Chris Silvestri on Mastering Persuasive Copywriting

Welcome to the B2B Growth Blueprint podcast, where we talk with entrepreneurs, founders, and leaders at B2B services, tech, and SaaS companies to learn the systems and processes they've used to create success for their business. We'll also learn from marketing and sales experts to provide insights to help you grow your business. I'm your host, Mark Osborne, and the podcast is brought to you by Modern Revenue strategies and their 10x ROI B2B growth guarantee. Schedule a free diagnostic call to learn the fastest path to double your B2B revenues in just 90 days. Hello and welcome to another episode of the B2B Growth Blueprint podcast. I'm your host, Mark Osborne, and today I'm very excited to welcome my guest, Chris Silvestri. Now I'm going to read Chris's bio to make sure I get this right. Chris is the Conversion Alchemist. He's a SaaS message market fit specialist and conversion copywriter. He worked 10 years as a software engineer in industrial automation, then took a sharp turn to enter the digital marketing world as the user experience lead at the usability testing startup known as Conversion Crimes, and previously at the conversion design agency Zeta Labs. Chris has been working as a copywriter and conversion rate optimization specialist for online brands since 2016. Chris, welcome to the show.

Hey, Mark, thanks so much for having me. It's a pleasure being here.

Yeah, I appreciate you coming on. So you've, you've done a lot of things and had a sort of an interesting path. Why don't you sort of tell us the story?

Yeah, so as you mentioned, it was— I spent 10 years as a software engineer, basically just stumbled on whatever was the most popular profession in back in my hometown in Italy where I'm from. I did that 10 years, got pretty, pretty tired of it. And I also had a band. So I'm a drummer, been playing drums for like 25 years now. And at the time, what do you do? No, it's hard to take time off to tour with your bands. So you always had to ask 3 months in advance for time off. So I started looking online for how to make money online.

Okay.

To get a flexible job, find something that I could do while I know traveling. And I tried different things. I tried an Amazon FBA business selling yoga straps. I tried SEO blogging. I tried lots of different things, but nothing really worked. Nothing really stuck until I stumbled on copywriting and I was instantly hooked by the psychology behind it, the understanding of the decision-making process behind how people take action online, especially and through the copy that you write, through the words that you write. And so I started studying on the side, weekends, evenings. Did that for a couple of years working on Upwork. And when I saw that, okay, I can make maybe $1,000 a month on the side with this thing. So maybe it's time to try and take the leap and take it full time. So at the same time, I moved from Italy to the UK where I now live, took this full time. Then I transitioned more into the UX role working for the startup agency as well. And I learned lots of things that I was kind of surprised how they intersect, how much they intersect with the research that's in copywriting, the UX research. A lot of things I saw the value in combining them. And so when I saw that, okay, it's time maybe to go back on my own, I decided to leave the agency and basically create Conversion Alchemy, which is my current business where I try to combine all the copywriting knowledge, messaging knowledge, the UX, the buyer psychology behind combined it into a mix to help my clients, which are B2B SaaS companies, typically Series A to C companies, improving their messaging in a way that makes their prospects want to take that desired action on either on the websites or emails or even just looking at a sales deck, right? It all stems from the core messaging.

Yeah, that makes a lot of sense. And so you talk about, and I love this phrase of message market fit. Talk to us a little bit about what you mean there, because folks are familiar with product market fit or maybe even go-to-market fit. What do you mean by message market fit?

Yeah, so message market fit, the way that I see it, is I think you should think in terms of how you address your prospects, not really as just telling them what you do and how you do it, but actually first thinking about matching a conversation that's already happening inside their heads when they land on one whatever sales marketing asset you're presenting to them. So that's the main important like mindset shift that you have to take, I think. Imagine them having a conversation. How do you match? How do you fit that conversation with your messaging? And then obviously your messaging stems from solid positioning, messaging strategy, and copy. So it's 3 components. that help you reach that message-market fit, which is basically when you have the right positioning for the right audience in the right language.

Yeah, it's so funny, like how often you see companies that, you know, all of their messaging is about me, me, me, me, me. And they don't take the time to really understand what their, you know, prospect is going through. And oftentimes they haven't even thought about who their prospect is. And so one of the things that we really like to do is really get crisp on like defining an ideal customer profile, not just, well, anybody with a checkbook, but specific types of companies that are going through a specific context in their business.

Yeah.

