Scaling Content in the AI Era (How To Create Fast without Losing Your Soul) by Vassilena Valchanova

Lecture Slides

About the Speaker

---

Lecture Transcript

Speaker Introduction

[00:00:00] 

For those of you that don’t know her, I don’t know if there’s anyone that doesn’t, but Vassy is an expert in content marketing, brand positioning, copywriting over 50 years of 15, sorry, years. Not that old.

Imagine like everyone was going to be asking about your beauty procedures. Sorry. 15 years of experience in building comprehensive marketing strategies, content planning, creating text that reveal the true value of the brand. And she’s helped SaaS platforms with over 500,000 registered users and over 12 million in annual turnover.

And she works with agencies and experts offering B2B services. Definitely check her out, book her if you need her, and even if you don’t so who need to strengthen their role as opinion leaders in the field. You can join me while I share your fun [00:01:00] fact with the world.

So she actually gave me fun, two fun facts, which I both love. The first one is amazing for all the women in the room, if there are any that she, the certified wine expert. And we have actually seen these skills firsthand because she, every time I’ve had dinner with her, she just completely bedazzle all of the waiters and the people that know a ton about wine.

And she just outsmarts them in a very like impressive, but not at all overwhelming way, but very great to see. But the second most interesting fact is that she told her husband she won’t change her last name because the domain authority of her website valchanova.me is too high. It’s, we love an authoritative queen.

We support you.

Lecture by Vassilena Valchanova

Thank you. What I wanna talk to you about today is scaling content.

And the most important bit on this slide [00:02:00] is the last one aligned. So it’s underlined. It’s how to create strategic high quality content fast without losing your soul meaning without losing what makes your brand you, what makes you yourself recognizable, differentiated in the market, no matter if you’re working for yourself and building a personal brand or if you’re working for a company and building a corporate brand.

Just before I start, I’ll go through this really quickly. Don’t worry, there’s a slide in the end where you’ll see a link where you can download the slides and also a ton of other goodies that I have prepared for you. So I first wanna set the stage and tell you what the session is not going to be about.

I’m not gonna show you like complex N8N automations and workflows and something that has like a gajillion steps. The reason for that is that we think that automation in theory makes a lot of sense. We’re gonna create an automation, it’s gonna save us a ton of time. [00:03:00] In fact, we spent more time than we would’ve originally spent on creating the content in troubleshooting and maintaining a complex stack.

So I’m not gonna be talking about automations because for me personally and with for the clients I work with, simple setups are always better and make way more sense. The second thing that I’m not gonna share with you is what all these guys and gals are doing. We’re not gonna generate 2000 ideas and bajillion pages of a content calendar together, because that makes no sense.

You can’t create content without putting in your sweat equity, your opinions, your points of view. So no to that. And finally, I’m not gonna share with you the only unique and only right way of using AI for content creation. This is my way, this is what I’ve seen works for me and for my clients. So it’s based on experience.

But if you have a different setup and if it works for you, come [00:04:00] to me after the talk. We can talk wine, of course, but we can also talk about what’s your setup and what you’ve learned, and maybe I’ll learn something from you as well. So having that outta the way, what am I actually gonna talk to you about?

I have tried to edit myself into sharing with you three key. Mindsets that we need when creating content with ai because I can’t share everything I have tried in the past. It doesn’t work well, especially in 15 minutes, so that’s why you have the additional materials after the talk. But here I’m gonna share with you three specific mindsets to use when using AI to create content and scale your content.

The first one is the mindset of your setup. So how are you creating your AI workflows for maximum impact and maximum ease in keeping these up and integrating them within the company? Because there are still people who are afraid of ai and the simpler we make it, the easier it’s gonna be for us to [00:05:00] onboard more people.

What I use is very simple. I usually, when it comes to content writing specifically, Claude is my guy. I use Claude for writing. I think it does a way better job at interpreting brand voice and making content that actually feels human. But you can also use custom GPT or Gemini Gems, whatever works for you.

But this is a super simple setup. And when it comes to Claude what I do is I give it a set of very detailed instructions, and this is something I’m gonna talk about later. And I also give it a lot of additional materials, which I’m also gonna talk about in a bit. So nothing more complex than that.

And this works for 95% of the use cases I’ve seen. The thing that I see a lot of people try to do and fail to do is using a single project or a single custom GPT for all of their content tasks. And I [00:06:00] don’t think that works well, I’ve tried it. I initially, when I set up my first Claude project, it was for all my writing, so it was jam packed with hooks for LinkedIn posts and then signature content from my newsletter to give it an idea of what my newsletter editorials look like, et cetera, et cetera.

