SaaS Conversion Alchemy - Using Copy, UX and Decision-Making Psychology to Gain More Customers with Chris Silvestri

Hey, what's up everyone? I'm Mike Sirius and this is Mastering Tech Growth. Every week we sit down with industry leaders to bring you real, no-fluff tips on how to grow faster. Today I got Chris Silvestri on the show. Chris is the founder of Conversion Alchemy. And honestly, he's the guy you call when your messaging is in a mess. Chris helps SaaS founders who know their product is great but can't figure out why people aren't buying. He's all about turning confusing websites into clear, punchy pages that convert. He blends UX, psychology, and conversion copywriting to help SaaS brands finally nail that message. And here's the thing. Chris doesn't just write nice. He digs deep into customer research, pulls out real customer insights, and helps founders figure out who their real buyer is, their ideal customer profile, so the message actually lands. Welcome to the show, Chris.

Hey, Mike. Thanks so much for having me. Glad to be here.

It's going to be an awesome one. People underestimate a good copy of a website when it comes to conversions. So really excited to have this one. Would love to kick off the show with this question, Chris. When I say mastering tech growth, what comes to mind?

Interesting. When I hear the word mastery, mastery for me is, especially this year, is one of my— I always have 3 words for the year, and mastery is a big one this year. I always think in terms of going deeper rather than shallower. And especially nowadays, I think it's important to to always be that human that checks over the tech. Yes, the tech is important, but especially now with AI, it's important to have that human in the loop that has the attention to detail, the, the curation aspect, and also the expertise that tech cannot bring to the table so that you can actually use tech to offer more value.

Okay, I love that. Going deeper instead of shallower. And first thing which came to my mind, Chris, was like, I'm not going— like, I'm not doing deep on anything. I just love learning new stuff. So I love to just tinker with something which I have no idea what that is, get this fundamentals, and I move on to something else. But I love tech. I love helping businesses, especially startups and founders, to actually look at the business and grow that and scale that well. So let me make this question more difficult on you, Chris. So from all the things you can choose when it comes to growth, I'll give you some examples. You got marketing, you know, you have sales funnel, you have people as well, you have timing, you have product market fits, you know, and all these other things. What do you think is number one? And if I would sort of like tell you choose one, like what would be at the top?

The top, probably, as you mentioned at the beginning, is I would definitely invest more in research, which comes before all of that, because I think the understanding part of the equation is what comes before everything. And once you do that well, then all the other pieces fall into place almost automatically.

I love that. So are we saying this is like a customer problem discovery situation here? Yeah.

Like, and also with that, from that, also comes aligning what you know about the customer with what your company does, with what your vision is and what the market also is doing. Right.

Okay. I love that. That actually takes us very nice into this customer research question, which I had. How do you do customer research? You personally, Chris?

Uh, all right. So yeah, I'm a big nerd on customer research. I mean, on research in general. So What I always tell my clients is that 70% of the work I do is going to be research. So the effective copy— yeah, you can write effective copy, but not without the research. And once you do a lot of research and all deep research, again, going into the depth aspect and the mastery, then the copy almost writes itself. And the way that I approach research, I have a pretty structured system that I always try to teach my clients, the companies, the B2B SaaS companies that I work with. And I, I always try to deliver, yes, the research, the insights, the strategy, the copy, but also kind of a system that they can install into their own organizations and then continue doing that research because it's not one like a set, set and done thing. Right. So the way that I see research, I think about it as a— think about it as an iceberg. Iceberg, right? So the part outside of the water, then the biggest part is underneath, is submerged. So the part outside of the water, I see it as the surface layer, which is what your audience is saying, which is all the stuff that typically company companies already have. It's all the materials that you already have. Maybe they have sales recordings. They can go look for reviews online. They can look at competitors' websites for insights as well. So stuff that's pretty easy to retrieve and collect, and that's the surface layer, and that tells you what users say, what kind of language they use. The next layer— so the first layer submerged is the way that I see it. It's kind of the structural layer, which is a component that a lot of companies miss when it comes to writing messaging, which is the experience layer, which tells you what customers or prospects do, whether it's on your website or in your product, the usability aspect. And for that, we typically look at heatmaps, user interaction data, we run usability testing, we look at how competitors structure their pages, how they how they structure the experience of consuming their messaging and all of those. The next and final layer, which is the deeper, deep-seated layer, it tells you why customers act, why they make those decisions that they make, how they make those decisions. And to understand that, you have to really go deep, which is more of the qualitative side of things, which is interviewing customers, looking at competitors' positioning, so the strategy underneath the copy, so deconstructing those, looking, running qualitative surveys, and really extracting the decision-making process. It's not only about asking them, hey, what do you like about this product? What would you want to see? No, it's actually about understanding why they make those decisions to buy from you or to sign up and how they made those decisions along the way. So I see the research in these three layers, and inside each layer. I also divided into internal research, so inside the company, team, or product. Then the external, which is the two sides of prospect and customer, so that you can map that transformation that needs to happen. And the market side of things, which is competitors or, yeah, competitors, messaging, reviews, all of that stuff. So three layers, and inside each layer, those three different facets.

This is an excellent framework to follow. So you're saying the business you work with, you basically, you give them that tool so they effectively would be able to repeat that themselves eventually in the future. Yeah, that's something which is easy to teach, easy to execute. It's, uh, or there is some intrinsic skill which you need to have to actually do that well?

Yeah, so like before even we work together, I start, always start like outlining my process And I always like telling them why I do what I do and the way that I do it so that they can start understanding, okay, this is the process. And once it's done, then we can repeat those same steps to kind of get the same result. And also when it comes to the strategy or the copy, I always like to leave them with some kind of how, how to use this document. So then once they, they get their messaging framework, for example, or their value proposition canvas, which are two messaging strategy documents that I deliver, they know exactly how to take those and write copy from those that, like, without me being there, right? So that's kind of like enabling them to repeat the work that they've done. Also because after all the work that I do, it would be a shame to see them write crappy copy, right?

