Meta Algorithm Update: How to Adjust Your Organic and Paid Media Strategy

0:00I think that whole bringing organic into the training algorithm is really interesting and something that's not really been surfaced.

0:12Welcome back everybody to another episode. Excited to dive into a few different things on this episode. I think we're going to start with some more big, big algorithm updates from Meta. Have you been using Twitter a lot recently? to be honest like just because of how busy things have been not not as much as I would like to we've seen I've just can't articles everywhere yeah it's gone from like including yourself including myself which is which is what we're going to touch on first but it seems to have gone from like 180 characters to long form essays when when did that happen but it's the it's like such a hack if you try to grow your Twitter audience throw out a few articles um pick a specific topic go deep on it for like a thousand words and the funny thing is everyone's just saving them and I'm 100 % confident no one's reading them it's just like yeah you've been racking up the saves it's like mental masturbation everyone just like saves everything it's like oh I'll come back to that later I can be guilty of that I'm terrible for it yeah do you think that spans outside of DTC Twitter or do you think in terms of like Twitter shifting towards this like more long form practicals yeah yeah way of consuming information I think they're using that as like a real big signal for like to push to push reach like that the one I wrote on um UTIS and GM which we're going to come on to like Meta's the next phase of Andromeda it's got like 40 ,000 views but the engagement on it is not even that good it's just got loads of saves um which isn't a huge amount of views but um it is for like a a relatively simple post and not like a resource giveaway way but yeah i'm not sure i'm not sure if it maybe works the same in sports and other things maybe maybe not but definitely in areas like tech and yeah business yeah um so yeah we can start there we've also got another article that i've shared which was sarah levin just deep dive into sarah yeah a really good deep dive into something we've been thinking about and talking about over the last few episodes so we'll touch on that but yeah i wanted to start with metas gem and utis updates as you've just called it I'll put the article I'm referencing in the description because it's a much longer form read, much more detail includes a few screenshots and resources and then another video that we touched on that we did in depth about creative variety which I think this is another it's that next evolution and it continues to crystallize this idea that creative is a targeting, that's the fundamental underpinning outcome of this it's just further crystallized um that creative is the way that meta delivers creative but it's also the way it learns about your brand and your customer um so what it means in is i'm gonna have to read some of this because i'm still trying to wrap my head around a lot of the terminology but meta has changed how it decides which brands deserve like reach um if you think about the system what the system previously rewarded a lot of it was engagement so if you think of a lot of those ads that really drove low cpm it was often the ones where you you'd get a lot of shares comments likes the the engagement on the ad would would drive better auction price and better reach um and it's what this this this update has brought is like this shift towards like an understanding of the quality of alignment with interests of the individual at the other end of the other end of the algorithm because they even came out in there and said that you know the the previous way was flawed and that i think they they had under 50 precision in terms of alignment to to who you're targeting yeah which is like bit of like stab in the dark right it's like they they didn't have a high match rate and therefore they've they were very fundamentally kind of rewarding the wrong outcome in in the auction so what they've brought in is this user true interest survey on the organic side um and i think what what it's kind of doing in the back what what it's asking is like how well does this content match your interest on a scale of one to five so it's literally a survey collected daily um and that is creating a new perception layer a new training vehicle for four meters algorithm on that organic content so reels being the main format have you seen this pop up before i haven't actually i've not actually personally seen it for those listening it's it's literally just a form that says to what extent does this video match your interest from not at all to a great deal yeah i haven't seen it i'm I don't really use, try not to use it still.

5:15I'm still off socials at the moment. Just love talking about it. Quite a juxtaposition, actually. What I have seen, just a bit of a segue, is have you seen the big, the notification that rolled out across the UK this week about paying for Instagram? Yes, yes, yes. Is it$4.99? $4.99 for no ads. For no ads, yeah. We haven't got any, we pushed our meta reps on expectations and data on that survey. so what do they expect Upcake take to be on people paying to have no ads? Their expectation was it was going to be, they couldn't share exacts. They've tested it in a couple of markets. The reason they're doing it is in response to EU privacy laws, I believe, to build a case around future changes.