That maybe they're growing or maybe they're shrinking, like both can be, you know, different kind of contexts that would trigger them to go to market or that would drive them to look for different things and solutions. And then really thinking through the personas of the different folks on the buying committee, from the actual decision maker to also the influencers and maybe the person who's out championing for this project and the end user, really understanding those folks from both a personal and professional lens gives you that empathy that you can then start to create this messaging for.

Yeah, yeah, that's a, that's a great point. Like, consider all of those different perspectives because People think that, okay, your homepage has a job of basically selling your product, but that's not true. Like anyone could land on your homepage from your CMO persona to the CFO, the procurement, all of those, or to IT, right? All of those have drastically different needs, wants, desires, struggles, and challenges. So you have to keep that into account when you write your messaging in order to address all of those different facets, different framings that people bring to the table. There's one thing that I kind of, one of my mantras that I say to kind of understand that people are different is that people don't just read words. They need to unpack meaning from those words. And for each of us, there's a different meaning. Same words could have totally different meaning for different people, right? So you have to know that and put that into context when you write your copy.

That's a, that's a great point. So what are some of the sort of tricks, tools, approaches that you use to build that empathy, to sort of really understand what the, you know, sort of significance of particular words are for particular audiences? Like, how do you get in the mind of, of those audiences? Because you work with lots of different businesses with lots of different contexts. So how do you How do you how do you do that?

Yeah, so what I always say to all my clients, most of them get it wrong when they start working with a copywriter, maybe for experiences that they've had in the past. But I always tell them seventy percent of the work is the research. So don't expect me to spend one week on research and then two months on writing because when you do the research part right, which is typically the longest part, then the writing. Can even take literally days because you have already all your language, all your different frameworks in place that you can just shape based on the persona that you're referring to. And so the way that I like to research, I see research kind of like an iceberg. Imagine like an iceberg, right? So you have the part outside of the water and then most of it is underneath. So the first part that a lot of companies already have access to, which is kind of the surface layer, tells you what users say. So the language, words that you can use, mirror, think of a review website that you can just go and skim and collect data from, or maybe you have sales conversations. So those are typically already available, especially if you record them. All of those is golden and tells you what users say, right? So second layer, it's the structural layer, I call it, which is what users do. And that's something that a lot of people miss, which is more of most of the experience layer. So that's the maybe some of the background that I have with UX research as well. But I like to think of my prospects or my clients' prospects consuming the website almost like they consume a product, right? So they have to go through different sections of the website. Each section contains a lot of micro decisions that they have to make. And each section, in each section, you are not selling the final call to action. You're selling them moving to the next section until they get to the final call to action, right? So, so that's the structural layer. And for that, I typically look at heat maps, user recordings of users actually using the website, interacting, seeing where they flow, what pages they navigate to. where they first click. I might also run usability tests. Obviously, for B2B, it's quite hard because you need to find the testers who match your demographic and psychographic. So there are some B2B message platforms. For testing, I also use AI, which maybe it's something that we can test later on.

Sure.

But that's a good approach, especially if you're a startup. And after the structural experience layer, then you have to look at why people act. And that's kind of the deep layer. And for that, the most valuable thing you can do is speak to your customers. So run customer interviews, but not like the interviews that you typically run to understand what they like about your product, what features they would like. No, no, no. Start understanding their decision-making process. So what were they doing before they chose you? What other solutions or alternative approaches they tried? Why did they go with you? And What are they thinking now they are with you? So mapping the whole process, the before, during, and after, that's super valuable to understand all of those struggles, desired outcomes, and motivations that tell you how actually people think when they land on your marketing asset.

So I love that, and that concept of goes beyond the words to— and I love the description of the structural layer. And we think about that, you know, oftentimes as part of like sort of the buyer's journey, you know, Even that sort of structural layer outside your website of, well, what was it that first kind of triggered in them, maybe this is a problem that I should solve? Then evaluating, is it worth solving? What's the return on investment of trying to solve this problem? Then what are different ways I could solve it? Then, well, now that I've picked a way, who are the providers that solve it that way? And then you actually are getting into evaluating provider A versus provider B.

Yeah.

So many companies just start at, here's why we're the best. But if they sort of understood that journey that the buyer is going through and then started to influence the journey up here of, hey, this is a problem worth solving because we have a unique way of doing it that, you know, delivers these unique results. And this is the right way to solve it because we've seen it delivers these unique results for our clients. And we're the right provider that solves it. That way because we have this proprietary process. And so when you guide all of those steps along the way, you build the credibility and trust that then they, they've come to adopt a worldview that's aligned with your product and buying vision. And if they then believe that, well, there's no choice but you. And so you've, you know, sort of led them through that education journey. Yeah.

