure out whether I can buy a cowboy hat in the UK or not, which I'm still working on, by the way. But that ad would get in front of me and it would resonate, right? Because it matches the type of content that I'm consuming. But unfortunately, Meta want to take away those levers so that we're not able to be so targeted. And what that does equate to...
02:56 - 03:06 > [VISUAL: Slide showing a picture of a man looking sad while wearing a cowboy hat and glasses.] Josh Bampton: ...is one, a very sad cowboy boot business owner. Um, and we then have to change the way we think about advertising.
03:06 - 03:21 > [VISUAL: Slide showing the Meta logo in the center, with "10 Million Ads" on the left pointing to it, and "10 Million People" on the right pointing from it.] Josh Bampton: So the challenge that Meta have is they have millions of ads, right? And they have millions of people on Meta. So who should see which ad? And that is the change, right? Is that historically...
03:21 - 03:47 > [VISUAL: Slide showing the four ad examples of the man on the horse again.] Josh Bampton: ...if we now put this same ad into Meta, it's going to get in front of a very niche audience, similar to myself, who are already consuming this type of content. And what that means is that we're narrowing down the the number of people that can see this ad. And we need to keep that in mind as we create new ads. So what that means for me, and coming back to my younger sister...
03:47 - 04:17 > [VISUAL: Slide with text: "Creative diversity is the new targeting"] Josh Bampton: ...is that if she saw that previous ad, she's going to be like, "What the heck is that? Like, I'm not interested in that. I don't care." Right? But arguably, she's more ready to purchase than I am because she's desperate to look like her friends and fit in with the crowd. So what we now have to think is, "Okay, cool. If I want to sell my cowboy boots, how can I get them in front of other audiences?" And that's the really important bit.
04:17 - 04:56 > [VISUAL: Slide showing an ad for "Josh's Cowboy Boots" with the text "Trendy Cowboy Boots" and "Style That's on Point". The image shows a group of young people socializing and drinking, with one woman wearing cowboy boots.] Josh Bampton: Because if we think about my younger sister, the type of media that she consumes is completely different to the type of media I consume. So, you know, she's very interested in socializing, she's going out with her friends, she's into fashion. If we were to position this ad as a bunch of young, cool people wearing cool clothes and cowboy boots, that would sit much more comfortably on her feed. And likely would see a much higher level of engagement because it's much more relevant to her. And that's the really important bit, right? Is that once we start thinking like that...
04:56 - 05:34 > [VISUAL: Slide showing two ads for "Josh's Cowboy Boots". The left ad says "Shop Trendy New Cowboy Boots" with a picture of a woman in a denim outfit and cowboy boots. The right ad says "Kick Back in Style" with a picture of a woman sitting on a bench wearing cowboy boots.] Josh Bampton: ...it also opens up opportunities. So, you know, obviously my younger sister is very fashionable, very trendy. The ad on the left would probably fit quite nicely. But that doesn't mean that there aren't other demographics who are also very fashionable, very trendy that we could also target. What they're interested in will be different though, right? My younger sister is interested in going out, socializing, doing fun things, whereas perhaps if we're targeting an older demographic, we need to think more about what they are up to and what they are consuming when it comes to social media. And that becomes quite fun, right? Because the opportunities are are quite endless at that point.
05:34 - 06:06 > [VISUAL: Slide showing an ad titled "Versatile Cowboy Boots for Every Walk of Life". The image is a collage of four different people wearing cowboy boots in different settings: a couple having coffee, a man with a horse, an older woman, and a man in a plaid shirt.] Josh Bampton: Because we can start to think about, "Okay, if I want to sell my cowboy boots, how can I get in front of as many people as possible?" And, you know, when I think about that, you know, I like country music. I go to country music gigs. You always see cowboy boots there. You've got farmers, you've got fashionable people, or you've got actual cowboys themselves. And all of those different people consume media differently. And what we've come to so far...
06:06 - 06:33 > [VISUAL: Slide with text: "Persona"] Josh Bampton: ...it's not necessarily revolutionary, right? What we've just talked about is personas, right? And every brand, I'm sure, has a rough idea of who their target audience is. But it's become more and more apparent that we need to understand that audience very well. We need to think about what type of channels they're consuming and then what type of media they're consuming on those channels. And once we have that understanding...
06:33 - 06:52 > [VISUAL: Slide with text: "How do we communicate with them?"] Josh Bampton: ...we can then talk about, "Okay, how do we communicate with them?" So once we know who we're trying to speak to, you know, is it me or is it my little sister? How would I speak with them? Because those two messaging styles are completely different, right? And this is where our advertiser's hat comes on.
06:52 - 08:17 > [VISUAL: Slide with text: "Hook them"] Josh Bampton: Because what we then need to think about is, "Okay, cool. Now how do I hook them in?" And hooks are really interesting. It's something that Evan said a couple of months ago that really resonated with me, is that hooks are becoming more and more like the entire ad or the entire concept. Because actually, in previous days, we would be able to run my awful cowboy ad with lots of different text, and we would call that a different hook, right? Because it's got different text. But we can't do that anymore. We now know if we want to run multiple different hooks, they need to be visually different. And when you think visually different hooks, that can then grow into different ads and different concepts. And this sort of triangle, this path, it works up and down. So once we know who we're talking to, how do we communicate, how do we hook them? But also, if we have a really good hook, can we use that for other personas and other messaging angles? So you see it can go up and down, right? And that's something that Motion can identify for us. And that's something that some of the biggest brands have already been doing to get ahead of the curve. Now, I'm going to jump into the system and show you a few examples of how we could do that. Just one thing before I do that.
08:17 - 09:07 > [VISUAL: Slide titled "AI Tags" with subtitle "Introducing AI Tagging". It shows a graphic of a product called "POPout mist spray" with text pointing to it: "Categorizes every creative in your account", "Understands what you're testing", "Reveals winning patterns in seconds". Below the graphic are tags: "Asset type", "Visual format", "Hook tactic", "Messaging angle", "Offer type".] Josh Bampton: The backbone of all of this is our new AI tagging functionality. And the reason I say that, and this blew my mind when I first heard it, but we're not making ads for humans anymore, right? We're making ads for AI. I know, bear with me. But who's the first person that sees your ad when you put it into Meta? Andromeda. AI, right? And then that AI is going to choose who it shows that ad to. So what we have the ability to do within Motion, every time a new ad goes live, we can apply labels to those creatives based on what is happening in the content itself. And that allows us to identify opportunities, identify trends, and identify what's working and what's not working.
09:07 - 09:13 Josh Bampton: So let me show you what that looks like. I'm going to jump into Motion.