6:04I'm surprised they hadn't done it sooner, to be honest, just if you look at YouTube Premium. Yeah. well they expect I can't remember the platform the reference when they were talking to us about it but like low single digits maybe even less which we hope for because then that won't impact people's performance because I do feel like Facebook ads and meta ads are quite non-intrusive compared to like YouTube pre-roll YouTube, Netflix, they're very unskippable etc so yeah hoping that plays out so yeah that's come out But yeah, this survey is, the result of this is that view of meta training its algorithm on what interest people have, what content resonates with that interest and moving away from engagement driven and therefore like general, generic, viral content that appeals to everybody.

7:04so broad hooks trends towards niche high quality content that speaks directly to people even if that's not a huge volume of people being surfaced in the algorithm more um from that they've seen big jumps in engagement big jumps in interest matching um but yeah i think that that whole bringing organic into the training algorithm is really interesting and something that's not really been surfaced like we weren't aware of that if that was done previously it wasn't clear how that maybe impacted both paid performance and your overall ability to leverage meta as a channel it's like the theory um aligns very similarly to that of paid you know you're looking at creative attributes as well like format style persona context which is exactly what we would do from a paid perspective and as a result you're thinking through those layers and thinking more intentionally versus what's going to get you know how we're going to drive as much reach as possible by going more generic yeah exactly and that's what the second half of the gm the generative expansion model kind of does from a paid perspective we'll come Come on to that in a second.

8:24But I think this Utis organic adjustment is a big shift in, brings about a big shift in organic strategy for brands where maybe they've been doing a lot of trend jacking, broad, relatable humor, like everybody will like this Hux or just even baiting watch time, which I think has been a trend at times just to generate huge reach and awareness and maybe following. I remember they're like used to do organic trends where if it was a video for three seconds and you just had the big text block and it would, like you say, hack reach and engagement because it would obviously, Meta would push it further out in the algorithm if someone replayed a video.

9:14Yeah, exactly. So it's like, but doing those things is like training Meta to think that your brand is for everybody. Exactly. And we know that the fundamental concept of marketing is like, if you're advertising to everybody, you're advertising to no one. So it's like, there is no direct statement from Meta about that system impacting paid performance. But a hypothesis would be that you've got organic here, paid here. I think those things are going to converge over time. Do we think it's possible that the understanding of your customer in an organic setting could support better delivery of your ads i think that's a why wouldn't they try and bridge that gap but they haven't said that so i can't say it's 100 happening but yeah i think it brings into focus like organic if that is the case and if we can sort of prove that and if we can see that playing out hopefully a lot of brands are able to that have done a lot of work on on community thinking about free soul thinking about um pure sport some of the some supplement brands off top of my head that have got a really identity driven um like cult-like following on socials like we should this should really allow them to even like multiply that impact from organic creative and resonate with that audience even more which i think is really interesting yeah exactly it's like you you're leaning more into like motivators behaviors which in turn is essentially what makes up a persona, right?

10:45So that then is going to likely propel. 100%. I think it's registering the topic of the ad, but just final point, but it's also ingesting audio, mood, style, motivation, production quality. So all of those signals into how it finds niche audiences or how it just delivers your content to more people. It's exactly what we spoke about in that creative diversity mastermind, like soulmate theory. There's a perfect piece of creative out there for everyone. It's like that next evolution of soulmate theory, moving on to organic. And I think it brings a lot of the creative strategy research process we've spoken about.

11:33We've always spoke about this for a paid context, but if I was a brand watching and you want to try and test out this new model and try and maybe improve your organic resonance, it's like it starts with messaging, it starts with persona, and then you can get creative about the ideas. I'm not an organic content expert, so I'm not going to go down that route. But I think the messaging and the persona, it's like go deep on that and try and maybe bridge the paid and organic teams together a bit more from that perspective and from a learnings and feedback loop perspective. Yeah. Yeah. I mean, it makes me think we used to have an organic service.

12:09That was actually, you know, since our inception, that was a big thing that we did before we moved over to paid. And the whole thing there was like, find the formats that work and double down. You know, diversity isn't necessary once you really find what works across maybe four or five different formats that you treat it like a TV channel and you run those same formats. Whereas what this is telling us through the lens of creative diversity and soulmate theory um from the perspective or through that perspective of organic is you know even if you're speaking directly to a specific persona with a specific angle in an organic creative you need to be running that against lots of different vehicles to ensure to to connect that person with the with their perfect creative yeah yeah and we've seen a lot of organic we've seen people who who believe in the concept of like test content through like you can do that test reels thing where you distribute reels to like five iterations of the same piece of content to a new audience then use that as a learning bed for paid.