And you had an excellent word that I loved, which is that your worldview and And that's important because your worldview or your point of view or that unique insight that only you have on maybe what the industry is doing wrong or what your customers are doing wrong or what are some myths around in the industry that you are actually addressing and solving for, that point of view is actually one of the strongest points that you can use to hook them when it comes to your sales pitch. Or any narrative or any marketing material. We typically use for sales decks, but also for any emails sometimes when you have to start maybe a bit higher level to address some education, some filling in the gaps that you need to do. That insight, which then opens up the field for addressing, okay, this is the insight that we have, why people, what people are doing wrong and why. Typically, customers try these alternative solutions and you kind of explain to them, maybe they already tried some of them, so they understand, oh wow, like this guy really gets me. They understand my situation. I should listen more. And then you paint them like a perfect picture. In this perfect ideal world, the solution would address all of these points that we've addressed before, but even better, uniquely. And then you can actually introduce your solution. So that's from that worldview that you mentioned, you can open up all of that narrative that lets you in inside their minds, basically.

Yeah. Well, and not only that, but in B2B especially, you know, people don't want to change vendors a lot or partners or providers. Like, they want to find someone that they can work with for a long time. But if they're gonna work with you for a long time, you have to have a shared vision for what's gonna— what the world's gonna be like 5, 10 years from now. And so if you can't enunciate Your company's vision for what the world's gonna be like 5 to 10 years from now in your industry, well, then you have very little chance of, you know, capturing what the customer thinks is gonna be true and then aligning so that they feel like they can really partner with you.

Yeah. Yeah. Switching costs are huge, especially when you have like infrastructures and deep-seated architecture that you have to replace completely.

So you've got this really robust process around, you know, and I also love you talked about, you know, the research is, you know, what drives it all. And I always come back to this phrase that, you know, Abraham Lincoln said, if he had an hour to cut down a tree, he'd spend 50 minutes sharpening the axe. Like, and I think of research that way of like really sharpening the axe. But you find this, you know, the words that they use, and then you look at the structure of how they engage with your website and Then you conduct these customer interviews. That starts to sound like a pretty involved process to maybe an early-stage company that doesn't have tons of resources. What are some ways that maybe a company that doesn't have the resources to invest in all that kind of research can get some of that insight so that they can start to build that empathy?

Yeah, I would say probably the least, like the MVP research that you can do, looking at those 3 layers that I mentioned earlier. Surface layer, definitely look at reviews for competitors, look at the language that customers use, try to collect them and divide it. Okay, these are the key themes that these people talk about, key themes for motivation, for value, and for anxiety. That's the simplest way that you can actually divide these 3 different areas. Then competitors, But I would be careful here when I say look at competitors. I don't mean copy competitors' websites, right? So what I mean is deconstruct competitors' websites' copy in the sense, again, you can use the same 3 areas. Motivation, what brings people here? How are competitors matching that thought process? What's their value proposition at the top of the page? What's their language? What are their claims and promises? What's the value that they offer? So what are the 2, 3 unique things that each competitor does? Try to list those. And how do they address the anxiety or objections? What kind of social proof they use? Just try to categorize those without actually copying the copy they use. For now, just categorize, organize all that research somewhere. That's going to be handy for messaging. Then the UX layer. Obviously, if you don't have a lot of traffic, it's hard to actually track experience. But in this case, I would probably combine this with interviewing prospects or customers or people who are the same demographics. So you can use— you can do that either on message testing platforms. So you pay a certain amount and have audience that fits your demographic and psychographic. Maybe look at messaging that you wrote or your website if you already have a website up. Those are pretty viable solutions. Or the other way, if you already have a few customers, you don't need huge amounts of customers. In fact, I typically like to interview 7, maximum 10 customers, and that gives you typically 80% or 85% of the insights. If you can interview them, then you could actually also use those interviews to kind of ask them their decision-making process, their process through looking at a website. Maybe you have them look at not a competitor's website, but maybe a draft of copy that you wrote that's already laid out, like even a wireframe or like a simple Google Doc using tables. Doesn't even need to be a website. I've done interviews testing copy on actual, like, very simple black and white wireframes just with tables. So that's something that you can do, anyone can do probably. And finally, yeah, for the deep layer, again, if you don't have a lot of customers, I would say probably the best way that you can do now, it's using AI as well. That's something that it's not really talked about often, but there's this kind of trend that's popping up a lot. It's called synthetic research. So that means using AI to simulate your customer personas basically and ask questions or try to have a conversation with them or even test messaging pieces with that. But there's typically a process that I as well use to actually create these personas the right way. Because if you think about AI, AI is access to the whole of the internet and you would basically just be getting generic insights if you just tell them, oh, you are a CMO at a startup that does, I don't know, marketing intelligence. You probably just get vague jargon stuff that's not really accurate. So if you want, we can dive into my process. But yeah, I would try to make it as precise as possible with these personas so that you can actually get not 100% accurate data, but at least 60%, 70% that can get you started all the way there if you were just starting out.