Yeah, it would be. I agree. Um, so a couple of things, Chris. Uh, what if, what we, what if we just kicking off? So you said the Check some sales calls, you know, potentially some information there. Check online reviews, what people said about your product or service, competition as well. What if I don't have that yet? Like, how do I approach that from -1 to 0 situation? Yeah.

So in that case, I would probably rely more on the internal side of things, as I mentioned. So like your company message and to understand that, you really need to dig deep into your own narrative, your own point of view. I think one of the most powerful positioning strategy, messaging strategies, especially if you're very new, you need to have a strong point of view, a strong narrative that differentiates you from what other companies are doing, especially when you— I typically speak with the founder, but also marketing might have some opinions that have passed on from the market. From the founder, I always like to ask them questions like, what is the— what do you see as being done wrong in the industry? Or what are your customers or competitors doing wrong that you solve in a different, unique way? Or I also like to ask them, what are some myths or lies that have been spread around in the industry and how do you, how do you address them? How do you see them? So these are some of the questions that give me this very differentiated, unique narrative, strategic narrative that then I weave into my messaging. And if you are new, that's probably where I would start. Clarify why you actually built the product that you built, right? You, especially if there's a lot of other competitors, if you came up with that idea, there must be a specific reason, otherwise you wouldn't have done it.

Okay.

So yeah, I would definitely start with that. And then probably the next best thing, especially for differentiation. I would look at competitors' website, but be careful. I wouldn't say copy competitors' website. That's probably— that's— you need to steer clear of copying anything. Also because you don't know whether they're doing the right things or not with their messaging. But there's ways that you can use competitors' website to deconstruct their, their strategy, their thought process in a way. So you can at least get some insights on how they're positioning themselves. What kind of audience they are targeting, they're speaking. So you can see, okay, does it make sense that we speak to the same audience? Maybe we should differentiate in a way. And to look at competitors' website, the easiest process or strategy that I've seen comes from actually a really good website. It's called MacLabs and they have a very solid process for analyzing, messaging, strategy, positioning. It's 3 simple, 3 simple areas. If you look at any page, any one of your competitors' pages, divide, try to divide the page in first motivation. What brings people to the page? What are their desired outcomes? What are their struggles or their purchase prompts? Second section is the value. What are they offering? That's unique, different, or done better? What are their features, for example, or maybe service offering? This, the third section, it's the anxiety section, which is how do they address objections or how do they prove their value through social proof? All of, all of these, when you look at these three areas, it's super easy to understand and deconstruct any marketing copy, any messaging that you find on websites, and then you can actually Take notes, understand what different buckets of competitors exist, and then try to, try to figure out, okay, how do we fit into those buckets and how can we differentiate? I think this is probably a very easy, quick process. So I would definitely say point to strong point of view, looking at competitors and differentiating. And third one, reviews. If you can, if you are in a niche maybe, or there are similar products already, look at competitor reviews and see how your potential customers that are already using competitors, look at how they speak, what kind of language they use, what kind of problems they have. Divide those into themes and then start addressing those, start using those and mirroring that language in your copy.

I love that. That's excellent, Chris. Very good guide to kick off. How you need to sort of think around that process. You mentioned a couple of questions you would ask to get, you know, that beginning happening. What are these, your questions, or are these based on something like— did you come up with them, or is it like some book source over there?

No, I think I come up with them. Maybe there are a couple of inspirations from other experts. There's a lot in my process. I take a lot from April Dunford, a lot from Joanna Wiebe. And so, yeah, I try to kind of mix and match all the things that I see work better for my process. But I'd say I always follow Joanna Wiebe, Copyhackers, April Dunford, the guys at Forget the Funnel have a really good and solid process and really good books as well. And so, yeah, some of those are kind of like mix and match of different things that I see working and I create my own process.

So how, how do you identify customer pain points within that process? Is it mostly looking at the reviews? Because when I think about looking at the copy of the competitors, it's— you, you sort of can backwards engineer the customer pain points they're trying to address. But if I want to go, you know, a level deeper, where do I actually find them? Like, what's the customer sort of research, you know, process to actually discover customer pain points with the niche I want to target?

Yeah, well, reviews is a very good starting point, especially if you go on websites like G2 or Capterra. Like, on G2, you can actually see the questions that they ask. One of the questions was, what are you trying to achieve, or what kind of problems you're trying to solve with this solution? So it literally gives you the framework of problem, benefit, actually motivation. So what brings you here? Benefit that you found value. Also, what were some of the pains, the things that you would see improved? And then finally, the use case as well. So you got a lot of ideas for use cases as well. So reviews and a lot of the companies, especially if you're like starting out, you might be having a lot of sales calls, right? So sales calls, if you record those, those are definitely the first place where I would go to. And that's why with all of my clients, typically I start there even before scheduling customer interviews. I always ask, hey, do you already have recordings of sales calls that I can look at? Obviously they aren't— they're probably— they don't probably follow the set of questions that I ask, but still you can get a lot of good value and you can see, for example, what competitors come up. In their deals. You can see, yeah, what kind of problems they have. So that's a very good one, especially because when you listen to calls and if they have videos as well, you can truly understand the emotional resonance when they say something. You can really understand, okay, well, they— I saw that they reacted in a specific way. So this means that it's really important. Let's note it down. Right. So yeah, qualitative data first, probably. Uh, but yeah, I would say if you're starting out, reviews, probably sales calls. Um, what else? Depending on, depending on the traffic, if you have some traffic, or especially if you're running ads at the beginning to just get some traffic coming, one thing that I like to do, which is typically used more for user experience improvements, uh, but from my background, I find it really helpful, especially to understand the personas and the pain points. It's running website polls or like website surveys. For example, Hotjar lets you run little website surveys that appear on your homepage, for example, or all the pages that you want it to appear. And you can ask simple, quick bullet point questions, actually multiple choice questions to ask, for example, hey, what brings you to the page today? And that could already tell you what the major pain points are. Or which one of these matches your current situation? And you can start asking different options for different personas that you might hypothesize for who they are. So you can start understanding, okay, these are, I don't know, IT people or these are CFOs. To truly understand the 2 or 3 top personas that you should address. And yeah, it was. And then generally expand from those. So when you get that, that is more quantitative data. Once you get there, then look for reviews that have specific themes that match those and try to validate those assumptions.