13:11We've obviously seen a lot of examples where great organic translates to paid maybe we'll see that go the other way definitely believe we could in terms of pulling maybe not the entire structure and framework but certainly some of the messages certain some of the personas. It's why I just think this like makes complete sense because for the last two years we're I just feel like a lot of our big swing winners that create step change in an account actually come from organic inspiration. There's an organic video or messaging vehicle that has resonated with a certain persona or a certain keyword that relates to a persona.

13:48Yeah, 100%. So yeah, I saw Cody from Jones Road saying that one of his big bets for 2026 is organic and as a channel and bringing that closer to his paid ecosystem. So you've seen a lot of the big brands taking big swings in this area and it'll be interesting to see how that plays out. Switching on to the GEM, which is the Generative Expansion Model, a lot of acronyms. Marketing loves acronyms, right? What GEM does is rather than looking at isolated user actions as a source of signals for paid delivery, of like what to match a certain individual with it's looking at behavior across instagram facebook reels whatsapps and across all pixeled websites and connecting that to both organic and paid interaction so the question that it asks in in theory is like does this brand's creative consistently align with what this person actually cares about which again i think is just that like persona pain point piece right it's like what who is this person what's their like what's the problem we're trying to solve in their their life what's the failed solutions they've maybe experienced how is that showing up on like a deep emotional level it's another like it's another source of signals and learning and move towards just deep customer understanding and deeply and using that to deliver creative to those individuals.

15:25I feel like this quote here captures it really well. Organic is no longer distribution. It's the training date of how meta understands your brand's relevance. Yeah, 100%. And I think this GM paid rewards against specific narrow language, so not broad, clear use cases of a product that resonate with that persona, context-rich demonstrations and then consistent storytelling from across the brand. So not just maybe on organic and on paid. And all of that is just going to ladder up to better signals and better relevancy. And then if you are able to lean into creative to maximize that better delivery and better results.

16:08So yeah, I think it's really interesting to see. Obviously, we had Andromeda and we've had the creative similarity wave. It's going to be interesting to see how these then roll out and from the data and from some of the proof points that Meta provide, the results seem to be really, really strong and it's just going to make delivery even more efficient for brands. 100%. And I think this is why we need to be thinking about diversity as something that is as important as volume, if not more. Because I think we were very heavy on volume and I think a lot of people when they hear Andromeda think creative volume but I think Meta doesn't necessarily want more ads it wants more distinct interpretations of who that product is for and how we need to speak to them and how we deliver that to them.

17:03Exactly. And I think we've been guilty at times last year of going too far down the volume route for certain brands and not having the right broad coverage of persona, angle, vehicle. And we've learned that. And that is strategy, right? That's good versus great strategy. How we look at being more intentional across those different building blocks of creative. I also think that when you do that well, it becomes a compounding engine because the diversity drives the delivery, which drives the signals, which allows you, and it's like the signals of new audience allow you to just constantly expand your reach and avoid being limited by first-time impression ratio over time.

17:46I think this is why it's just super, super important that if you're spending, say, over 100k a month, it's less important with the smaller budgets, but the importance of going much deeper with your persona mapping. And I think people look at personas as we have this one persona that relates to this one demographic or this one type of person. Well, that's not enough really, because although you and I may actually fit under the same persona, we've got different desires, core beliefs and fears. And as a result, how someone speaks to me is going to be different to how they need to speak to you. So actually, we would both be sub-personas of a macro, right?

18:31So that's why we're looking at it through that lens now. like for every one of our brands we will have macro personas and then at least three micro personas that are tapping into different primary desires core fears and beliefs as well yeah and i think you pull that process we talk we focus all on creative but you pull that through the customer journey as well like each one of those sub personas should be spoken about differently from a messaging perspective an angle perspective they'll all have the ability to work across multiple vehicles, that should be pulled through to the landing page, which should then have the same fears, common objections, beliefs, just that and building that congruency is where I think the brands, like we did a Kind Patches breakdown, where they are really good at persona-driven, sub-persona-driven messaging and they've got such diversity.

19:24I think Loop were very good at that as well as a brand who were very early mover to volume and diversity. It's super important. Yeah, 100%. I think that's why, you know, when you're thinking, we talk a lot about paid here, we don't really talk about organic, but I think even if you are going to be doubling down on that organic strategy this year, previously, like when we were building an organic strategy, you know, similarly to how we would set funnel positions in a creative strategy on paid, we would set slightly different setup in terms of like how we would name it, but it would be entertain at the top, inspire, educate, convert.