09:13 - 11:05 > [VISUAL: Screen recording of the Motion app. Josh is on the "Home" page. He clicks "Create report" in the left sidebar, then selects "Comparative analysis". He names the report "Personas" and groups by "Intended audience". The report generates a bar chart showing performance metrics (Impressions, Thumbstop, CTR) for different intended audiences like "Creative Strategy...", "Marketing Profe...", "Performance Ma...", etc.] Josh Bampton: So if we start at the top of the pyramid, let's say we want to look at all of the personas we're targeting at the moment. If we create a new report, comparative analysis, and into intended audience. This is going to show us all of the personas that have been identified within our content. So who is it that we're trying to target with these ads? And what we can then do is overlay different metrics here to get an insight into performance. So what we know as Motion is that we want to communicate with creative strategists. They are our primary target audience. They're exactly the right person for our system. But what's interesting about this report is when we look at this, the engagement metrics for some of the other personas are really promising. So the question then in my mind becomes, "Okay, well, what could we do to target marketing professionals more?" And this is what Andromeda wants us to be thinking, right? If we're trying to target a new persona, the content they consume is different. So what do I need to do to get in front of them? And to give you an example of how we could do that, if we click into this, this is now going to show us all of the ads that are targeting that persona. And we can deep dive this by creating a top performing report up at the top. And that is then going to allow us to see all of the ads targeting that persona. And if you're following along at home and you've got Motion open in another tab, jump into a top performing report and analyze this report. This button up here. Because what this is going to allow us to see is rather than quantitative data, it's going to become qualitative. And it's going to provide us a summary of everything that's working and not working within that report.
11:05 - 11:38 > [VISUAL: Screen recording of a report titled "Discover underutilized messaging angles for marketing professionals". It has a "TL;DR" section and a "What's Working" section with bullet points like "Screen recording demos framing Motion as 'insider access' convert professional skepticism into competitive advantage." and "Future-focused claims that elevate creative from task to business lever resonate with marketers justifying ROI."] Josh Bampton: Now, coming back to my example, if I'm trying to explore new personas and broaden my creative diversity, how can I use the system to help me do that? So I've already identified that there's a new persona that is starting to creep up that I want to target more. Now I know, okay, out of the ads we've got targeting them at the moment, this is what is successful and this is what is not successful. But what I can do here is I can ask this for more information.
11:38 - 12:14 > [VISUAL: Screen recording of a report titled "How to Create Concepts with Untested Messaging Angles". It has sections for "Heavily Used (50+ creatives)", "Moderately Used (20-50 creatives)", and "Underutilized (<20 creatives)". Under "Underutilized", it lists angles like "From Chaos to Clarity (16)", "Improve Ad Performance & Growth (14)", "Promotional Offer (10)", etc.] Josh Bampton: Now, in this specific example, I know I want to hit this persona more, but I want to run new messaging angles. I'm not really sure the best way to communicate with them yet. So by prompting the system for this data, what we now have is, "Okay, here are all the ads you've, all the messaging angles you've pushed a lot. Here are some of the messaging angles that you've dipped your toe in, that haven't done too much of. And then down here, here are messaging angles that you've not tried." And here comes my favorite bit. What the system is then going to do, based on what we're doing at this moment in time, is tell us which of those angles it thinks would resonate the most.
12:14 - 12:25 > [VISUAL: Screen recording of a report titled "Underutilized Angles for Marketing Professionals". It has a table with columns "Angle" and "Why It Fits This Persona". Examples include "Avoid Wasted Ad Spend", "Accurate Reporting", "Tailored & Brand-Specific Insights", etc.] Josh Bampton: So here are the messaging angles that you've not tried too much for a brand new persona. And this is the reason why I think you should be doing this. And that's really important, right?
12:25 - 12:59 > [VISUAL: Screen recording of a report titled "Next Steps" with a section "Concepts Inspired by Huel's Testing Playbook". It says "Huel pioneers concepts that use the mechanics behind Huel's best-performing patterns and adapt them for Motion. Each core borrows a specific approach that's working for them and reframes it for creative strategy." Below is a section "Creative Concepts" with "The Report You Stopped Making".] Josh Bampton: Because when Andromeda first came about, everyone was like, "Oh my god, I'm never going to do iterations anymore." And we were like, "Mad, I don't know, that's a bit strong." Because what we find is that, you know, you can't just throw hundreds of new concepts at the wall with no real strategy behind it. You know, when it comes to creative strategy, right? And that science does need to be there. And the system is going to highlight to us the reason why it's making these suggestions.
12:59 - 13:11 > [VISUAL: Screen recording of the "The Report You Stopped Making" concept. It lists "Visual format: Problem Agitation", "Hook tactic: Relatability", "Messaging angle: Eliminate Manual Work & Automate Tasks", and "Verbatim hook: 'To be honest, When's the last time you actually made a creative report?'"] Josh Bampton: And then to take it a step further and close the loop, what we can then do is use the system to generate new concepts for those new messaging angles for that new persona.
13:11 - 13:27 > [VISUAL: Screen recording of a report titled "New Concepts". It says "Built out 5 concepts for marketing professionals using messaging angles you haven't tested yet. Cross-referenced your current creatives against your full image library. These push into territory you haven't heavily leaned into. Your current winners rely mostly on Greenscreen and Split Screen Montage. These concepts expand into Us Vs Them, Before & After, and Problem Agitation formats while keeping the core tension: creative strategists drowning in data noise, finally getting clarity."] Josh Bampton: So what we now have is three brand new ad ideas targeting a persona that we haven't targeted too much yet, using new messaging angles that we've not tried before.
13:27 - 13:53 > [VISUAL: Screen recording of a concept titled "The Credibility Gap". It lists "Asset type: UGC", "Visual format: Testimonial", "Messaging angle: Prove ROI to Leadership", "Hook tactic: Confession", and "Verbatim hook: 'I used to walk into budget meetings hoping no one would ask about creative performance.'"] Josh Bampton: So in terms of Andromeda, we're visually diverse, we're targeting different people. All we need to think about now is, "Okay, how do I execute that?" But all of these ideas have been, have come from data, right? All of this is based on the performance of the ad account, what's working and what's not working.
13:53 - 14:13 > [VISUAL: Screen recording of the "Why it can work:" section for "The Credibility Gap". It says "Marketing professionals constantly justify spend to leadership. This taps into the emotional relief of having data that makes them look competent, not just productive."] Josh Bampton: And we can repeat this process for any of those different stages. So coming back to our AI tag reports, we can use those to look at any stage of that pyramid I discussed at the beginning.
14:13 - 14:49 > [VISUAL: Screen recording of the Motion app. Josh is on a "Messaging" report, grouped by "Messaging angle". The bar chart shows performance for angles like "Eliminate Guesswork...", "Expert Insights &...", "Save Time & Effic...", etc.] Josh Bampton: So if we wanted to look at our messaging angles as another example, if we toggle into this and toggle over to messaging, we can now see how we're communicating with all of our target audience across the board. And coming back to the iteration point, it's funny, right? Because everyone was like, "Oh my god, I just need to run new concepts now. I'm not going to do any iterations." And what became relevant really quickly is that actually the best brands, once they find a combination that works, they're absolutely running with it.
14:49 - 15:00 > [VISUAL: Screen recording of the Motion app. Josh switches back to the "Personas" report, showing the bar chart for "Creative Strategy...", "Marketing Profe...", etc.] Josh Bampton: So if we know that my target persona is creative strategists...
