Evolving Creative Strategy To Grow To 9 Figures

Here's the thing. When brands start to scale, the creative floodgates open. Suddenly, you're cranking out more ads, more ideas, and more variations than ever before. But with all that growth comes a new problem, creative chaos.

What worked at six figures can break at eight. And keeping strategy aligned across teams, channels, and campaigns becomes a real full-time job. That's where our next speaker comes in. Oi is co-founder of Soul with Us, the performance agency behind some of the fastest growing DTC brands.

They've helped generate over 400 million in revenue by bringing creative strategy and performance under one roof. Today, Oie's breaking down how to evolve your creative strategy as you grow, keeping structure, momentum, and performance as your operation scales. So, if your brand has outpaced your creative process, this one is for you. So, drop some fire emojis in the chat and give a big creative con welcome to Ollie Hudson.

Hello everyone. Thanks for sticking with us so far. Um, I'm Ollie from Saw Group and I'm going to be going through evolving creative strategy to grow to nine figures. As mentioned, I'm the founder of of a group of agencies working out of the UK.

Um, we're Meta's leading independent agency, fastest growing over the last 12 months, working with a ton of the market's biggest brands and helping them make hundreds of assets per month. I think um going into November we'll be making 5 to 7,000 unique ads. So yeah, huge amount of volume um across a really wide range of categories. And I'm wanting to spend today talking through um how we are leaning into the most recent meta updates in order to ensure that we can continue to scale brands on platform who are spending multi6 seven figures a month without seeing CPA and CAC increase.

As we're all aware, social media feeds have changed a lot over the last 3 to 5 years. Um, we've gone from a world of social delivery, social graph, passive discovery, feeds based on those who you follow and interact with on a regular basis to intentbased delivery, interest graphs, intent modeling, um, a deeper understanding of personal level preference across these platforms and the ability for them to deliver uh, content that resonates with your interests, your content uh, format preferences, your past purchasing and future purch purchasing uh desires and yeah really that that shift in discovery as a result of that and I feel like when you look forward um the world is moving towards this AIdriven hyperpersonalization ecosystem so agentic uh deep machine learning deep understanding of your you as an individual and how to serve you the best possible experience to maximize your time on platform and therefore the amount of um revenue that these these social channels can generate from you as a user. More recently, we've seen paid as a channel and as a as a as an algorithm on Meta change, too. Um we've been around through multiple different iterations of of Meta's ecosystem.

Um, we we we started in in late 2017 back in the days of running 52 sometimes up to a thousand different lookalike audiences in an account. super granulated retargeting, very little creative production relative to now through to iOS 14 through to the rise of ASC and then into 2025 we've seen the launch and roll out of Meta Andromeda. What is Andromeda? Um Andromeda is is a new algorithmic approach to ad delivery within Meta's ecosystem.

So, this somewhat ugly screenshot on screen ultimately is a machine learning and matching system that helps Meta better serve its ad candidates. The, uh, ad inventory we upload into the platform to its user requests, the requests from people scrolling on Instagram, Facebook, uh, etc. in a way that maximizes action rates for advertisers and and ultimately drives better results. So what does this mean?

Um Meta is really leaning and we we are really leaning as a provider and you should be as advertisers into this idea that creative andor creator slash the identity that the ad is sitting on is the targeting. Meta in theory can now support 5,000 ads a week, not 50. It is far more sophisticated in ad delivery. It can it can handle a much greater volume of uploads without seeing diminishing return on performance and effectively rank and deliver those ads to the right individuals.

The algorithm has shifted from deterministic to probabilistic. What that means is far less the algorithm is is less based on a if then um way of working. So, uh, demographic based delivery and more based on a ranking method of reviewing assets, um, reviewing all of the possible, uh, delivery combinations for that asset within the context of its platform and finding the right individuals to serve that in front of. So, far more sophisticated way of of delivering assets.

There's a lot of articles online that are going to explain the difference between those two things better than I just did there. But I don't want to dwell on that because I want to drive into like how we can translate this update and how we can capitalize on this update to separate and drive far better returns for the brands that are really leaning into this way of thinking. Um, Soulmate theory. This this game breeds this idea that there is actually the a perfect ad for every user.

Um in in an ideal world, it's just about creating enough highquality creative to appeal to every individual on the platform. And ultimately, we need to supply the algorithm with volume and variety. The theory is if we 10x output for a brand whilst retaining quality and diversity, it will result in greater performance and and scalability. It's all about how do we get there?