20:03But then you set your splits through that lens. One thing we didn't do but we would do now is the same with personas, right? And then you're layering your organic creative in a way like you would approach creative strategy on a paid level. 100%. I think another thing we've started to see Meta do over the last few months put a lot more time and tech, like they're just rolling out loads of features in this creative marketplace, creator marketplace. So they rolled one out today, I think at the time of recording this, which literally just pulls out of your, all, pulls from your followers, all of your, all, all your, pulls from your followers, all of the creators who follow you.

20:48So influencers, people who are posting high volume of content on, on Instagram. to go and leverage. But I think we're seeing a big focus on creating marketplace and we're seeing a huge, huge, huge impact from partnership ads still. And we've talked about this on previous episodes. We'll put below a refreshed partnership ads guide that I built over the weekend for people to have a look at. But I think partnership, I think about partnership as like the multiplier on creative diversity. You've got the building blocks and you've got like this multiplier effect of partnership ads identity because it allows you to just do even more volume across all of those areas.

21:38and the thing that makes it again it plays into signals like new running ads through a third party identity allows you to leverage the signals from that person's pre-built audience that is going to be very different to your brand and that drives new and over over indexes on net new impressions relative to if you've got a lot of content coming for your own brand already and I think a really good benchmark to look for for partnership ad spend is like 20-30 % in an account I'd say most accounts that we audit are sub 10 % last 12 months and I think just bringing that up like 10 % towards 20 % of account spend whilst maintaining quality is just such a huge unlock and yeah I think Mets is obviously pushing it really hard this year it's one of the key objectives our reps are giving us as an agency like getting partnership ads up to a certain level of spend across our whole portfolio.

22:42And you know when Meta's trying to push something It's driving results. It's not just driving results but it's almost like this OP thing. Yeah, it's like first mover advantage. Yeah, exactly. It's similar to what we were talking about with Twitter earlier. They're going to push it more in the algorithm. You want to get onto a feature, yeah, for sure. So it'll get to a point where everybody's numb to it and it just doesn't create that level of impact. But I think there's a real nice window right now and there has been for two or three months of if you can do it well, it will drive a big change in ability to spend.

23:26I would say the biggest area there, and obviously across all demos it's super important, as you've just said but i think the biggest opportunity there lies in like 50 plus people because i think it's the in terms of like partnership ads it's the type of it's the persona that do it the least um just naturally younger people and more inclined to to to agree to that but secondly older audiences are the least naturally the least trusting and require the most trust building as a result, which as we know, this is something that really helps support that. Yeah, 100%. I haven't actually, thinking about it, I haven't seen a huge volume of partnership ads in that demo.

24:14I have seen quite a few, I wanted to come on to a few quick wins across this area, but one of them is like third party and persona driven pages. So creating a page that's, and creating the identity yourself, bit grey heart. I actually told Meta about this in a meeting with our rep, so they definitely didn't approve, but we won't. We're a bit quiet in the room and I think as soon as Ollie said it, we were like, maybe shouldn't have said that. It's a little bit, yeah. It's a bit like DTC grey heart, but it's like identify a persona or niche that matters to your ideal customer, create an Instagram and Facebook page built around that persona um build out the page and relevant content so like don't you don't need to be super extensive but just populate it and then align that with the the ads that are being driven through that persona so supplements for focus could be adhd life hacks baby products could be first time mums dog food could be kitchen critic menopause supplements could be menopausal mums and it's and it again it's resonance it's a multiplier it'll probably help with trust it'll help with with with delivery i think another couple founders ads like so simple every every founder ad pair it with a with a founder's page their personal instagram if they're open to it but on private i think you can even and they have to change it to a professional account so you can't uh yeah founders founders leveraging your team so i think internal team members that's whether that's a creator or whether that's like staff members um like doing bts interviewing people about the favorite products in the business like just leverage them for delivery through third party pages obviously taking top performing ugc creators and putting them on retainers i also think looking at like your influencer seeding program finding like the top 10 percent of content quality across that getting on whitelisting maybe even looking at customer base um you can use things like refundal for activating customers and getting whitelisting rights looking at tiktok shop uh top tiktok shop creators if you're on there asking them if you can pull the content and their identity across to meta doesn't have to be not that complex but massive massive leverage the one that I've just seen work really well into and it's kind of becomes that creator slash creative flywheel is like tiktok shop like even with creators if you can bake that into contracts where if they if they get top spenders with you to then look at moving that creative to meta and doing it as partnership ads yeah I've seen I've seen a lot of brands like build those relationships at scale 100 it's not that expensive either like for a lot of tiktok shop creators you pay over 100 quid for a 30 day whitelisting.