Okay. I actually like, I love that very much. Any, any discovery which you can do with, with surveys and you can actually go buy services to, to spread the surveys not only on your website but to the general public just to get insights that— and it's not that actually expensive. I feel like that's a highly overlooked way of gathering quantitative data to make some decisions. Okay, so we've got reviews looking at— we've got some quizzes going on. So we're capturing all of the data. At which point in time an ideal customer profile comes in place? Like, do we construct the persona before we start everything and we're trying to sort of validate against that? Or do we actually just gather the info and then we look at that ideal persona which came out of that?

Yeah, I would say I typically have a very good idea after I've covered those three areas, the internal side, the external, so prospect, customer, and market side. If I covered at least in one or two ways all of those three areas, Then I start getting a sense of the actual customer personas and I draft like full comprehensive customer personas. I actually have like my framework that I like to use for mapping those out in, in actually 5 different areas. So the way that I define it is, okay, the demographics, which is more of like the identity, so who they are. Then the motivation, which is what they want. Then the other third step is evaluation, how they decide. Then we have resistance, what friction points, what objections do they have? And finally, the vision, which is where they want to be or who they want to be if there's transformation once they bought. So all of these like 5 different areas and I have subpoints inside each of them in my customer personas. But basically they give you a really good idea of their thought process, decision-making, the psychology going through the whole process, the alternatives that they were comparing, what brings them to the page. And I find that once I've covered all of those research areas, I typically have a clear idea of these points, which enables me to say, okay, now we can move on to the strategy part. Which is the actual positioning, then messaging strategy, and finally with the copy. And the first time, especially if it's, if it's a launch of a new product or if it's a repositioning, these will be assumptions, obviously. And once you write the copy, you can test it in the market and see if you need to change anything, if you need to adjust. So it's, it's, it's a constant loop, I would say.

How many personas did you actually end up with on average when you work with businesses? I'm just interested to understand. Well, there's a lot of data coming into you. It's like, do you try to distill to a couple of personas? Because having too much potentially, you know, it's just like, it's too much, right? You don't even know who you're targeting. So it's like 1, 2, 3.

Yeah, that's a great question, actually. So it depends. I think it depends on the It depends on the stage where you are at, right? So I think we— I just finished working with the kind of an early stage company, earlier stage. So they, they got acquired, they basically need to redo the whole product marketing and relaunch everything from basically from scratch. In their case, especially because we, we weren't able to speak with a lot of customers, we spoke with some customers, but they also had the system integrators partner side because this is a data integration platform, right? And we weren't able to really speak with a lot of partners. So what we did in that case was draft two personas, but we didn't really have like specific roles for each persona. We had like some groups. Okay. Like for example, we had the value seeker, which is the more of like the business side of the persona they want. Like the business outcomes from the platform. And then we had more of like the technical influencer. And I say influencer because we knew that in deals, the business persona was the first one coming in. And then the technical influencer was the second one who actually jumped in after and influenced the decision. So in that case, we didn't really have specific, I don't know, like you have a name and she's the CMO person and then you have CTO. We had groups. So we say, okay, the value seeker, the business persona might be a CMO or CFO, HR, procurement, all of those. And then we had the value seeker, which could be any like technical position that we saw came coming up in deals. So in that case, we had more generic personas. In other cases, when you are already established and maybe you're just repositioning, I typically have at least 3 personas just because I like to always cover who is the check signer, who, who, who actually signs the check, then who is the, who's the manager, and then who's the influencer and who's the daily user. So typically it's between 3 and 4. And in that case, I know that we are pretty accurate and we just need to flesh it all out and test it with the first round of copy.

I love that. So you got the daily user, that's going to be my biggest fan in a way, because they're going to be using whatever I'm selling every single day. Then have a local influencer within the business, a person who effectively trying to get manager to agree to, to, to buy into my tool. You've got the manager, which effectively will put the case to the check signer that we need this tool. And then you have an actual person who can give you the money to buy it, right?

Yeah.

So you have these 4 personas as a very healthy way to think about this. So you construct these 4 personas and I imagine the copy targets them in separate landing pages or like how do you use the personas then to actually drive the messaging? Yeah.

Yeah. So Obviously, when you write a homepage, for example, the job of the homepage is not to, to actually sell to any of those specific personas. I think the job of the homepage is to actually match what they want when they land on the page and direct them to specific areas of the website when you can actually start creating and going deeper into your argument. But there's different ways of addressing specific personas, whether it's in a specific section of a page or whether it's writing a standalone landing page where you send paid ads to. But another very useful way is to have specific either use case pages or persona pages. So you have, for example, I know data integration for marketing, data integration for sales, integration for IT. All of those need to actually speak to that specific persona. Right. And One, one thing that I also like to always pay a lot of attention to when I create the visual hierarchy of the website and the UX, the experience of how people consume the messaging, the copy on the website, it's also to pay attention to your navigation. And I probably got this from my background as I worked at a usability testing startup in working a lot with e-commerce brands where it Finding your product, your specific product among a lot of products is the main goal. I've, from running usability tests, I see that a lot of people just hover, for example, on navigation to just orient themselves. Orientation is the second step that everyone, if you think about it, you land, okay, this is for me. Okay, it's got my attention. Then I need to orient myself. Where do I go? Where, what do I click on? Right? So the, one of the very first steps is hovering on the navigation. If you see those use case pages or personas page that speaks directly to you, then you've already won another click, which is already a big achievement because it's a series of micro steps and micro decisions.