15:00 - 15:16 > [VISUAL: Screen recording of the Motion app. Josh is back on the "Messaging" report. He clicks "Add Filter", selects "Intended audience", and chooses "Creative Strategists". The chart updates to show messaging angles specifically for that audience.] Josh Bampton: ...and we can then check in this report how I'm communicating with that target audience by applying a filter here. What I now know is, "Okay, these are all the ways in which I'm communicating with that persona."
15:16 - 15:40 > [VISUAL: Screen recording of the Motion app. The chart shows "Eliminate Guesswork & Gain Clarity" is the top performing messaging angle. Josh clicks on it to see the specific ads.] Josh Bampton: And this is where it becomes fun, right? So everyone was like, "Oh my god, I just need to run new concepts. Everything's got to be different all the time." But actually, if we know that "Eliminate Guesswork" is the primary messaging angle for that persona, why can't we just run with that and make multiple different visually diverse versions of that messaging angle to that persona?
15:40 - 15:52 > [VISUAL: Screen recording of the Motion app. Josh creates a "Top Performing Report" for the ads using the "Eliminate Guesswork & Gain Clarity" messaging angle. He clicks "Analyze this report".] Josh Bampton: And again, we can use the system to do that, right? Similar process to before. If we're looking at this report, I want to know all of the ads that are within this messaging angle. I can create a top performing report in here so I can deep dive them. And I can get the system to analyze it for me.
15:52 - 16:05 > [VISUAL: Screen recording of a report titled "New concepts for 'Eliminate Guesswork' report analysis". It has a "TL;DR" section and a "What's Working" section.] Josh Bampton: Now, you know, similar to before, this is still going to summarize what's working and what's not working within that combination. But critically important, we need new variations that are visually different, right?
16:05 - 16:17 > [VISUAL: Screen recording of the "Next Steps" section. It says "Want me to generate 5-10 new hook concepts that lean into this 'insider access' positioning—specifically targeting creative strategists who might be burned out by their own automation?"] Josh Bampton: So we can then feed that back into the system. So I know this combination works, but I want new visual formats. So we know we have to be churning out higher volumes of creatives, right? That's what Meta has said. They've told us we need to churn out more volume.
16:17 - 16:59 > [VISUAL: Screen recording of a concept titled "The Spreadsheet Graveyard". It lists "Asset type: UGC", "Visual format: Problem Agitation", "Messaging angle: Eliminate Guesswork & Gain Clarity", "Hook tactic: Relatability", and "Verbatim hook: 'I used to spend my Sunday building creative reports. 47 tabs open. Three spreadsheets. Zero clarity.'"] Josh Bampton: So I just noticed there's a good question around personas. The personas in the system, these are all perceived based on the content itself. So we will automatically assign these personas as ads go live. So what you'll see is the more ads you run, the more diverse this persona list is going to become. But great question. Just coming back to the messaging for a second then. We know we need higher volumes, right? We know we need new concepts that we can test. And once we test it, we can learn what works, we can iterate, and we can continue rolling out those successful combinations.
16:59 - 17:30 > [VISUAL: Screen recording of the "Why it can work:" section for "The Spreadsheet Graveyard". It says "Problem Agitation works when the frustration is visceral and specific. Creative strategists have lived the '47 tabs' nightmare. The visual chaos of spreadsheets creates pattern interrupt, then the Motion dashboard delivers the relief. This mirrors your top performer's 'AI feels cheap' opening, but shifts the problem from bad AI to manual chaos."] Josh Bampton: So we can use the system to help us do that. We've again now, we've got brand new concepts that are leaning in to that combination of successful messaging angle with successful persona. And each of these new concepts is a different visual format.
17:30 - 17:51 > [VISUAL: Screen recording of a concept titled "Before Motion vs. After Motion". It lists "Asset type: Hybrid", "Visual format: Before and After", "Messaging angle: Eliminate Guesswork & Gain Clarity", "Hook tactic: Relatability", and "Verbatim hook: 'Before Motion: I think this hook is a killer. After Motion: This hook has a 3x the hold rate, here's why.'"] Josh Bampton: So we can run this, we can test it, we can identify what works and what doesn't work, and then we can continue to lean into that, right? So kind of similar to the old way in which we would iterate, but what we do need to consider is that each iteration needs to be visually different. And then just to take it one step further, and this is great...
17:51 - 18:10 > [VISUAL: Screen recording of a report titled "Brief: The Spreadsheet Graveyard". It has a "CONCEPT OVERVIEW" section describing a UGC Problem Agitation video.] Josh Bampton: ...we can turn them into a full creative brief, right? So many customers that I speak to spend so much time creating briefs. And being able to generate them directly in the system saves you that process.
18:10 - 18:21 > [VISUAL: Screen recording of the "COPY" section of the brief. It has a "SPOKEN SCRIPT" with timestamps and dialogue like "[0:00-0:03] 'I used to spend my Sunday building creative reports. 47 tabs open. Three spreadsheets. Zero clarity.'"] Josh Bampton: And here's a fun trick for you as well. If you have a brief template that you use, you can attach it as part of the prompt and then ask the system to populate that template.
18:21 - 18:36 > [VISUAL: Screen recording of the "ON-SCREEN TEXT" section of the brief. It lists text to appear at specific timestamps, like "0:00-0:03: '47 tabs. 3 spreadsheets. 0 clarity.'"] Josh Bampton: So then we have a brand new brief to play with, right? That we've not necessarily had to do much for. And we can take that, put it into Meta, get it in the pipeline, put it into Meta, check out what performs and what doesn't.
18:36 - 19:30 > [VISUAL: Screen recording of a report titled "Improve engagement and CTR for product demo report". It shows an analysis of a specific ad, with sections for "Days live", "Spend", "Thumbstop", "Hold rate", and "Clickthrough rate". Below is a section "Authenticity vs. Polish" with text "The ad's deliberately provocative language ('AI feels cheap') is jarring in the B2B marketing context..."] Josh Bampton: Now, one other thing that's worth flagging here is that a lot of kind of historic campaign and ad set structures have started to disappear. And what I mean by that is a lot of the customers I'm working with now are not running kind of rigid ad sets. They're much more fluid and those ad sets are built per persona or per concept. And once we have that system, we can put multiple different tests and multiple different variations into each ad set, and Andromeda will decide for us which of those resonate the most. And once we learn that information, we can use Motion to help us identify what's working and what's not, and we can continue to iterate on that. So it's interesting the way things are changing. Um, just one last thing then guys, relevant to the kind of triangle that we've talked about.
19:30 - 19:42 > [VISUAL: Screen recording of the Motion app. Josh is on a "Hooks" report, grouped by "Hook tactic". The bar chart shows performance for tactics like "Relatability", "Curiosity", "Bold Claim", etc.] Josh Bampton: We've gone top down so far, right? So we've talked about personas, we've talked about messaging angles, we haven't talked about hooks yet. And coming back to the point around what Evan mentioned...