27:10I've actually got to get in the guide there's some negotiation tactics because it's a bit of a sales exercise convincing some people of getting to allow you to whitelist their creative, especially if they've got more followers and you start to move up that chain. But yeah, I think if you're a brand watching, take trailing six months spend, segment out how much of that has come through non-brand page, so not your core identity, if it's under 5 % big red flag if it's under 10 % between 5 % and 10 % like you're in a you're probably above most right now I'd say you want to be trying to get up towards 20 % of spend maybe beyond if you start to see that really work the only caveat to it is it's not a it's not a magic bullet so creative is still the key variable just slapping partnership ads on everything isn't going to work so you've got to do it with intention and the creative starts to be really good.

28:10Yeah. Yeah. I think that's the two big meta updates at the moment. Yeah. Wow. Very, very in-depth. Full of value. Back to Twitter articles, I guess. If that's where you want to go next. Yeah, I think it's super interesting. We'll link all of this down below, but there's a really interesting article written by Sarah Levenger who really is an expert when it comes to D2C psychology and behavior um so she's she's super super she looks at creative strategy in a in a in a slightly different way to most and she's essentially built a guide as to like the different levels of creative strategy which i think is is really really interesting and almost what you need to be to reach each different level yeah very gamified pokemon evolution i think that was her theme right exactly so you've got level one creative strategist which is the foundation builder and i just think she's actually the way she's described it is actually perfect from us working so closely with strategists at every one of these levels like i i think the way she's personified it is is great yeah i sent it across to you and i was like this is just weird timing because we're having so many conversations in our in the business right now about like how do we codify what great strategy looks like and this was a really strong articulate codification of of the of the progression framework for for cs yeah 100 well i want to get like i want to get to level three first because that's what we that's the north star right where we should all be aiming for but i guess to give context on level one firstly the level one creative strategist typically has six to 18 months of experience within the role they're learning the language of performance marketing and beginning to understand that creative isn't just about aesthetic, but also conversion.

30:04So essentially, it's like very basic platform literacy, fundamental understanding of frameworks, basic copywriting, fundamentals of key metrics, et cetera, et cetera. And what she does really interestingly is she goes into like the tech stack and how that evolves at each level, which is very true really, and how you can start to systemize your role as a creative strategist and create the environment in which you should be working with him. Yeah. um we get to level two which she calls the pattern decoder so the level two strategists have moved beyond limitation into interpretation they've seen enough campaigns live and die that they're starting to understand why things work not just that they work so this is all around like being able to actually hypothesize why something is working and understanding the strategic elements and layers as to what drove success like pattern recognition as well across contexts right whether that's in an account and across campaigns across personas or across winning assets losing assets or even across different businesses commonality in in inform like in various building blocks exactly and i think with ai you can actually now get to level two quicker because it is just advanced pattern recognition in in many ways yeah i've actually been doing quite a lot of that this week i was doing um i built a process around uh again so starting with the persona because we document everything right it's like persona uploading top ads across that persona doing pattern recognition and gap analysis and then using that to review congruency of message on ad to landing page and how to like drive that closer together but use just gemini for all of it and it is really good at extracting patterns in a way that you probably would miss yourself in some ways yeah it's multi-layered isn't it it's difficult um but yeah i think you know this at the level two you're thinking more through psychological frameworks advanced data interpretation obviously you're able to understand like testing um testing methods, platform translations, etc, etc.