Okay. I actually love you calling that out. We had an episode with a couple of e-commerce people and, and it is such a brutal game. You know, you have like 1.7 seconds when someone lands on your page, majority are from mobile. To actually, you know, make sure they stay.

Yeah.

So everything needs to be pretty much perfect if you're talking getting the conversions to the next level. So I want to take us into AI a little bit, Chris. So we have all of this data and everything I was thinking when you were talking is just, I'm going to just put this in a page. I'm going to upload it to, you know, ChatGPT as an example, and I'm just going to ask the questions and try to simulate my personas, try to distill that information, you know, get some efficiency in that process. Then rather than looking at it myself, what's, what's your take on this? Like, is that possible? Does that work? Or are we talking hallucination and actually bad idea?

I think it works. So I typically do it and I always say, especially if you're just starting out, simulating customer personas obviously doesn't get you to 100% of accuracy. But even if you get like 60, 70% of the way there, it's already a major advantage compared to companies who never do customer research, never actually have any, anything to actually bounce their ideas off of. So they just basically throw stuff out and guess at what might work, right? So it saves you at least some, some, it gives you some ideas that you can at least get started with and take action on. Rather than just stalling. But the problem is you have to kind of do it the right way. There's also— so there's a very specific process that I at least follow. Obviously anyone using AI has got their own processes, but I think the most important thing, you still have to at least start with some real human data, whether it's just some reviews, as we mentioned, some competitor messaging, or at least an extractions of insights that you've gotten from looking at competitors. Those sales recordings, they're super valuable if you have them. So the first thing that I do is I start my own kind of, let's call it like marketing assistant chat where I share all of those documents and I start having a conversation with it back and forth, back and forth, giving it all the context that I have as well. So it's always up to date as we go. And for that step, until you are kind of the strategy step, I would definitely use a model that has a big context window because if you ChatGPT is probably not great, but something like Gemini, which is going to probably the biggest context window, it's super helpful because it can keep all of that data in, in memory and kind of give you a retrieve the right information at the right time. If you want to be even more specific, something like NotebookLM, which can actually— it's limited to the documents that you share with it, so you can actually make sure that everything that it gives you is accurate and doesn't hallucinate, and you can actually jump to the, the rotation that it gives you. So definitely stay like broad and big context first to start. So you can instruct and have generate that kind of marketing assistant that can give you ideas and help you synthesize a lot of information and make that process efficient. Then when you start moving into the strategy, then you can start using models that are a bit better in terms of reasoning and writing as well. So something like Claude is excellent, I find. And with that, you basically go back and forth with your own human insight and knowledge and the, the insights that you've extracted from real research and start writing down the first strategy documents, whether it's the positioning or the messaging framework, so that you actually have AI-synthesized data but with supervision from a human. After that, once you have those kind of crystallized or version 1 drafts document, then when it comes to writing, the, the best thing that I always like to do with any model to make sure that it's consistent, clear, and, and that maintains all the guidelines that I give it. I have my positioning document, my messaging document, I have my research an insights document, which is all the research extracted. So these 3 documents, and typically I also like to do an extra step, which is sharing voice of customer. So especially like from reviews or from customer interviews, I call it the VOC injection, which is basically, okay, you ask it to write some copy taken from all these documents, but then probably the language is not yet, uh, super human or the way that you want it to be, right? So what you can do is actually, I don't know, take an extract from a couple of reviews that speak to a specific theme that you want to talk about in that piece of copy, share it with the AI and ask it, hey, can you inject some of this language into your— into the copy that you just wrote? And you will see that AI would start mirror that language in a way that's like very natural and very human, especially if it's some other like Claude, which writes very well. And that way you can actually write copy that sounds human with AI at scale faster than you would do any other way. And also then if you have other tools which are shared with people. So for example, I use one that's called TeamGPT with my team. You can actually share chats so you can share the same chats with the same documents with other team members and everyone basically writes from the same voice and tone and knowledge of the customer that you collected, right? So that's very useful to scale things up in your marketing team.

I mean, it's a, it's a, everything you just gave very clearly answers the question how to automate customer research using AI, which is pretty much step-by-step process. So thanks for that, Chris. Uh, one thing which came to my mind is can I actually fake the data? And what came to my mind is deep search, right? So ChatGPT has a deep search or deep research. I think Claude has it. I think Gemini has it. I think all the big engines now have it where, you know, you give like a huge thing you want to receive and it gives you, you know, 24 pages of information. So I feel like, can we actually go in there and get this fake or demo type of information about our customers? So we sort of maybe start with the customer persona or hypothesis and we go like, we don't want to do reviews, we don't want to do the sales calls. Can you just generate this for us? Like, can we somehow, you know, create this information and not actually gather and still be on point to some extent?

Yeah. So the main problem with AI is that it has access to all of the internet, right? So you have to kind of be able to restrict its field of vision and direct it in the way that you want it to go with, with its vast knowledge. So if you have, if you can guide it with a specific prompt and ask it, for example, search for, I don't know, these, the top 3 pain points, like something very specific. Obviously I wouldn't ask, give me the customer, customer profiles or ICP. For my 3 customer personas, that's super generic, super vague. So I would probably give you a lot of vague generic data, which is not useful. But if you ask for the right thing, so I don't know, pain points or give me the specific motivations as you find them online, especially with some tools that can search social like Perplexity, Deep Research, for example, it's very good because it can search on Reddit as well. I find it very helpful. Reddit has a lot of discussions that can actually give you some useful data, and we do some of that research to complement all the other research that we do. But I would definitely say, yeah, you can use, you can use it as a guideline to augment all the other research that you do, but it would never just rely on a quick prompt to ask AI to give me all of the research. I still need to be the one especially conducting customer interviews because I want to get that, that human sense, the human empathy with speaking with people, kind of getting a sense for asking the right questions, follow the conversation where it goes when I speak to them. And so being an external, just asking an AI to give you material can be a good starting point, good guideline, but I would never just rely on it.