More recently, as in literally in the last seven days, we've seen updated guidance from Meta on creative similarity. So, this screenshot is from a creative session we ran with ran with our reps. Um, and it shows how the algorithm is viewing creative uh differentiation. So on the left we've got a similar ad with a similar message and a similar visual style and a small change in copy.

And on the right we've got a diff an ad that's got a different differentiated message and a differentiated style. So very visually different maybe a different angle, a different persona. And what this is saying is that for you in order to maximize Meta's delivery, reach new audience consistently and and drive consistent results and efficient results as you push more spend through their system. Really thinking and keeping creative similarity and differentiation in mind is is super important.

If we're producing a lot of ads on the left hand side of this spe of this spectrum, we're going to find those ads are grouped together. They are they are bucketed. They are layered up into the same entity ID and they are delivered to the same audience as a result. And you can see how if you're doing a lot of um incremental iteration that was maybe best practice at the start of this year or the middle of 2024, this will be limiting reach over time and therefore it's likely that your CAC will be going up over time as you try to spend more through Facebook.

Um it's a really common thing we see is is we on board brands that have been making more and more creative. they've been trying to spend more on platform but we're we're not able to do that without seeing an increase in CPA because we are not effectively differentiating our assets and therefore we're not reaching new audience to build on this this is a real world example from a client of ours space goods so this is a report we got from meta recently and it's a it's a review of which of our assets in our ASC campaigns are identified as identical or highly similar. So you can see these assets on the bottom right which I think are most interesting. These two assets are actually two different creators delivering a quite a different script.

Similar persona, a painoint and angle, but they have been grouped by Andromeda is highly similar despite them being different individuals, despite them being dressed differently. Um, and you can see that I think this is a real proof point that that that differentiation is important. What this is saying is these two assets are going to be bucketed to the same audience. We're going to deliver those to the same individuals and therefore we're not getting the incremental reach we want from introducing new assets into the ecosystem if they're being grouped together in this way.

So yeah, creative similarity for us is is a real bottleneck for for brands trying to spend multi7 multi6 seven figures on platform and it's something that you need to keep in mind when reviewing your creative strategy process when reviewing your output on a monthly basis. The way we look at this from a metric perspective is through the lens of incremental reach and first time impression ratio. So first time impression ratio is ultimately the the number the total percentage of impressions you are getting on meta in a given period that are truly incremental or new. So this graph on the left the blue bar um the bottom at the bottom is previously reached individuals.

The turquoisey green bar above that is incremental reach in that given period. Here we're looking at months and then the rolling percentage orange number is the percentage of incremental reach um across our account. So on the right we're we're we're reaching 13% of our impressions in a given month are are truly incremental or going to people who have never seen the brand before. When you're spending multiple six figures a month and above, it's super important to keep this in mind.

We use this data to inform our media planning tactics. Um, we want to be keeping this between 10 and 20%, it's never going to be 50%. Not every previously reached impression is a negative. Obviously, it can take two to five impressions to convert someone.

However, we see this as a leading indicator of our ability to grow an account over time. If this is below benchmark, the first tactic we look at for a brand is creative diversity, messaging, format, angle, and creator, etc. And only when we've exhausted these options and we feel like we've got a real strong process and consistency around creative diversity for a for a brand, do we look at non- conversion objectives. And this this re this um report on the left is is is one you can pull from your meta reps.

You can get it on a rolling weekly or monthly basis, but it's really important to be keeping incremental reach at the forefront of your process as a brand within Meta. Um and ultimately we see it uh we see these two leading indicators um as as the key metrics to review uh that provides a a proof point of declining incremental reach over time. So firstly you you generate the report I shared. Secondly it's about looking at CPMR.

So the absolute cost to reach a thousand accounts on meta. So CPM times frequency. Many accounts may have really strong CPM, but their frequency can be going up over time because they're getting an increasing amount of creative recognized as similar, therefore being delivered to the same audience, similar in message, similar in format, similar in visuals, etc., like I've just touched on. So we want to be keeping C CPMR at the forefront at an account level like how we how we tracking this over time what does this look like on a monthly basis but also on a campaign level and even on an asset level if we start introducing new presentations or new angles often as you find things that successfully unlock new audiences you'll see big shifts in CPM and frequency it'll underindex compared to your account average which is a positive Secondly is is new visitor percentage.

So the absolute percentage of people who click through an ad that is that are new to you as a as a brand. Um again you can review this on an account campaign ad set and ad level. Um and I think CPMR is your ability to get in front of new audience. New visitor percentage is the ability to to is is a measure of how effective those ads are at generating an outcome or traffic for your brand.