32:19And I think, as I say, the level two is probably where we see a lot of strategists sitting. But I think what is really interesting and what we've been speaking a lot about before you'd sent me this article is, is that idea of, you know, a great creative strategist moving from really being a driver of executional throughput to the strategic spine in the process and i think where ai is going that is going to be essential and if you aren't able to understand pattern recognition psychology which we'll get into in level three strategy but also like moving from actually executing within a process to building the system around you building the creative strategy operating system i also think another layer of that is like your a better awareness of your position within a broader commercial process so like when when does a business need to be creating higher volumes of assets when does return on assets spend like yeah how to think about like forecast and the commercial side of return on investment in creative process at a business level yeah and i think this is actually why like claude using creating claude skills is actually an amazing exercise to actually allow you to reflect on logic and reasoning because like you know you sometimes you you will do something in an ad account you're not you don't really think about it in the moment but you just have the instinct to be able to know what to do because you built up pattern recognition to some degree whereas if you actually reflect on okay so if if x happens then i do y yeah if you had to contextualize that as to the the actual reasoning for it is it intuition or is there is there is it a science yeah i think that's where a lot of the that's that's the hard bit of building like internal ai tooling as well is like pulling that context graph that decision making framework out of your head because i find it quite hard to articulate your own actions sometimes especially when it's in that like nuanced area yeah and that's where we are right now like we are we're going through an intelligent project where we are thinking through that lens and i think a lot of the things that you might think just are based on intuition what we're actually starting to do with our creative strategy team and sam and i had a call about this today is starting to create diary entries of almost why we made certain decisions or did certain things as a creative strategist um because usually that's stuff that isn't necessarily documented whereas if that is you you would say on an isolated case that that is down to intuition of just having the context, having gone through the failure of being in enough ad accounts and reviewing enough creative.

35:11Whereas actually on a macro level, if you have 200 of those use cases, there's pattern recognition there that can be tied together as part of that context engine. Yeah, 100%. It's hard. It is hard to, that's a really interesting way to think about documenting that, like diary entries. Makes a lot sense. Yeah, 100%. Doing it in the moment rather than trying to reflect. It's harder to reflect, isn't it? Yeah. So I think that's something, that's a behavior that we're starting to get our team to execute on essentially and start thinking through. But yeah, anyway, getting to level three, creative strategist, the strategic architect.

35:49So essentially at this level, they've expanded their skills beyond execution and optimization into building sophisticated systems predicting trends and creating frameworks that others will follow they don't just respond to data they understand the market forces cultural currents and psychological patterns that will shape performance so i love this idea of like creating your ecosystem creating the operating system that you you work within the environment in which you work within which is super interesting so the competencies here fall into like actually building proprietary frameworks which again is something that we are we are really super super bullish on um you know predictive creative strategy actually building the system in which we work within cultural trend forecasting and this is kind of in in many ways like documenting that context right or that that data layer that we can sit within.

36:54Yeah. I think that's where we try to take, and I think that's where everybody should be trying to take their creative resource, the people involved in that process. Because I think it kind of codifies the way in which AI is going to sort of erode some of that value creation up to, and compress that value creation up towards people who are really good at level three, and automate away some of the things at maybe level one. that it is currently doing, such as just building swipe files, data collection, 100 % in a context library. That's the, yeah. And I think if you're a founder, if you're a creative strategist, whoever you are watching, listening to this, really start to download your mind into context documents.

37:42And we're going to give some examples of where and why you will need to do that. But you've essentially got two different types of context. You've got domain and you've got brand. So what sits under domain is external learnings and insights, which I know you do a lot of this. You might watch an hour-long YouTube video, transcribe it, put that into a context document, tweet, tweets, blogs, whatever it might be, actually PDFs of books. This is the brains behind how we're able to understand how we actually enter the point of theory. Yeah, a lot of that is really useful. a lot of that's really useful in every all of your interactions with AIs like that context engineering from like telling it to use some of the biggest advertisers like Ogilvy breakthrough advertising like some of those big seminal like texts and reference those in the process works here but you can also do that in other areas like business strategy, management, referencing certain people's frameworks.

38:54So I'd encourage people to do that across the board. You also don't actually have to have access to the resource all the time. Sometimes it can just retrieve it without having to upload it, which is very valuable. Yeah, so you've got that, which is one half of domain context, and then the other half is what we were talking about, internal knowledge uploading. It's like IP, really. Exactly, it is IP, yeah. And then the other side of context is brand context. So that's your guidelines, you know, do's and don'ts. I guess like a lot of the data behind what's working for your brand or not working. It could be beliefs, processes, etc, etc.

39:30But I think the importance of getting there, I guess, which feeds into this like level three strategist, especially when I'm like, you know, what we're starting to do with Claude Skills and Claude Cowork is this idea of moving from prompt engineering to where we maybe were last year which is context engineering to where we're going to be this year which is workflow engineering so essentially you know we are set we're laying the foundations for it to engineer a workflow for to achieve a certain objective yeah i think we're going to do a mastermind on claude code claude skills claude code aren't you yeah i say we lucas is going to do a mastermind perhaps with sam that we've we've run with our team but that's definitely where we're heading and then it's moving that from there into our own proprietary proprietary system uh but the documentation is the the boring foundational part that you just kind of have to get through and the sooner you start the better yeah in that process yeah 100 100%.