Yeah, I love that. I love you saying that you still need to command their orchestra. You know, even if AI is the orchestra, you still need to tell them what exactly the music they should be playing, right? You can't just—

Exactly. It was actually— someone asked, someone asked me, okay, but like now that you delegate, you can basically enable clients to do all of this stuff with AI. Like, what's your job going to be as a copywriter, messaging strategist? And what I said was exactly I think copywriters will become more of like orchestrators, kind of architects behind the, the strategy first. So if you're a copywriter, look into learning strategy as well and not just writing copy, because then the strategy guides all the copy, especially now it's super important. And also big editors. So you still need to have that sense of what sounds good and what is, what copy is aligned with how customers speak and write. So then you can jump into the copy that maybe AI wrote and edit it and fine-tune it. Because as a copywriter myself, I 100% know what is AI copy, especially from different tools, versus what is actually human copy.

Yeah, any, any power user can sense AI miles away. Yeah, like that. That's what I feel like. If you use the tools day in, day out, you can definitely give a strong context, a strong style of writing things which will not be generic spitout, which ChatGPT or others give to you. But when you have, you know, that generic one where you don't feed in like, you know, a book or your own writing style and you go like, write on it like this, you can definitely see them miles away. Like, I actually had a post on LinkedIn where, where I was like, everything I see in my feed right now, it seems like it just has been throttled through ChatGPT, like 90% of that. Yeah. So everyone's looking at that efficiency, but at the same time we started looking the same, right?

Yeah, totally. And the big component of doing it right, I think it's doing it a lot first. So you understand what sounds AI, but also because there's this big part of evaluations, which is they call evals, right? Which is you can do it with AI agents, which they take an output from an AI, evaluate it and refeed it back and continuously improve the output. But you can do it as a human as well. So ask the AI something, see the output that you get, give it feedback like if you were like a mentor or teacher and tell it how you want it to be, what you want it to sound, what words to avoid, what kind of frameworks or structures you see are very redundant and very AI-sounding like, and kind of evaluate the output constantly until you get to the final, final result that actually sounds more, more like you or more like the company you want to speak, speak at. And nowadays there's a lot of tools that allow you to set your own specific style, like for example Claude or System Instructions, for example. So once you've found something that sounds right, then you can keep it quite consistent over time.

Yeah, we're getting better. We're definitely getting better. When ChatGPT came out, I've seen a couple of people trying to jump and create quick services where they, they were saying, you know, we'll give you your brand language within the outputs of ChatGPT. And that suddenly died off because ChatGPT just became so much better. And I feel like, you know, every week we're getting just better and better with that. Potentially, you know, it's not going to be a problem moving, moving forward. So, Chris, this question: can you simulate customer behavior with AI?

I think I would say you can simulate persona behavior, but not real customer behavior, at least not yet. And the, like, the reason I say this is that As I said, yes, simulating customer personas, so feeding AI all of your research material and instructing it to be your customer persona without breaking character. It works to get you to that 60, 70% maybe insight. I think especially if the input that you give it is actual real customer human research, that's very good. And in fact, one thing that I've done was also sometimes testing it to simulate my clients because I know, for example, I knew, for example, some of my clients had a very strong, like, opinion on maybe copy. So I wanted to see if I simulate my client's persona and I share the copy before sharing it with my clients, would I get the same kind of feedback? And it was super impressive. Like, it was pretty accurate. I would say maybe 80, 90% of the same thing that they, they, I said was kind of in line with how my client reacted. So I would say it can work. But yeah, obviously you need to actually still validate with humans and still see if it resonates with humans because we are probably too complex for now with compared to what AI can actually achieve.

Yeah, 100%. I asked this one because I was interested in what was coming through my mind is some tools I saw. Where they're trying to simulate, you know, customer behavior exactly for that, for just quantitative data inputs.

Yeah, I think, you know, I know, yeah, I know a couple of ones. There's Synthetic, Synthetic Users is one that I tried. There's a couple of other ones. And I think they're really good for, especially when it comes to like product insights, maybe like features and stuff. But when it comes to actual messaging that needs to resonate emotionally, We are still— it can be a useful tool to get started, but then like we're not still 100% of the way there, I think.

Yeah. Love you calling that out. All right, let's talk about website copy then. So we are at this stage now. We did the research, we used AI as well to just distill the research into something actionable. We have our personas, we know the motivations, we know the evaluations, the resistance, we have the vision. So we have all of these things. So now we're at the point where we just want to write the copy for a website. And according to HubSpot, 72% of SaaS websites fail to clearly explain what they offer, which means they leave so much money on the table. People come in and they go like, I have no idea what that is. A lot of features, features, this and that. I'm, you know, I'm off, right? That's not a great copy. And I see that all the time, by the way. I see that all the time. And as well, a lot of founders I work with and they argue like they will go in a defensive mode instantly and they're going to be like, what are you saying? You know, we copied our sales page from the biggest competitor. Like it's pretty much the same. We just rewrote our features. This clearly has to be good, right? So they need that education, right? They need to understand that. So I want to open this with the Jobs to Be Done framework. I know you use this. So how do you apply Jobs to Be Done framework to website messaging?