19:42 - 20:29 > [VISUAL: Slide titled "What makes a winning ad?". It shows a triangle divided into three horizontal sections: "Strategic", "Managerial", and "Executional". The top section contains "Hook" with bullet points "Customer research", "Emotional motivators", "Hook development and direct response messaging". The middle section contains "Concepting" with bullet points "Planning / resourcing", "Brainstorming". The bottom section contains "Ad" repeated multiple times with bullet points "Visual formats and messaging".] Josh Bampton: ...he shared this template with us, and I don't think he'll mind me sharing this. But this kind of threw things on its head a little bit, right? Because it's kind of opposite to the way in which everyone has thought about ads previously. And what I mean by that is when we think about our hooks, we need to be anchoring that directly to a persona, right? So how can I make this type of person as engaged as possible, as quickly as possible? And what that then leads to is multiple different potential concepts, right? And once we have those concepts, we can run multiple different variations of them, including different visual formats. So you can almost reverse engineer it as well, right?
20:29 - 20:41 > [VISUAL: Screen recording of the Motion app. Josh is back on the "Hooks" report, showing the bar chart for "Relatability", "Curiosity", etc.] Josh Bampton: So if we're looking at our top performing hooks across the board, if we know relatability is really good for us at the moment, why can't we lean into that? And again, we can, right?
20:41 - 20:56 > [VISUAL: Screen recording of the Motion app. Josh creates a "Top Performing Report" for the ads using the "Relatability" hook tactic. He names it "Top Hook Style" and clicks "Analyze this report".] Josh Bampton: We can create a report to show us all of those relatability style hooks, and we can use the system and reverse engineer it backwards. And we can do that through the use of the AI, right? So once we have our AI analysis...
20:56 - 21:06 > [VISUAL: Screen recording of a report titled "Exploring relatability hooks for diverse visuals". It has a "TL;DR" section and a "What's Working" section.] Josh Bampton: ...we can then ask the system, "Okay, how can I run relatability style hooks for multiple different visual formats?"
21:06 - 21:22 > [VISUAL: Screen recording of the "Next Steps" section. It says "Want me to generate 5-10 new hook concepts that lean into this 'everyone says you should do X, but here's the messy reality' relatability angle—specifically calling out other AI tools, creative testing advice, or reporting struggles your audience is definitely hearing but not seeing results from?"] Josh Bampton: And what we now have is the other way around, right? So we've got multiple different visuals that we could run for the same hook. And it sounds crazy, right? I'm just going to jump back to the slides for a second.
21:22 - 22:22 > [VISUAL: Slide titled "Fyxer analogy". It shows a grid of images. The top row is "Hooks" with images labeled "Young Adults", "Busy Moms", "Blue Collar Workers", "Senior Citizens". The middle row is "Visuals" with images labeled "Good vs Bad", "High Quality", "Latest Trend", "Latest Trend", "Limited Time Offer", "Financial Savings". The bottom row is "Messaging" with images of cowboy boots labeled "Visual Appeal", "Performance & Durability", "Family-Friendly", "Financial Savings".] Josh Bampton: But I was talking with Fyxer a little while ago, and I don't think they'll mind me saying this. Guy or Lucy, if you're on the call, shout out to those guys, they're legends. They make fantastic ads. But we were talking about the processes they follow for generating new ideas. Because they have extremely high volumes of ads, right? And I'm like, "Guys, how do you keep on top of it in such a small team?" And honestly, what Guy described, I can only share with you as like a Batman villain style wall, right? With Post-it notes everywhere. And each Post-it note would be a persona or a hook or a messaging angle or something that they have found that is successful for them. And what they're then doing is they're pinning those combinations to new variants they think could work. And the reason I wanted to share that is that what we are trying to achieve is to lay everything down—hooks, visuals, messaging, personas—and identify those combinations and lean into them to increase our success across our new ads. So I wanted to share it just so that everyone doesn't think I'm completely crazy.
22:22 - 22:27 Josh Bampton: But just to jump back into the system.
22:27 - 23:53 > [VISUAL: Slide with large text "84%" and smaller text "Of Marketers say that AI will give teams a competitive edge over the next year." Below is a shoutout: "Shoutout Kosta the genius, and the incredible report - Winners, Losers, and the Next A Decade in Creative Strategy. Thumbstop Quarterly. Motion Creative Analytics. 2025"] Josh Bampton: Once we have those new ideas, we then need to think about how we execute them, right? And visual formats are so important here. Because we know if we're not visually different, then our ads are not going to get in front of new audiences. So we can use this report to help us identify what's already working. And you know, the big standout for me here is split screen ads, right? And I'm sure Carissa is more than aware of this and probably making loads of split screen ads at the moment. But for whatever reason, the engagement metrics are solid here. So whatever is happening in this style of ad is very successful. And once we know that information, the question then becomes, "Okay, cool. Well, could I use a relatability hook here on a split screen style ad to try and lean into those trends?" And again, it's being able to keep on top of all of those different combinations, high volumes of testing, and then we iterate from that. So you know, everything that we've talked about so far is very much high level, and it's very much around the different variables and the different combinations that will work for you. The other thing that is always going to be important is knowing what is working within the individual ads themselves. And we know that Meta is saying we need 50 optimization events in order for our ad to get out of learning. And you know, if we're running hundreds of ads, that's more difficult than it seems, right? But what we can still do is configure reports in Motion, similar to this, that are looking at all of our new launches, and we can overlay different metrics here to identify what we think is working even within those early stage ads. So even if they don't skyrocket and become our top performers, we can still take learnings from this, right? And you know, personally, when I look at numbers too long, it gives me a bit of a headache. So what I like to do is turn on color formatting in here, and this is going to show us where we have above average and below average performance. And the real thing I'm looking for whenever I look at this is what kind of patterns are here. So you know, where do we have similarities? And you know, one thing that's interesting about this particular report is four out of seven of these ads have successful hold rate and click-through rates. And the question then becomes, "Okay, is that because the messaging angle is good, but the hook is maybe not quite tailored enough to who we needed it to go to?" But that's an interesting piece of information that we can further investigate, right? So we can click into these ads, we can use the AI to help us, we can ask the system to analyze the ad itself to generate its thoughts on what's working and what's not. But we can also use the performance data in here to help us generate ideas ourselves around placements, demographics, and how they're interacting with our content. Because of course, this is also really useful information.