So yeah, like you opened the door. I'm definitely going to walk through that. So tell us about this process for generating synthetic data to get an understanding of your customers, even maybe before you have access to a large number of them.

Yeah. So some of the problems that I see here that if you are in B2B, you're probably familiar with hiring a consultant, like a research strategy agency, it's super expensive. Some of those message testing platforms are also expensive. They're typically targeted enterprise. So it's even harder, even for me as a consultant, to justify using them with some of my clients. So I have this 4-part framework, which is called the PATH framework, P-A-T-H, which also stands for empath, empathy.

Okay.

And the first step is prepare. And by that, I mean using actual real research or the stems of it, the like very beginning, if you have some survey data or even just very little sales calls recordings for demos, all of those transcripts, you can use them and feed them to the AI so that you can start shaping this persona like an idea of the persona. And here when I in the prepare stage, I like to typically use 2 different chats. So I have one chat in an LLM. That's kind of like the marketing strategy. So this is a chat that I keep up to date at all stages, and it knows basically everything about my marketing efforts, about the strategy that I'm working on, right? So you can give me insights on specific materials that I share with it, or I can ask ideas for copy, all this stuff. And the other chat that we are— we're creating after in the second step, which is articulate step, is the actual persona ICP chat. And that's, I keep it separate so that it can wear the shoes of the persona. It also never breaks character. It should always stay in this persona's shoes so that you can actually ask questions as if you were asking questions to like a typical user or customer. You use all of that real research. So that's a step that many skip. You don't start with generic questions. AI knowledge, you use real research or as little as possible, but at least something that's useful to point the AI in a specific direction. And then you articulate this persona. I use a simple prompt, but by like, even today, LLMs are so good that you don't really need prompt engineering to actually do this. So the more conversational it can be, as long as you feed it the most context as possible, That's always pretty good. The next step, once you've created this marketing chat, the AI persona chat, then it's the test step, which is where you can actually have these conversations with the AI. So once the AI persona is created, think about asking questions about anything like, I don't know, what objections would you have, right? Or test for different angles or different resonance for emotion, different emotions that you can evoke with your copy or Maybe, I don't know, anything that you want to understand about your customers they are not clear on, you can ask the AI and see what it gets. The other thing is useful for testing messaging angles, right? So you share a headline or even just like one time I even, I tested an entire email sequence with like 2 variants for each email. So it was like 3 A, B, C for each email in sequence. And I feed that, I fed that to the AI. I asked them for each email that you stumble on, give me your opinion, your like first impression. Give me even a score. You can ask it and compare all of these emails. What do you think if you were in this specific context, the specific condition, right? You want to give as much context as possible. And the persona can actually tell you, okay, I like this first email. This would actually convince me to buy, or this would convince me to click on the next. The email in the sequence. And by that, you can actually start getting real good insights that you wouldn't have gotten otherwise, right? Because if you're starting out, it's hard to actually have real people have eyes on your assets. The final step, which is the harmonization step, is with all these insights that you've gotten, you can actually use that to refeed into the whole cycle. It's actually a cycle. And that can point you into doing better real research. And that better real research then can be refed into the AI. And then the cycle continues. It's basically exponential improvement. And there are a lot of companies that do AI agents now. So you could actually, if you wanted to be a bit more complex, you could build your AI agentic workflows that continuously iterate on this process. But even just like the manual processes with the commercial AI platforms, it's already pretty good and super valuable, I found.

That's awesome. So I want to say that back because I want to make sure I capture these. So PATH. So the first part is prepare, and you prepare by using whatever you've got from your current kind of theory of marketing, and you use one AI To sort of talk through that. And then another to feed it really sort of your proprietary research that you've done with the 1 or 2 customers or prospects that you have talked to and you've recorded some calls with or conversations or whatever.

Or even reviews, like even materials that you find online as if you don't have real customers.

Got it.

That's already good.

So you prepare both of those kind of datasets. And then the next, the A is what was the A for again?

Articulation. So basically preparing those AI personas. Yeah.

And so then, yeah, you prepare the actual personas from that. And one persona is essentially the CMO of your company.