So you want to keep both of these in mind when looking at combating and and keeping an eye on incremental reach over time through Facebook. And we see declining first-time impression ratio as a leading leading indicator of increased CPA. So it might not be an issue today, but if we aren't keeping this top of mind, it's it can become a really big issue a month, two, three months down the line if we aren't combating it due to the work that needs to be done in order to generate a step change in incremental reach through meta. And therefore we fundamentally believe that those who are able to create a diverse range of hardworking assets at scale will win in today's ecosystem.

And most advertisers have not adapted to this way of thinking. They've not adapted to this idea of providing meta with a diverse range of content. that diverse range of content picking up signals through engagement, interaction, those signals, providing meta with the context it needs to better deliver those assets to a new audience. And I think if you can keep that way of thinking in mind and and really think through the lens of signal engineering in order to drive incremental reach, you can step you can you can set yourself up for taking advantage of this opportunity that exists right now.

Those who are maximizing firsttime impression rate rate will create a future power law. The impact of focusing on this over time will compound early adopters. We're seeing our top 10% of of brands. It it feels as easy as ever to scale them on meta currently because we're compounding the impact of focusing on the right variables and the right approach to the algorithm.

The algorithm learns from this this this increase in signals, this increase in diversity. you gain more share of voice and ultimately the top 10% of advertisers who really lean into this will capture the majority of impressions and drive the best results for their brand. Now I want to touch on how how do we translate this into a change in our approach to creative strategy and I think we we like to think of creative diversity through the lens of a pyramid. Um, and the base of this pyramid is is is the foundation that can and that's where our biggest swings can come from.

So a change in product, a change in brand positioning ultimately gives us a completely new way to go to market. Above that we've got a persona angle or message. So how can we sell a product to a completely different audience? Take creatine for example.

We could sell that to gym bros or we could send sell that to menopausal women. Both of those are completely different segments of the market and therefore then breed a huge amount of opportunity to diversify format once we've found a winning message. So format being high production static UGC and then vehicle on top of that look we're looking at podcast ad we're looking at VSSL we're looking at founder story how can we take the format and diversify again and then at the very top of this this pyramid we've got those things that a lot of brands have spent time testing and optimizing over the last three five three to five years. So small incremental changes in color, font, even copy that that have been sort of I think pushed down the road a bit in terms of how impactful and how much focus should be put on these.

And we want to be focusing on the 80/20 of creative and we want to be focusing on the things that can provide those step changes into the account and from a from an incremental reach and performance perspective. So we want to be spending a lot of time towards the bottom of this pyramid. We see this as obviously an always on and evolving process. So research, strategy, execution, test, learn, iterate and optimize.

Um the brands that can work this process the hardest are the ones that are getting that compounding um those compounding results and that future power law and driving accelerating returns over time through meta which is still very possible in today's ecosystem. But I wanted to focus heavily initially on research. Um, internally we believe that research is the 80/20 of of creative strategy. It's where we find the biggest unlocks.

It's where we get the best insights and it's where we we really build up our strategic approach to execution. Um, we want to be majoring in strategy and and and really spending a lot of time here on a brand to extract the right learnings and increase our hit rate and success rate and ultimately the impact that our creative has on an account. So, yeah, great performance content comes from great research. I'm going to touch on a few key areas of how you can conduct highquality research for your brand in the pursuit of unlocking net new eyeballs and further accelerated growth through Facebook.

So for we we break this down into a range of areas. We break it down into an ad account audit, creative audit, and um growth audit, which I'm not going to touch on as much. And then we look at research through the lens of competitor, audience, social, and market. And I'm going to touch on each one of those briefly over the next few slides.

So, um I'm sure many of the many of you watching will have will have done a lot of of AI leverage research. I'm sure many of the the sessions on during this um event have touched on this and and I think that audience is is where we really want to start. So on screen right now is a review mining prompt. Uh take a picture of this, download your CSVs, upload your reviews and and prompt to to get a really good good starting point for value propositions, benefits, pain points, desired outcomes.

I think a second building block of this, if you're at scale and you're trying to find that next level or that next angle, that next route to to really driving a step change in um in incremental reach and performance and growth as a result of that is is prompting around like what are the the lesser mentioned benefits maybe what are the benefits or or value or um pain points that appear in less than 5% of the reviews like we can really use the use AI try to to drive us to new um angles and personas if we're prompting it in in the right way. So yeah, starting with audience research, starting with the customer and really looking for those lesser mentioned but uh angles and and benefits and opportunities that that could could lead to us finding a completely new way to take the product to market. We we do a lot of our audience research through automation. So this is our N workflow that scrapes tons of different data points across the internet to pull us pull context together for us to analyze through claude or chat GPT to produce a clear structured output output of research.