23:53 - 24:10 > [VISUAL: Slide titled "Where should I be focusing?" with bullet points: "Trend Insights via AI Tag reports", "Top Performers (by Persona, Messaging, Visual and Hook)", "Early Stage Potentials", "Iteration opportunities", "Inspo", "Analyse this report, Analyse this ad, Tell me what this brand is testing, Spot Gaps and Opportunities in this brand.", "Use the AI outputs to create new ideas"] Josh Bampton: The other thing that's always important, and I've mentioned it a few times already, is iterations, right? So when we're looking at our new ads, I want to know if I have standout performance in certain creative areas. Because alongside knowing the trends in the account, this is a great way to keep on top of new emerging trends, right? So you know, when we're looking at a report like this, we can start to see where do we have strong performance and where do we have weak performance. And what we can also do is we can use our new compare functionality to compare the performance with the previous period. So now if I do this, the system is going to show me the difference in the performance of the ads compared to where it was two weeks ago. And this is really cool. Because what this enables us to do is identify what's scaling. Shout out Tara, by the way, because this ad is awesome. So we can see that this ad is doing really well. It's growing. It's getting a lot of good budget. Click-through rate is increasing. But we can also see we've got a couple indications here where things are starting to decline. And you know, the question then becomes, "Okay, is it declining because it's just not good enough? Or is it declining because it's been up and it's fatiguing and coming down?" And we need to be able to get ahead of that, right? So by using that comparison functionality, it enables us to see the changes in our data. And once we have this, this is really useful information. So if I'd said to Ed, "Look, I love this ad. Tara's ad is killing it at the moment, but the click-through rate's just not quite where I want it to be." I can ping this straight over to Ed. If I click into Slack, I can pick a channel that I know Ed's in, and I can say, "Dude, we've got to fix the click-through rate here. What can we do?" And that will go straight through to Ed, and he'll then be able to prioritize it. If we include the AI insights here, that's even better because it's going to provide a summary of what's going on. But if we continue running with that example, if we click into Tara's ad here, come into creative insights, and ask the system to analyze it for us, we don't have to completely change the whole ad, right? Like we know the engagement here is really good. It's scaling, it's doing really well. I'm just wondering if I overlaid a different CTA, whether that might increase the performance. And questions like that that are much more bespoke, that will also generate new ideas for us relevant to that. So to show you what I mean, if we come back to this analysis of Tara's ad, we can see where the ad is from a metric perspective, and we can see why the ad thinks, or why the AI, sorry, thinks this ad is performing in the way that it is. And as mentioned, I don't want to change the whole ad. I don't want to completely scrap it and run a new concept. All I want to do is try and improve the click-through rate. So what can I do to just tweak perhaps the back end of the video to increase the CTA performance? And the system will also generate those ideas for us, right? So we can keep an eye on the trends within the account. We can also keep an eye on individual ad performance to see what's going up and what's coming down to enable us to get ahead of that. I've got one final thing in the system to share. The other thing that we haven't talked about yet is we can also use the inspo page to help us see what other brands are doing. And this is so important, right? Knowing what your competition is doing, seeing what other brands are doing that make cool ads can help influence us to make more visually diverse content to get in front of new people. Because ultimately, we need to be Andromeda friendly, right? And those larger brands like Fyxer, for example, are very good at churning out high volumes of ads that are all different. So we, through the use of inspo, we can see all of the, all brands that we followed, right? So we can search for brands in here. We can see all of the ads within their account. And to take it a step further, we can start to build collections of what we like from other brands. And there are some really fun updates coming in this page. I'm not going to give it away too much, but keep an eye on it because I'm really excited for that to come out. One thing that we can do from here is we can again, we can utilize the system to help us make those new ideas for us. So if I'm looking at Huel, and I think, you know, it's a completely different industry. They're not selling cowboy boots, but I'm sure I could get some cool ideas from here. I can use the system to summarize for me what they're doing at the moment. And what I mean by that is the system is then going to summarize what they're doing from a testing perspective, what it looks like they're testing at the moment, what looks to be working, and also what looks to not be working. And then I love this bit. It's a little bit sneaky, but it's also going to tell us what we could steal from here. Now, it's going to be difficult for me to steal meal stories considering I'm a cowboy boot brand, but quick and easy UGC reviews is a good idea, right? But what we can do again, coming back to new concepts, is we can use the system to generate those new concepts for us based on other brands. And again, we now have brand new concept ideas that are completely different to what we've tried before, based on a different brand's ad portfolio. So just to summarize, guys, and I appreciate I've gone through a lot of information today. The real critical parts of Motion that you need to be keeping an eye on are your trend reports through AI tagging. So, you know, knowing who our personas are, knowing how we can branch out across different messaging angles, and then using hooks and visuals to build those new diverse stories. And then keeping on top of what's working and not working within your own portfolio. So, just to jump back to the slides for a second then. And I appreciate you bearing with me, guys. Just a couple of things just to finish on. Last year in Q4, Kosta, who is a genius in our team, he released a really interesting report. And there was two things in that report that really hit home for me. The first one is this stat, right? 84% of marketers think that AI is going to give them a competitive edge this year. But the other part that I think is really interesting is this quote from Brennan, right? Is that you can use generative AI to make crappy cowboy adverts like I've made. And, you know, that will come out and you can use that. But if you don't provide the strategy and the process behind the ads that you need, those ads are never going to be successful. And the reason I found that so relevant is that it relates perfectly to Motion, right? Because all of our system is built off of the creative strategy knowledge that we have and also your own ad account's performance. So all of the rigor and the thinking, as Brennan mentioned, has already been input into the system. And I can't hit home enough how time-saving that can be. Some of the customers I speak with, especially the new customers, when I'm showing them how they can do that, they're like, "Oh my god, I spent hours doing this." And it's like, "Dude, I know." So just one last thing then, guys. I have got a folder of reports. The support team might shoot me for saying this, but if you want a copy of the reports, drop your CSM a message, give support a message in your channel, and we'll duplicate them into your account for you. And one final thing.
24:10 - 24:19 > [VISUAL: Slide showing a picture of a man in a cowboy hat, denim shirt, and cowboy boots holding a lasso. Text: "To be a cowboy, you must think like a cowboy. Yeehaw."] Josh Bampton: If you want to buy some cowboy boots, give me a shout because I'm going into cowboy boot business.
24:19 - 24:20 Josh Bampton: Thanks, guys.
24:20 - 25:31 Ed Inicki: Josh, my man, that was awesome. The, y'all, I first met Josh in Texas last year. I'm pretty sure it was his first time stateside ever. And for him to weave that back into his first event a year later, yeehaw indeed. So great. So great. Thank you, Josh. That was excellent. Uh, y'all, hope you, hope you took a lot away from that. One thing I wanted to call out that I love that Josh talks about, he calls Motion "the system" all the time when he's jumping into the app. I don't know if that's a UK thing or if it's just the way your brain is tuned to this, Josh. Like, thinking about how do we connect all of these aspects of doing organic brand research, pulling that into our creative brainstorming sessions, pulling that into production, and all of that weaving into a system. But Motion is more and more becoming that for a lot of people. And I think you did an excellent job showcasing all those different touch points of the process that lean into this. So thank you so much. Um, you grabbed a glass of water, I hope. That was a lot of talking. But I want to jump into Q&A, uh, with you if you're ready for that.
25:31 - 25:32 Josh Bampton: Let's go.
25:32 - 25:57 Ed Inicki: Sweet. Okay. One thing I saw, um, I apologize, I forget who asked this in chat, but Josh, would love for you to make the distinction again for everybody of where you're talking about making big creative changes for Andromeda kind of at that top line, and then how that feeds into the iteration side. Because I think that might get a little confusing for people when were talking about creative diversity, but we still talk about the importance of iterations.