Yeah, so there's a really good book actually by Bob, by Bob Maester, which is more about using Jobs to Be Done for sales. It's called Demand Side Sales 101, I think. And it gives you a very good framework for understanding how people make decisions and how they have to make progress to those decisions. So with Jobs to be Done, it's important to find the job to be done, the functional, emotional, social job that you need to accomplish. And to understand that again, it's first doing that deep customer research. And the things that I like to understand as well are what's the awareness level of my prospects? So what do they know? And then sophistication. So how much do they know about my product, about the industry? And then to understand their decision-making style and to understand that, I try to ask the before, during, and after in any of my interviews or surveys. And that's what— when I was talking about understanding the decision-making process rather than product features or ideas for future product updates, decision-making process for me is understanding the before, during, and after. And I try to ask all those questions like, I don't know, what were What were you using before you tried this solution? Or when it came time to actually evaluate this solution, what other alternatives were you comparing? Which is also about understanding things like competitive alternative approaches. Doesn't mean that it's a direct competitor. Could be another way of doing the same thing, right? For example, if you have a tool like Productivity Tool, another way of doing it could be using spreadsheet or paper. So understanding all of those, then when it comes down to like the during, asking questions about like the main value, I always like to ask what's the big problem that this solution helps you solve, which gives you a lot of very good language that you can use in your value proposition when it comes to defining what you are or who you help and how. And then going into kind of the, the after, now that you are using this solution, what are you able to do that you weren't able to before? Right. So those kind of questions help me map out those before, during, and after. And then we have to kind of try to make sense of all of that in our process and map it out on the— first, I always start from the homepage. And one of the biggest things that I see B2B SaaS companies doing wrong, especially enterprise, but that's a whole other thing. It's they mess up their value proposition at the top of the page because they always try to tell people basically to communicate their positioning right away. We want to tell them this is what we do, this is how we do it right away, like from the get-go in the above the fold section. But there's a very good approach that I've heard from, and I always remember from Matt Learner. He's got a great book called Growth Levers, I think. He said, imagine that your copy needs to continue the phrase I want to, and then you basically complete the phrase and that's your value proposition. Like a super simple, super simple way of doing it. So imagine that you, I don't know, help people, help marketing teams save 10 hours a day. Your value proposition could be save 10 hours a day because it completes the sentence I want to. Say 10 hours a day, right? You're not telling them we are a productivity software tool. Uh, yeah, you have to phrase it in terms that helps them understand that you, you know what's going through their head when they land on the page. So you, you have to try and match that. And okay, and that's typically the main job that you need to solve for, right? That's what they want to achieve. And then throughout the page you can expand into all of those functional, emotional, social, more like secondary jobs. But I would definitely approach in terms of understanding first decision-making process and then mapping out that first job in the, in the value proposition.

Okay, that's excellent. There's a lot of content on our homepage, Chris. So you mentioned above the fold and on mobile is going to be different as well on desktop. Some of the desktops will be different depending on the resolution. So what's the science and thinking from, from your perspective? Like what's, what goes above the fold? What's the main message? What goes below the fold? You know, do you have an approach where you go like, that's going to be H1, like, you know, bold, big letters, H2, smaller? It's like, how do you structure all of that? Because the homepage itself, right, is going to have— yeah, it's going to have some content.

Yeah, I always like for the above the fold section, I always like to have an H1 subtitle and then a call to action, which the call to action doesn't mean that we need people to click on that. It's more for, as I was saying earlier, for orientation so they know what the next step would be. But still, our goal when we land, when they land on the page, our goal for them is to actually make the decision of scrolling and learning more because we don't want unqualified people. We want them to land on the page, understand that this is for them, but we also want to understand how we do it and what we do so that we qualify them and we know that they are the right prospects. Right. So that's probably the main goal. And when we go through the page, then I typically have the What We Do section. So this is actually telling them this is our position. Positioning. This is what we do, who we do it for, and how uniquely better. So I typically divide it into sections. The what section, what we do, then how we do it, which you can go into the benefits, the differentiators, features. And you mentioned, yeah, a lot of companies use features instead of benefits, which is good in most cases, I would say. But still, you need to, especially in B2B, when you're speaking sometimes very— writing about very technical topic, very, very technical concepts or products, you have to still include the features. And I always like to, from a usability standpoint, I always like to use features as like what I call eyebrow copy. So above the H2, let's say, that talks about the benefits. So you would have feature benefits H2, and then you can go in more of like how the features connects with the benefits and to the— and with the value that you offer user in the paragraph copy. I would say you have to kind of understand what balance to use there because it depends on your audience, right? On, as I mentioned, the awareness, sophistication, how much do they know, how much they need to know to bridge that gap. And you have to balance features, benefits, and the differentiators. That you use.

I love that.

And after that, probably after the, the how that you explain, then it goes into the anxiety section that was mentioned before. So we are again motivation, value, anxiety. How do you address anxiety? What kind of copy do you have to address objections? Maybe bringing some objection objections up proactively before they actually come up with those and social proof and the final call to action.

Yeah, this is excellent homepage. Thanks for that structure, Chris. That's excellent. Um, I love, I love you blending features with benefits. I feel like features alone is very cold page, especially like you have these sales pages where they go like Plan A, blah blah blah, features list, £60. Plan B, blah blah blah, features list. I mean, if you already have me and I coming to that page, potentially good, but if that's your the only page That's like, there's no benefits. I, you know, you're forcing me to think how this feature is going to benefit me. Uh, you know, you, you're forcing me to think too much. I'm gonna go to a more, you know, better converting page effectively.

Yeah, especially so, okay, so when you, yeah, when you consider like the messaging that you come from, like if you landed on an ad that was all about benefits and then you land on a page that's all about features, that's like too much of a gap.

Yeah, I love that. We've got a page now, right? So effectively, this is where we go. We got big research, some AI in the middle to allow us to get some answers, but effectively still lots of human work or lots of human oversight. We wrote the copy, hopefully it's good. So we at this point now where we launching the website, and effectively when— that's a hypothesis, we're going forward, but by no means is that we're going to nail this in the, in the, you know, first time we do this is going to be perfect copy. Millions are coming. Good. That's just probably not what's going to happen. So how do you set up the environment in such a way? I mean, tools to capture information potentially. How do you get the feedback on your copy back to you so you can actually iterate on that copy and make it better and better and better and more converting over time?