40:41So how are you, talk to me, I know you've been using Claude Code a lot. Talk to me about some of the stuff that you've been doing there. Or like a use case. Claude a lot in general. I feel like Claude's having its moment, right? They're absolutely ripping new tools, new features, new integrations, like the recent Excel update, Claude Codework, Claude Code. they've just released a connector where it can interface with like Slack Figma etc and I think a lot of the getting more value out of it is actually just the simple things like hosting it locally on your computer first so you get the reason Cloud Code is powerful I think I mentioned this in a previous episode is it's really good at tool calling and it can go and create if it needs to access something it'll write an API and go get the data for you so you could but I think the thing that people struggle with is when you put Cloud Code onto your computer it's in like a terminal interface and it's really like if you're non-technical it can be a bit intimidating you can get quite good wrappers which is what Cowork is now Cowork is the Cloud Code wrapper that makes it just a bit more UX UI friendly and a bit more enjoyable for non-technicals so you can either use Cowork I prefer to still use Claude Code with a third-party wrapper on top, because I think it's just still better.

42:11Claude Code works a bit buggy and slow at times. I've been using that for just crazy long pieces of work. I wrote the partnership guide, which we'll share down below. I wrote in about 20 minutes. I asked it to go and interface with all of our Google, Notion MetaGox, basically one prompt one big macro prompt I asked it to condense information write in a certain tone and then produce this document and it just went from top to bottom in like 20 minutes but it codifies referencing a lot of our context, a lot of online context, it's really powerful at sequential steps sometimes it'll spin up a 40-50 step process to get to an outcome.

42:59So it's very agentic rather than like single input-output question-answer. I guess the one thing I've used it for most recently is we're moving offices

43:13and obviously it can code. It's ridiculously powerful at code-based output. I think that's what AI is really good at right now. It's really good at driving more value across anything that is written, like anything that requires, if a higher volume of written output will produce better value creation, AI is really good at producing that because that's what it's really good at. Yeah. So I built a big, we're moving offices, so it was like, well, maybe we could build our own like office management app. So for like attendance, replenishment, meeting room bookings, connect it with Slack, connect it with Google Meets, birthday notification, social calendar reporting any issues people or things and I did plan mode built a plan, it took a while to like you can go into a thing called planning mode it helps you work through the plan produces a plan and then you just press accept and continue and then I left it overnight and it coded for like a few hours and then I got up and it just built an iOS app perfectly.

44:30Like perfect, didn't have to change anything. Fully functional. I needed to add a few. I then did another planning mode round and add like a better onboarding sequence. I need to brand it up, but I reckon that would have taken someone. I have no idea what I'm doing with coding, so I have to do a few debugging, etc. But that would have taken a team quite a while to build less than six months ago. Yeah. And cost a few quid as well. And cost a lot of money. But I think that's what I really am enjoying thinking about is not doing any huge projects. It's just like, what are those niggly little things in your life or in work that a tool like that would help with?

45:07That's an obvious one. It's like having an auto name. I built an auto name generator in there every week for those people who are in the office. Someone's going to get picked to do the bins, which I think is just like novelty, but it's like... yeah yeah it's just it just makes that solve the problem right um built a forecasting app for for tiktok shop because that was taking us like two hours to build a forecast and i built a application you put stuff in and it spits it out and you can export it those are the things that i think cloud code can facilitate in the business you don't what you don't want to do though is get lost creating things that drive no value because it's very start with the outcome yeah it's super like real enjoyable it's really fun that's the thing you can you can go down rabbit holes yeah it got to about midnight last night and i was just so deep in um yeah co-working it was just yeah it got a bit out of hand should we say i've seen so many people talking about that though like um on twitter where they're like i was up to like 5 a.m just coding it's like so i know it's like playing video games when you're like 14 it literally is yeah that's sort of dope for me you should just leave the screen to do what it needs to do but you just stare at it just watch it so yeah i think it's the the thing i would say about like if you're going to spend a lot of time on are you going to go down that like the thing i've learned is a lot of it's all around having a really good plan an idea of what you want to build and articulating it yeah very very well because small changes in the way you articulate something can create a very different outcome.