25:57 - 26:03 Josh Bampton: Yeah, it's a really good question. It's a really good call out. Let me jump into the system again for a second.
26:03 - 26:16 > [VISUAL: Screen recording of the Motion app. Josh is on a "Top Iteration Opportunities" report, showing a table of ads with metrics like Impressions, Hook score, Match score, and Click score.] Josh Bampton: So the beauty of Motion, and one of my favorite things about Motion now especially, is how we can use the different data to influence different things, right? So to give you an example of that, if we come back to our visual formats...
26:16 - 26:44 > [VISUAL: Screen recording of the Motion app. Josh switches to the "Visual Formats" report, showing the bar chart for "Split Screen", "Greenscreen", etc.] Josh Bampton: ...for us at the moment, split screens are performing really well. Obviously, you know, the spend is probably lower here. It's not scaled quite as significantly. But that to me suggests that that could be a good thing to test more of. So what we can then do, when we look at minor iterations or ads that are showing signs of potential but they're not perfect...
26:44 - 27:08 > [VISUAL: Screen recording of the Motion app. Josh switches back to the "Top Iteration Opportunities" report.] Josh Bampton: ...is we can use that prior knowledge of, "Okay, I know I want to do more split screen ads. What can I potentially iterate on or what can I take learnings from and iterate in that new format?" And you know, split screen might not be the best example because it's quite a dramatic change. But if we were looking at a creative like this one as an example...
27:08 - 27:13 > [VISUAL: Screen recording of the Motion app. Josh highlights a specific ad in the "Top Iteration Opportunities" report where Impressions and Hook score are low, but Match score and Click score are high.] Josh Bampton: ...where the impressions and hook rate aren't great, but the messaging is obviously good here because the hold rate and click-through rate is high. So my mind then goes to, "Okay, would it be possible to iterate on this ad, but lean into the trends within the account?"
27:13 - 27:25 > [VISUAL: Screen recording of the Motion app. Josh switches back to the "Visual Formats" report.] Josh Bampton: So do it as a split screen or even a green screen, something that we know is working for us at this moment in time, and generate that new iteration idea.
27:25 - 27:34 > [VISUAL: Screen recording of the Motion app. Josh switches back to the "Top Iteration Opportunities" report.] Josh Bampton: And again, we can use the AI to help us actually come up with the concept. If we analyze it, we can tell the system we want it in a certain visual, and that will give us that new concept.
27:34 - 28:22 Ed Inicki: Cool. So you're blending reports and blending creatives together that way, rather than having like a, "Hey, this ad did well, let's iterate on it," and just brainstorming from there. Yeah. I mean, that's still important though. It's worth noting that is still important. If you see an ad that is really good, how can we create new variations of it? Obviously, it does need to be visually different, so we still need to lean into some of those trends. But if we see something that's successful, we want to make new variations of it. Awesome. Love it. Okay, Josh, we're going to keep hammering through. Folks, lots of great questions. Thank you so much for the engagement on that. We're going to go by, um, ranking them. So if you jump over to Q&A and you upvote the questions you really want to see us get to, that would be super, super helpful. Um, I'm going to pull one up here for you, Josh. You can stop sharing screen too. But talk about, talk about this approach here, Kyle's approach here, and how you would critique this, maybe slight tweaks you would think about related to this.
28:22 - 28:26 > [VISUAL: Q&A overlay appears on screen: "Kyle Wedemeyer: Is the approach to: (1) test multiple messaging angles within a single ad design, (2) identify the best-performing message, and (3) produce multiple creative variations featuring that winning message?"] Ed Inicki: Yeah, a good question. So let me just make sure I understand. We test multiple messaging angles within a single visual format, identify the best performing message, and produce multiple creative variations featuring that winning message.
28:26 - 28:35 Ed Inicki: Yeah, definitely. I mean, we need to know how we want to communicate with the persona that we're targeting, right?
28:35 - 29:39 Josh Bampton: So if we are testing, let's say, 10 different messaging angles to one persona, the likelihood is not all 10 of them are going to succeed. And that's okay. That's fine. That's a learning for us. But if we have three of them that are good, two of them that are okay, and five of them suck, great. Let's scrap the five that suck. We're not going to do that. The two that are okay, how do we need to tweak it? You know, is there anything that we could change within it to increase the performance? And the three that work, awesome. How can I make new visually different ads for those three messaging angles that are successful? So yeah, Kyle, you're absolutely on the right track there with that.
29:39 - 29:42 Ed Inicki: Love it. Love it. Okay. I got one, this will be a good one.
29:42 - 29:45 > [VISUAL: Q&A overlay appears on screen: "Meagan Kay: Do you have insight into how Andromeda identifies personas? Is it age/gender/interest? Or is it more story telling as Motion is showing?"] Ed Inicki: Let's talk about how Andromeda, and again, Josh and I don't work for Meta, so we have to make a bit of a, take a bit of our best guess at this. But Josh, from what you've seen, can you talk about the personas that your customers are seeing in Meta versus how they show up in Motion?
29:45 - 30:02 Ed Inicki: It's a good question. And again, I will just caveat this that I don't work for Meta. My understanding is that it's based on creative signals from the end user in Meta, right? So what is that end user interacting with? Like everything from impressions to watching content to scrolling away from, what are they interacting with?
30:02 - 31:01 Josh Bampton: And then those interactions then become almost like a group, right, of people that have those similar interests. And when we upload ads into Meta now, they're grouping entities by ID. Meaning if you have ads that are all the same, they're going to be grouped together as one entity ID, and that will then be showed to the persona they've deemed as most relevant or receptive to that style of ad. Love it. It's a tough question though. I wish I had a better answer for you. That's all good. That's all good. Great one. Okay, bring this one up from Michelle. I thought this was really great. Static versus video. We have a lot more creative metrics and engagement metrics we can think about on video versus static. Um, yeah, what are your thoughts on that one, Josh?
31:01 - 31:05 > [VISUAL: Q&A overlay appears on screen: "Michelle de Maat: How, if at all, does your approach change with statics vs. videos. Creative metrics on a static tell us much less on hook / engagement & ultimately conversion than a video does. Which might make it trickier to figure out exactly what was working?"] Ed Inicki: Yeah, it's a good question. And static analysis is always going to be more difficult than video analysis, for sure.
31:05 - 31:16 Ed Inicki: We will still be able to identify trends. And the other thing that Motion can help you do is when you look at a headline report via AI tags, what that actually means, it's a little bit confusing, but it's not actually the headline of the ad, it's the text within the ad itself.
31:16 - 32:28 Josh Bampton: So for static analysis, that text variation trend input can be really useful to determine, okay, what sort of message resonates through the static. Because really when we talk about images, what we really mean by that is just the messaging angle, right? Sometimes the design and the color can be important, and we can use, you know, naming conventions to build those types of reports. Although hopefully the team's working on that, sneak peek. Fingers crossed. It's a little bit more difficult with static, but I would definitely recommend headline analysis, persona analysis, messaging analysis, because you can also filter messaging analysis by image, right? So we can still see all the messages we're portraying for statics only. Love it. Love it. Okay. One about, and I think this is one you see a lot, Josh, is how do you, what do you suggest for creative strategists and paid social teams working better on this? You know, where do paid social teams need to adjust in order to like make these audience conversations work a little bit better? What are your thoughts here?