Yeah, well, obviously there's A/B testing that you can run, but if you don't have the traffic, it's probably a no-no. But even before getting to launching the copy, what we do sometimes with clients, especially if they're sales-led rather than product-led, we might create their sales collaterals, their sales deck presentations, and start already testing the messaging in their sales conversations. So when they have those sales pitches that follow the narrative the, the value themes that we created, then their salespeople can already start testing them with prospects and we can start getting data that we can use to improve our copy. Another way you can even test in emails. So if you have the first couple of signups or already have a list, you can start testing your messaging in subject lines. There's probably the fastest way to test the specific, especially if you want to test specific, um, points that you want to hit. Maybe it's their pain points or their motivations or different features if you want to see how they resonate. And then obviously there are some tools. So they're probably the biggest one in B2B especially, it's Winter, which is a message testing tool. And I haven't really heard about many others, actual message testing tool in B2B. I always recommend usability testing, but more for If you are in B2C, just because it's hard to get the right tester panel in B2B, like it's hard to get a CMO or CFO to spend 20, 30 minutes looking at a website to give you feedback, right? Unless you pay them very well. So I like to use usability testing if I know that my SaaS company, maybe sales, has a segment that's B2C as well. So we want to test that. Or, or if I want to see how maybe how they move through existing copy or signup process, if everything works well, just because maybe I need to optimize some message or some product messaging as well inside the product. So in that case, I might use it. But yeah, I would say once you put it out, then it's more of like constantly iterating, but being clear on what kind of points you're iterating on. You obviously don't want to rewrite the whole copy. You need to start small and divide it into those sections. Like, again, let's test first our motivations for each page. There's a different motivation. Let's test our value and let's test our anxiety sections one by one without like reinventing the wheel. And yeah, start getting feedback as much as possible. Sometimes I might even do on customer calls, I might even ask them, hey, like, what do you think? If I have video as well, sharing my screen, hey, 'What do you think about this? Is it clear? Does it resonate with you?' Yeah, especially if it's someone that's not yet a customer.

And, uh, yeah, I find people— I get these sometimes asked, by the way, uh, where like, you know, 'Hey, we're starting a new product, you know, do you want to be a beta tester? You know, here's a new copy, can you give us a feedback?' I mean, ultimately, I, I don't have time, so I, I just keep pretty much through 99.9% of them. So I would imagine a lot of people do the same thing. That being said, though, if you have your angel list, you know, or a group of other people doing the same work, any feedback is great feedback. And then once you have that A/B testing, that's where real data comes in because that's— you're probably buying ads. Maybe you have some organic SEO or whatever, but people are coming to your landing pages and you're seeing what's happening. So if you're doing the test tests. You're constantly iterating on these messages. And to put this— to put more weight on this, Bain and Company found that companies who refine their messaging regularly grow revenue by 8 times faster. So 8 times faster, imagine this, right? That's, that's a huge, huge X. 8X is a huge amount.

Yeah, I will also say another maybe underrated way to test your messaging. Obviously, they can't go qualitative, but if you want more quantitative data or more of like an understanding at least of what works versus what doesn't, not the why, but if you want to want to know what, looking at user interaction data, so user recordings, for example, with Hotjar, those, those that you see are all people that are your prospects because they land on your website. So same audience that does the— that could do maybe testing. But you can see whether or not they hover on specific piece of copy, whether they like do some rage clicking because they maybe they didn't understand something, or they, they're confused in any way, or they go back and forth between sections because they don't understand. So you can start getting some sense at least. Yeah, from those.

A good shout. Yeah, I love that. All right, Chris, let's go into quick questions round because I have a couple more and we had the time and I still want to pick your brains on some of these. So quick questions. What should a message contain for online marketing to actually work? One thing, which is the most important?

It needs to contain that message market fit, which is basically having the right positioning with the right language for the right audience.

Excellent. How do you create an email copy that actually drives action?

Well, first it's about understanding who we are speaking to. And once you have that, then use again that sequence that I mentioned— motivation, value, and anxiety.

Would you put that in an email like, hey Tom, motivation?

Yeah, because if you think about it, if you think about it, motivation can be several things. Could be their desired outcome, but it could also be their pain point, their struggle, right?

Yeah.

You need to understand which, which one are they more— are they thinking more in terms of going away from a pain or moving towards an outcome? And then using the ones that you see works better with them.

Okay. And we still want to have effectively the, the, the anxiety in that email. So it's almost like the same copy, but doesn't that make you longer as an email?

No, no. I mean, anxiety could be something as simple as saying in one sentence, our clients saw 30% increase in conversion. Boom, that's done. You basically provide social proof in a couple of words, right? So it's just— and I would always actually try, especially if you include some value. I would include value in any email. Needs to be valuable of some kind, especially to provide value. If you can back up that value with some social proof, it's always best, especially if it's contextual. Contextual. So if it's related to that value that you just mentioned, same thing that I would do on a page. I would always have— if I have a paragraph that talks about features and benefits, then I try to back it up with the testimonial that speaks to those specific features and benefits. The contextual aspect matters a lot.

Love that. How can you speed up copywriting without losing quality with AI?

Context. Context is huge context. So sharing all the information, real human research, and yeah, directing the AI in a specific direction rather than just letting it access the whole of the internet.

Okay. What's your best tip for cold call copywriting?

Cold call. I never actually, it's not really my, really my, not really my forte. Yeah. I will. Yeah.

Well, imagine you're getting a cold call, like what would make you to stick on a call? Call if someone like first sentence they say, you know, I mean, they have like 5 seconds.

Yeah.

You know, because they always talk like very like fast and they have no gaps where you can go like, thank you, not me. So they just blah blah blah, you know. So they don't even give you a chance to sort of go like, and they try to shove a couple of sentences in your brain.

Yeah.

What do you imagine should have been said like to make that work?

Yeah, I would probably have to, to, I would want them to use the exact words that I'm thinking Like, say, if you are expressing a pain point, how do they express it? What kind of language? Try to mirror your customer's language in whatever pitch you're using, because I get a lot of pitch slaps, I call them, on LinkedIn. So people who add you and instantly pitch you in the first DM. And the thing that they don't really get, especially at the top, first, sometimes they say, you are a 'LinkedIn lead generation agency.' No, I'm not. Gone. I don't even pay attention. So it's like using the same language, same specific points that address what I do and the motivations they have.

Yeah, it's called show you— show me that you know me.