Yeah.

And the other is the customer, the potential customer of your company. And then T, you test different messaging strategies with the buyer persona. That are generated from the sort of CMO or marketer persona. And then the final step is to sort of close the loop and harmonize that throughout and have it retweak everything again. Well, if this is true, then how would you change all of this? Wow, that's great. And do you do this typically within like a ChatGPT or another LLM?

Yeah. So I like to start with, so the first like preparation stage, I like to start with LLMs that have a huge context window. So right now, Gemini 2 or 2.5 just came out, I think. So 2.5 Pro, I think it's got like 2 million token context window. The 2 has got 1 million, which is already huge. Like I've never surpassed the 1 million tokens with tens, like dozens of interview transcripts of like 1-hour interview transcripts with the team, with my clients, review mine, survey results. So with 1 million, it's pretty much endless already. So I would, I like to use that at first because it's got the biggest context and it can basically like the, it's got like the freshest memory to retrieve all those information that you share with it. The second step, the articulation, I like to use something a bit more polished, maybe like Claude, just because it's better at writing and Writing with like from a human standpoint. So I use that to create those personas. And yeah, to test, I either keep on using Claude or with my team, I started using this tool, which it's helpful for teams especially. It's called TeamGPT, but there are others online. There's a TypingMind is another one. They basically use the APIs of companies like OpenAI or Anthropic. And you can have shared chats with your team so that everyone basically uses the same voice and tone. Everyone speaks to the same AI persona if you have that chat created and shared. So it's pretty handy. Like I use it with my junior copywriter to both research and write copy so that we can basically act as the same copywriter using our client's voice and tone based on all the research.

Uh-huh.

And so yeah, this is pretty much, I start with big context and then use the more like polished and refined LLMs for the writing and the simulation.

This is so valuable. Like, this is such a sort of insight into a real cutting-edge solution that you developed. So I appreciate you sharing that. That's amazing.

Of course. It was actually like a fun experiment that I tried once because I had a client that was pretty opinionated, right? You stumble on some of those that they have pretty strong opinions on what they want to write. And so I was okay with it because most of the stuff that he was saying was right. So I just basically needed to adapt my research to the way that they thought. So one thing that I did was simulate my client's persona and I fed the copy before sharing it with my clients. I shared it with this AI persona and it was pretty crazy to see that the responses that I got were pretty much aligned with what I got from the client after. So it's a good way to to, yeah, set your— set expectations for yourself and also prepare yourself in general for, I don't know, even just presentations that you need to run. It's a good playground.

Wow, that's awesome. That's awesome. So we could talk about this all day. You're a super expert and I love learning from you, but I got to be mindful of your time. But before I let you go, what's something that's, you know, sort of on the horizon for you that you're really excited about either with your business or personally?

Yeah, as I mentioned, I want to test those AI agents. I just finished working through a bootcamp, AI bootcamp that I studied like for a month. So I want to try and use those AI agents to kind of automate some of this process. But also there's a, there's a process called evals, evaluation for AI, which basically just means that after you get an output, another AI agent takes that output and evaluates it and then refeeds it to the AI. So we basically have this cycle acting between 2 AIs that continuously improve and that avoids all things like hallucinations. So basically AI shouting like random stuff at you. And yeah, in general, it can align better if you have voice brand guidelines on stone. So all of that, It's something that I'm pretty excited and curious to try and see if it really works for the creative work that we do, which actually has become a bit less creative and more of like an architect or like an engineer of sorts.

Wow, that's amazing. So for our audience that's been listening and is as blown away as I am with, you know, sort of your application of some of these amazing tools and, and the sort of approach and framework that you use, what's the best way for them to get in touch with you?

Yeah, so I have my website conversionalchemy.net where you can sign up for my newsletter. I also have a podcast which is all about message-market fit, and I'm generally on LinkedIn just chatting about this stuff, nerding about AI and all sorts of things. So you can find me there. Awesome.

Well, Chris, thank you so much for joining us today. I know I learned a lot. I'm sure our audience did too.

Thank you so much, Mark. It's been a pleasure and a super interesting chat.

Alright, talk to you soon.

Cheers.

Thank you for joining this episode of the B2B Growth Blueprint podcast. We hope you've gained insights to grow your business. If you enjoyed this discussion, please rate, comment, like on your favorite platform and share to others who will find it useful. Do you know a CEO, founder, or leader at an early-stage B2B services, tech, or SaaS company who would like to be a guest and share their story, send us a note. Let's schedule a conversation.