32:28 - 32:31 > [VISUAL: Q&A overlay appears on screen: "Hannah Johnston: How would you suggest creative strategist and paid social team work together to achieve this? Does paid social media team need to step off the gas when it comes to audience targeting for the campaigns for this to work?"] Ed Inicki: Yeah, it's a really good question. And I saw some, when I was doing my homework ready for this event, I was reading up about Andromeda and like what people are talking about in the industry.
32:31 - 32:48 Ed Inicki: And one thing that became more and more apparent is that, you know, Meta eventually wants to phase out some of those old media buying tricks in order to target people. And what that means for paid media teams is that there needs to be a deeper understanding of why creative is so important now and the pillars of success behind those creatives.
32:48 - 33:47 Josh Bampton: So if we're going to the paid team and saying, "Look, I'm going to run this ad, it's a bit obscure, but this is the reason why and this is the data behind it that validates that decision." So for me, when I'm training new paid media teams, the one thing that we talk about a lot is why creative is so important and why the data in Motion, or sorry, how the data in Motion validates decisions for us. Love it. Okay, I'm just scanning through. We got a bunch of questions here. Um, also shout out Michelle, who was the last question. Very good. Okay. Here's one. Gemma, I believe I'm pronouncing it correctly. Don't hate on me if I got it wrong. Uh, so different brands have very different personas. When you're thinking about inspo, Josh, how are you thinking about going in and analyzing inspo to match a persona base?
33:47 - 33:55 Ed Inicki: I'm going to answer this question at the risk of potentially being told off by the product team. But we're working on some really interesting developments within our own inspo page that would allow for filtering and searching of different variables.
33:55 - 33:58 > [VISUAL: Q&A overlay appears on screen: "Gemma Sayer: Different brands will likely have very different personas. Is their a way in the platform to analyse inspo similar to your current persona base?"] Ed Inicki: Um, so, you know, rather than just searching for brands, we could say, "Okay, show me any ads that target cowboys." And obviously it would then return that. And we can also do that for visual formats. Kira, please don't shout at me for saying that.
33:58 - 34:15 Ed Inicki: The other way in which I would do that, the more sort of historic way, when I work with new customers who aren't sure about their personas and brands, ChatGPT will do a great job of providing you a list of businesses that target a similar target audience to you.
34:15 - 35:26 Josh Bampton: And once we have that target audience, we can come into inspo, follow all those different brands, and then we can see all of the ads they're running to get in front of a similar audience to us. Why don't you pull that one up, Josh, really quick? You could screen share that and we could, and we could talk through running some of the AI tasks in there too. Because if people haven't seen those, I think that would be really helpful. Yeah, absolutely. I don't know how I steal the screen though, Ed. It won't let me share while you've got this up. Yeah, give me one sec. Okay, go for it. I felt like such an old man there scrolling to find the stop share button. So if we're in the system and we have a list of new brands that target a similar audience to ourselves, we can then come into here, search for that brand, find that brand, and we can then see all of the ads that that brand has in their ads library.
35:26 - 35:31 Ed Inicki: And this is going to include active and inactive ads. And it's also going to give us the ability to filter this page on a couple of different variables.
35:31 - 36:50 > [VISUAL: Screen recording of the Motion app. Josh is on the "Discover" page, searching for brands. He clicks on "Fyxer" and views their ads. He uses the "Status" filter to select "Active" and the "Days active" filter to select "More than 3 months".] Josh Bampton: So for me, a little trick that I like to do is use this days active filter to say, "Show me ads that have been out for more than three months." Because obviously anyone in this industry knows you're not going to have an ad out for that long unless it's relatively successful, right? So what we now know is that these ads are some of their long-standing, likely successful ads. Obviously, you know, in the ads library, we don't have access to their performance data, but you're not going to have an ad out for that period of time unless it's good. And what we can then do to help us generate new ideas is at the top here, the system is going to suggest a couple of different tasks. So if we were to look at gaps and opportunities based on this brand, that is going to pull out a summary of what they're doing, what they're not doing, and how we could position ourselves to be in competition with them. Or testing is similar to the output I showed earlier, which is going to show us exactly what they're doing, what looks to be tested and what looks to not be tested, and what's working and what's not. So getting that list of personas, getting it into Motion, and then using the AI to help us generate new ideas based on what they're already doing.
36:50 - 36:53 Ed Inicki: Love it. Love it. Thank you for walking through that, Josh.
36:53 - 37:26 Josh Bampton: Um, okay. I got one more here. This is a good one from Samuel. So I think you mentioned this at one point in the persona section. You have some customers that are running ad groups based on personas now. Can you talk about that a little bit more? Yeah. So it's a really interesting one. And some of the bigger customers have already been doing it, but I'm also noticing now that new customers coming through are also doing it, which is an interesting sign that people are moving into this Andromeda model, right?
37:26 - 37:29 > [VISUAL: Q&A overlay appears on screen: "Samuel Mangialavori: You mentioned clients running ad groups based on personas now: can you please elaborate on that?"] Ed Inicki: And what I mean by that, one of the customers I spoke with probably last week, maybe the week before, they've configured their accounts, so you've got an evergreen campaign, a testing campaign, and within both of those, you've got multiple ad sets that are targeting multiple personas.
37:29 - 37:35 Ed Inicki: So for example, if you've got nine personas, you've got nine different ad sets, and you can then use those ad sets to be really fluid, right?
37:35 - 38:51 Josh Bampton: I know I want to target this persona. I'm going to chuck in loads of different ads that I think are going to work for this persona. When I identify what does, it moves into the other campaign, same ad set, and then you repeat the process. It's a lot more fluid than the traditional kind of testing top of the funnel, middle of the funnel, bottom of the funnel. I'm not saying funnel analysis isn't important, it definitely is, but I think being a little bit more fluid and less rigid around how we configure that can only be a good thing in light of the fact we don't have quite as many targeting powers anymore. Love it. Okay. Let me see what else that we haven't... Oh, this is a good one. Um, so we get to talk about another Meta thing that we don't 100% know how it works. Uh, from Denise, Dennis. Um, you mentioned that within Andromeda, we should be designing creative for AI, Josh. That kind of big headline here of AI and Meta is going to see our ads first. But how do you, how do you see that fitting in with the GEM updates? Those came out in November. People have probably heard this as a bit of a buzzword as well. And talking about what you've seen, the conversations you've had with customers specifically to GEM.
38:51 - 38:59 Ed Inicki: It's interesting. And I have to be completely honest, I haven't specifically dived into the sort of algorithm itself within Meta.