Yeah, exactly.

Yeah, I get pitched. I get pitch slapped all the time, and it's just so annoying. Right now I'm in this mode where I'm like, if I get this, I just instantly unconnect. Yeah, like, I'm like, like, it's clearly we connected. Like, I accepted your connection because that first message was great. Yeah, right? You said something, I'm like, okay, cool, you know, let's be in a network, see where that takes us. And then, and then it's like, boom, it's left.

Yeah, sometimes like even before that, like, if you— if they ask you for a connection and you see that in the kind of bio line that I have under the name. If, for example, I see, I help agency owners get 10 clients a day, I don't even add you. I know that you're going to pitch me.

So yeah, that's quite clear. I get you. Right. How do you write ad copy that actually sells? Ad copy, not a website, but ad copy.

Specific ad copy. I'm thinking because I'm working with a client now on actually looking, validating some of their ads. And I would say, yeah, be clear on the angles that you're using, both in terms of what personas you are addressing with those ads. So if you have different ads for different personas, make them very distinct and clear. And also going back to the awareness and sophistication, how much do they need to know? How much do you already know? And try to bridge that gap in the copy.

Okay. I love that. So it still means that you have to do homework. Like you don't just go in and go like, I created this website, I did this product, I'm going to buy now ads.

Yeah.

And just sort of off we go, fingers crossed, you know, hope things work out.

Totally. Yeah.

Because as you mentioned in the beginning on the episode, like you do the research, before you do anything else.

Yeah, exactly. Because if you think about it, like what I say, it's people don't just read words, they need to unpack meaning out of those words. So if you think about it, when you read some piece of copy, you have some work to do on your part to understand what they mean. If that copy is confusing or unclear, you have more work to do. So you want your prospects to do as little work as possible when they read your copy. They basically need to speak their same language. It needs to come here and stay in their minds as long as possible to— for them to actually click on that ad.

Don't make them think.

Yeah, yeah. The, the, the best copy is the copy that doesn't look like copy.

Yeah, that's true. I love how Reddit does their, their advertisement because it's like, um, I was recently in their webinar And, and effectively they're suggesting when you do ad copies, actually make an ad copy like it would be a post in the feed.

Yeah.

Because that, you know, ad comes up in a feed when people scroll down. And if you're targeting that person, that ad should just be just as an, as a Reddit post. It should just add value to whatever is the information they're trying to consume rather than being an actual ad.

Yeah. And you made a good point, like it needs to add value. Because otherwise you'll end in those dark patterns which, yes, grab attention, but then you— people realize that it's an ad and they report it.

Exactly. And then they get annoyed and they go, oh, you sneaky bastard.

Exactly.

They just, you know, not only that, they bounce, they probably have adverse feelings now towards your brand. Yeah, right. How do you clarify your marketing message when you got too much to say?

Yeah, it's, I think it's, it's first understanding what the messaging strategy is and how it's composed, what the different elements are. So first understand you have your positioning, which is what you do, how you do it, and who you do it for. Then the messaging strategy, which is how you say it and why it matters for them. And then the copy, which is actually the words they use to express all of that. So I would say to be clear, as clear as possible, understand these different elements and then connect them together. Because if you start with any random copy, but then there's nothing before that, uh, the strategy is not clear, then that's when your copy is actually diluted, confusing, vague, and jargon. So yeah, understand how good copy is formed and composed.

I had that problem. I still have that problem. I just want to say everything. I just want to, you know, my mind goes like, what? Well, what if you don't know? So like, I want to tell you, right? And it just becomes like 6 A4 pages of one landing page copy. That's a real struggle I have, Chris. Seriously. Right. So I think you answered this just now, but let me ask, what's the best way to position your message so it sticks?

Yeah. So as I mentioned, defining, being clear on what, what you do, how you do it, and, and who we do it for, but also injecting some of that point of view and in that insights that I was talking about at the beginning before, because when you have that strategic narrative, the product history as well, and the difference, what makes you different, then it's the actual messaging that sticks with prospects because they start realizing, especially if you use it as an insight at the beginning of a pitch, for example, it's what actually hooks them and makes them stay because they realize, oh wow, like these guys see things differently. I want to know more. I want to know why this thing is better for me because of the way that they see the industry, the market, the customers, um, solutions.

I love that. I mean, I love all of this, Chris. I mean, that's packed with gold. So thanks for sharing all of this with us today. And one thing which I take out from all of this is that we need to show the customers that we get their problems and effectively we need to make it easy for them to say yes.

Yeah, exactly.

That's, that's like my, my, my, what I've learned.

Very good encapsulation.

Big science though, big science, lots of process to actually get this done well. But ultimately, if we do that well, and according to the quote above mentioned, 8 times more conversions, that's a big number. Just think what that means if you actually right now making, say, half a million, that can be 8 times more if you get your message better, right? That's not a joke. So tell us, Chris, What's up with Conversional, Hemi, and what's happening next and where people can find you?

Yeah, currently I'm always working with clients, B2B clients, mostly SaaS, but also just hired a junior copywriter. So kind of expanding, optimizing my process is always, always learning. I use it every day. So that's a big, big part of a big thing that I'm always interested in talking about and nerding on. And yeah, if people want to chat about messaging, copy, positioning, AI, they can find me on my website at conversionalchemy.net and I have a newsletter there and a podcast and typically just on LinkedIn, mostly on LinkedIn, just chatting about these things, sharing things that I learned.

And yeah, I love, I love you saying always learning. That's exactly who we are as well. Always learning. That's why we have the master in the title. It's just never-ending journey, right? Especially tech landscape tools and innovation come left and right and you always need to be effectively trying all of them out, distilling the best ones, throwing up, you know, the shit away and giving your clients the best you can. So people interested in Chris and Conversion Owl, let me check the show notes. You'll see all the links there and If you enjoyed the show, as always, hit that subscribe button and send us your biggest growth question at masteringtechgrowth.com. And as always, until the next time, stay safe and see you online.