However, you can do this manually and I'd recommend you going through um a whole host of different places to to create to create a context doc that you can then analyze with AI. So Reddit Quora, Mumsnet, searching on Google for PR and articles that mention your product or the brand's product, looking at organic comments across all of your channels and building them into a database. Um, and then looking at customer surveys through something like nocommerce to to really ask questions that you can feed into your creative strategy. We want we do this and we we populate an audience flash sheet.

We do that for benefits, for pain points, for features, for angles, and for objections. And this really gets us in the mind of the customer. Again, it's been amazing how how often we've done this for a brand and we found a completely new way to take a product to market. Maybe it's a supplement that's been previously pushed for energy and that it's and people are calling out the benefits of that supplement for their gut um in comparison to a previous failed solution.

And that's provided a completely new way for us to to take the product to market and allowed us to get in front of a completely new audience on meta and drive a step change in results as a as as a as an outcome of that. The result of this for us when we do when we're doing that NA AI research is a is a huge amount of insights and um o like um messaging frameworks that we can then layer into our creative strategy uh pain points, benefits, uh failed solutions, emotional triggers, features, objections, etc. And that becomes the building block for our creative. Another area we're finding a lot of a lot of quick wins at the moment is comment analysis.

Uh Tik Tok recently launched a a tool to support with this. Obviously, we're doing this manually in the previous slides, but we can we can leverage Tik Tok's comment analysis tool if if your brand's active and has a good community on there. And it'll provide a really clear um insight into commonly used terms uh the the the sentiment behind those terms and how we can and extracting insights from from this can again help us help us speak more like the customer and maybe find new ways of taking product to market. Thirdly, we need to be looking at where do your customers sit in the in the in the funnel.

Where where are they with everything with all of the context in mind from from a brand perspective with your uh existing audience uh spend level uh brand journey like where how how large is your spend? How thirdly, we're looking heavily at awareness phases. And we're looking at this both when a brand on boards with us, but we're also periodically reviewing this because the way we approach awareness phases for a brand can develop over time depending on the scale, depending on how we're seeing incremental reach um play out for them on platform as spend is increasing. A really common problem we're finding with brands is we'll on board them, we'll analyze their um creative output and they'll be producing far too much content towards solution, product and most aware towards those lowerfunnel awareness phases that don't as effectively support demand creation and are far more focused on demand capture.

This works really effectively in a market where there's a lot of awareness around a problem. Um there's a lot of awareness around other solutions to that problem maybe other products um be that direct competitors or indirect competitors. However, you reach a point where it becomes difficult to solely do demand capture and we really need to be focused on how do we do education first and selling the product second. An overwhelming majority of our top performing ads over the last 12 months sit in problem aware and product problem unaware segments of the funnel.

I like to use an example for this. So looking at AG1 in 2019, I think we'd all agree that there was very little awareness and education on green powder and its benefits. There was very little market sophistication. Whereas you compare that to now 2025, everybody knows what a green powder is.

there's many players in the market. There's a huge amount of saturation and there's a huge amount of education on the benefits of this solution. Uh the product itself, the mechanism in which it solves your problem. So if I was to look at entering this market today, I'd do it heavily through the lens of demand capture at least in the initial phase of my creative strategy.

I'd be looking at solution and product aware. How do we compete on USPS benefits uh quality? How do we sweep up on like price etc. However, if we were entering the market in 2019, it would have required a lot more demand creation.

So, a lot more focus on problem unaware, problem aware, education before we're driving people to conversion. I think that's a really interesting mental frame because markets can shift over time and your approach to creative strategy and awareness phases need to shift with it as that market sophistication change grows or or develops. So once we've once we've really been intentional about populating our our research databases. So this is an example of a research database for competitors.

We've taken the ads, we've analyzed the angle of those ads, we've analyzed the format, the framework they're using, what funnel stage is it, the hook, and then we're really writing out a hypothesis around why we think that's worked for a brand. And we're doing this with with all of our different um research points to to I think that hypothesis process is really important. We're then breaking that into a a core insights breakdown. So, what are the core personas?

what are the claw demographics, what are the um commonly used creator creator demos, etc. And then we're taking this into an ICE impact, confidence, effort, strategic prioritization pipeline. So from all of our research in each of these buckets of audience, competitor, market, and I forgot the fourth one off the top of my head. um we're we're building out uh our opportunity bank and we're scoring those by priority using that ICE framework.