38:59 - 39:03 > [VISUAL: Q&A overlay appears on screen: "Denis Grigas: You mentioned that within Andromeda, we should be designing creative for AI rather than people. How does this also fit in with the GEM updates that were rolled out in November, specifically regarding its ability to process multimodal data and diverse creative formats to power intent based user journeys?"] Ed Inicki: But my understanding here is that Meta itself, it wants to have more control over our creatives, right? So creative formats that are diversified is essentially what it wants.
39:03 - 39:31 Ed Inicki: And what it's asking for from a lot of customers is more control around flexi ads as well, right? So when you put an ad in, let Meta run 100 iterations of that ad in different formats. Do I know if that's a good thing or not? I'm not sure. And have we seen success from that yet? Again, I'm not 100% sure.
39:31 - 40:43 Josh Bampton: But it is really important that we have an understanding of how that AI perceives the content because that is ultimately the first thing that's going to see our ad. And we can't change, once that AI has made that decision, we can't change that without changing the ad itself. So, you know, having an understanding of what it thinks and then being able to decide, "Okay, I want to move away from that or I want to move into that," is really important. Love it. Love it. Okay. One about, and I think this is one you see a lot, Josh, is any plans to implement TikTok, YouTube? Product team, if you're here, read Sarah's question. We need to get on this. I agree, Sarah. Thank you for calling it out. We know TikTok and YouTube need a bit of love. And we're excited to, we're excited to hopefully have something to share with you soon. But I'm going to put the pressure on for you on this one. Josh, what would you add there?
40:43 - 40:45 Ed Inicki: Just keep shouting about it, Sarah.
40:45 - 40:48 > [VISUAL: Q&A overlay appears on screen: "Sarah Hamilton: Any plans to implement TikTok/Youtube in Motion?"] Ed Inicki: Yeah. Totally. Keep shouting about it. It gives us more ammunition to go to the product team. We know we need to focus on those areas. We're just trying to get that planned into the roadmap at the moment.
40:48 - 41:07 Ed Inicki: Absolutely. Okay. This one came up. I don't know if you've seen this much, Josh, yourself. Something that's showing up when somebody's seeing "None" related to it. William, there might be a few reasons depending on how old the creative is. It might not have gotten a tag applied to it because we did have, when we first rolled it out, it was only in the last 90 days with spend. So that could be something.
41:07 - 41:21 Josh Bampton: Um, Josh, I don't know if you've seen it for new creative popping up and how that would work out for a customer? So this actually happens, it happens quite a lot, right? And I caught up with Alicia who helped the team design AI tags and she talked me through this.
41:21 - 41:24 > [VISUAL: Q&A overlay appears on screen: "William Jones: What would you recommend for a portion of ads that show as "None" for personas?"] Ed Inicki: If it's been labeled as "None", that means that that creative doesn't have a distinct persona that it's targeting. So to give you a good example of that, where I see this most commonly are in catalog style ads or like very retail focused brands where they're just portraying like a video of a jumper.
41:24 - 41:50 Ed Inicki: Like that doesn't necessarily have a clear persona. So those ads will be labeled as "None" because there isn't a persona that's clearly able to be perceived. And whenever I talk about that, I always say that's a really interesting thing to know because in light of the new Meta update, really we need to be able to answer that question. So if we do have ads that are coming out as "None", what can we change to make sure they are targeting someone?
41:50 - 42:48 Josh Bampton: Love it. Okay. I think we're going to run up on time here pretty shortly, but see if we can knock out a few more of these, Josh. This is a good one from Michelle. Um, talk about, and maybe a bit more broadly than this, how would you combine, Josh, teams that are using like pretty robust naming conventions with AI tags? I know we've talked about this in the past of using your AI tagging as a bit of a gut check to see is our AI tagging matching up with what your team visualizes or identifies in your naming convention. Because maybe you think you're speaking to an intended audience and you put that in your naming convention, but Motion's reading that as a different audience. Yeah. I think that's a really good point because one of my favorite things around naming and AI tags is either validation or opportunity, right?
42:48 - 42:52 Ed Inicki: Because if we have named it that we want to target wannabe cowboys, and then we see an AI tag that is similar, we know we're doing the right thing.
42:52 - 42:56 > [VISUAL: Q&A overlay appears on screen: "Michelle Murphy: If we wanted to manually create these reports using naming conventions instead of AI tags. Can we still use this messaging feature to create concepts etc?"] Ed Inicki: But, you know, nine times out of 10, there's probably going to be some nuances, there's probably going to be some differences. And one of the conversations that we have a lot, because, you know, AI tags are so great, is how important is naming conventions?
42:56 - 43:27 Ed Inicki: And, you know, in my opinion, I would say naming is always important. Having an understanding of the ad is critical. Totally. But what we really need is some of that more unique data that we don't currently have. So like influencers, concepts, products, things that, you know, AI tags isn't picking up. Because we can then filter our AI tag reports by those variables. So not only could we say, "Show me the messaging angles I'm running for this persona," we could say, "Show me the messaging angles I'm running for this influencer and this product through the use of naming conventions."
43:27 - 44:37 Josh Bampton: So they work hand in hand in my opinion. Love it. Love it. Okay. I think we're going to call it there just because we're running up on time and I want to be respectful of everybody's calendar. Folks, I did see a few other questions we weren't able to get to. Um, if you have more questions, please reach out to us. You can hit [email protected]. You can reach out to us in chat. If you've got Slack with us, fire questions to us there. Really appreciate all the engagement. 350 of you are still hanging out with us and I really appreciate that. We're stoked. I know this is early morning for some folks on our West Coast, so thank you so much for joining us. If it's late where you are, thank you for joining us. If it's the middle of your day, I hope we gave you some juice going through the rest of your day and got you excited about some things. Um, just wanted to call out again, Josh said it, uh, the team's probably going to get on us about this, but if you are looking to add the reports that Josh highlighted in today's session and you want to get those added, reach out to us in chat, hit us up over email, hit us up in Slack. We'll get those for you. Might not be able to do it instantly, but we'll see how big the list is. Thank you so much. And Josh, I just want to give you final word to sign off as you close out your Motion event debut.
44:37 - 45:53 Ed Inicki: Yeah, thank you so much, guys. Thank you for listening. It's been super fun. Like Ed said, give us a shout if you want the reports. Give us a shout if you have any questions. Super happy to help. But yeah, it's been great.
45:53 - 46:04 Josh Bampton: [No speech]
46:04 - 46:07 > [VISUAL: Motion logo animation on a black background.]
46:07 - 46:35 > [VISUAL: Motion promo video. Text on screen: "Ship more winning creative". Shows a grid of various video ads. Then shows the Motion app interface with a "Sprints" report. Then shows individual ads with tags like "Unicorn", "Top clicked", "UGC Success", "Try new hook", "Fix ending", "Improve CTA", "Try new offer". Text on screen: "Join 2,100+ teams shipping winning ads with Motion" followed by logos of brands like Vuori, True Classic, Hexclad, Jones Road, Mud\Wtr, MuteSix, Ridge, Wpromote, Power. Text on screen: "Book a demo for a VIP tour". Ends with the Motion logo and "motionapp.com".] [No speech]