And that's where our research becomes strategy. That's where we take all of these insights and turn them into what is the best and most important next step for this brand based on where they are today and where they want to get to in 3 months. And this process will give the foundational pillars of your strategy, messaging, formats and angles. Um, and we want to be re reapproaching this process every three months in pursuit of finding new ways to take product to market that supports us in consistently addressing declining incremental reach over time as you scale spend.

And I want to touch on iteration the right way. Um, and how do we add fuel to the fire once we've found something that works? And I think this is the the workflow that in light of Meta's creative similarity changes is shifting the most within our workflow today and within brands. And and I believe this is it's really important to to do this within your business as well.

So it's key still to be testing one variable at a time. However, previously this was very much done through the lens of iterative change. So maybe we're changing the hook, we're changing the text overlay, we're swapping a comment reply for a for a written call out, we're swapping the creator demographic or we're swapping the creative from one fe one female to another. And I would say they are now incremental changes and we want to be focusing on much bigger swings.

So how do we look at this? Firstly, it's angle like we've got problem versus desireled. So, still waking up tired in the more tired than you went to bed. Maybe we can swap that from a problem to a desire.

Imagine waking up actually feeling rested. We could go from rational to emotional. So, clinically tested to reduce wakes ups by 43%. To emotional, I cried the first night I actually slept through.

And you can see these are these are very different ways in to package up the same message and therefore increase the likelihood that you're going to be able to deliver that creative output to a completely new audience and drive that step change in performance that we're looking for. And I would categorize these in that lens of of big swing. Secondly, we've got you can layer on top of this pain points. So, can't fall asleep, waking up in the middle of the night, low energy in the morning, sleep affects everything.

Again, it's like how do we take a overarching pain and package it up in in different ways that resonate with a slightly different persona, pain point, combination. And thirdly is is visual uh iteration. So founders, this founders concept on screen is a perfect example of how we think visual iteration should be done today. The asset in the middle was our first founder story ad for the brand 47 skin.

And we've then iterated on this asset over 25 times whilst retaining performance. I've only brought on screen three examples here. The right is a static, the middle is a is is the initial high production and the left is a repackaged high production with a slightly different um visual format and a slightly different message. Um but you can see how the iteration from a from a format perspective is is very significant here.

Each one of these allowed us to retain what made the first asset work but allowed us to get in front of net new audience and drive performance over time. And this generated over2 million pounds in revenue in under six months through through iteration. And I think this is a re this is how we want to be thinking about iterative output today. We want to be think thinking it through the lens of big changes either in message like I touched on the previous slides or in format and visual presentation that this slide shows.

and a few thought proven formats for you to try in pursuit of of retaining incremental reach and and retaining impression share as you spend more on platform today. I think all of these are proven uh in our accounts as of this quarter. VSSLs, podcasts, founder ads. Um these are very scriptbased.

Like if you get the script right, they will work. I think if you've tried these and they haven't worked for your brand, it's usually an execution problem. AI timeline. So this is walking someone through the timeline of utilizing a pro product um how to go from what to expect over the first week, month, uh 3 months Tik Tok shop style.

So these are those are longer form YouTube uh UGC videos that are that are feel very native on feed uh from and very Tik Tok shopesque. So maybe approaching 45 seconds long. do a lot of education and storytelling before we sell the product. Top of funnel native statics, whiteboard education, and then a new one that we found really recently that's working very well is industry exposed.

So, tons of different ideas here for how you can take a message or an angle and package them up into different formats that will help you reach completely new audience on on meta. So yeah, I think obviously touched on the theory of incremental reach, how to measure it in the account, the research process very very high level but due to time and then showing some new new ways to think about iteration within within this new ecosystem. But I think the key is if you're seeing growth through meta decline or performance weighed over time, we need to be repeating this process of research um reflection and resetting our our approach to creative output with a focus on how do we inject visual diversity and message diversity into the account? How do we find insights that support that?

Um, so we want to be repeating this process on a on a monthly or three monthly basis to ensure we're still getting in front of enough new people to drive day one growth and future growth for a brand. So yeah, thank you for listening to me today. Um, if you're looking for creative support, if you've been seeing declining incremental reach and and you're looking for a partner who can come in and and inject new life, new formats, new approaches into the ad account, please drop me an email at oiesawwiththers.co. I'm sure we can help with that.

But otherwise, I wish everybody a very productive Q4 and thank you for sticking with me throughout this presentation.