The thing is that it can’t keep up with all of these in the same place. So I prefer to set up individual tools, let’s call them like individual gems or individual Claude projects for the specific tasks that you have. In my personal setup, I’ve split this up according to the type of content I’m creating.

So I have one project for my newsletter. I have one project for my LinkedIn writing and so on and so forth. By the way, this is subliminal messaging Marketing Matters newsletter subscribe. Very important for some of the teams I work with, when they have more people working with content and they need to hand off processes from one person who writes the brief, [00:07:00] for example a product marketing person who defines their target ICP and then the key messaging pillars that they need to include in the content and then hands it off to their best buddy in the content marketing and content writing team, we would have individual tools for ideation and then individual tools for drafting and content writing and editing and polishing and so on and so forth.

So this can vary, but the important bit is try to stick with one thing per project or per custom GPT. And the other like big hack I use is what some people call knowledge blocks. So if you check out over here, you have in the cloud project a lot of additional resources that I’ve uploaded. So this is what you can think about as knowledge blocks.

These are additional materials that give your AI context as to whatever makes sense. So in most cases, this would include a company overview, some audience [00:08:00] persona details so you can actually tailor your content to the people you’re talking to. Guidelines on structure, your brand voice your key messaging pillar.

Words and phrases not to use because if I see delve into one more time, I swear to God. So I give it a blacklist of AI cliches that I don’t wanna use. And I also give it some of my existing content to give it an idea of what I used to write, like before I started using ai. So this is the most natural, unvarnished, like organic biodynamic version of my writing.

And I also give it access to something that’s called a scoring rubric. I’m not gonna spend time on this. In the additional materials what a scoring rubric actually is and how to set up one. Suffice to say it’s a great way to create an improvement loop with your AI where it can write something, then score it on its own and give you an improved version without you having to go through the process of [00:09:00] that.

So using knowledge blocks gives a lot of additional context to your AI and makes the output like significantly better. The second mindset I wanna talk about is how do you actually communicate with your ai? The art of having difficult conversations there are. Two things that I see has amazing impact on my writing.

And the first one is what you can call priming or warming up the algorithm or something like that. You can think about it like sharpening the axe before you actually start cutting the tree. And this is a way for you to give the specific chat within your project, some additional context on what you’re trying to do and what it needs to keep in mind when it helps you, do your individual marketing tasks.

This is especially useful when you’re first setting up the best practices for your writing. So what this comes down to is you [00:10:00] give it a role, you tell it you’ll be acting as this type of person. Now you add a lot of jargon and references to spark these semantic connections, giving the AI an idea about what you wanna talk about.

It’s not enough to just say, you are gonna help me, with email marketing. You tell it a lot of other things. I’m gonna show you an example in a second so that it just starts creating these connections and gives you a better output because it knows the context better. And then finally, you ask it what it really knows about the topic.

And frankly, I don’t really read most of the time what it actually says at this point. The reason you’re gonna do that is to give it a lot of ideas as to what you’ll be doing as next steps. So here’s an example. I know that you’re not hawkeyes over here, so I’m just gonna give you an overview of the best bits and you can see the full the full chat in the additional [00:11:00] resources.

In this case, I was trying to analyze the email nurturing sequence of a client. So I told it, you’re gonna be acting as a lead nurturing strategist. I used here a, I was about to say shit ton, but that’s probably not allowed. A ton of different words like elite scoring, a BM nurturing behavior based email targeting, deliverability, et cetera, et cetera, just to give it more words to work with.

And then I ask it like, what are the best practices? What are some common mistakes, et cetera, et cetera. Only after it’s provided like a really long laundry list of things, did I actually tell it? Here’s the nurturing sequence that I want us to analyze. Like how will we go about that and actually doing the work itself.

And another example here, just to show you that it doesn’t just need to be buzzwords. It can be stuff like concepts like in [00:12:00] this case it’s about positioning and differentiation. So Marty Meyer’s Onlyness statement is a common one when it comes to differentiation. So I give it bits and pieces to know what we’re talking about here and try it for yourselves.

I think you’re gonna like the results. And the next thing is how do we actually write when it comes to stuff that you’re gonna be using again and again, how do you write your master prompt? So the instructions within your project or within your custom GPT, this looks I don’t know, like the, the receipt from a bachelorette party I guess, or something like that.

It’s super, super long. And the reason for that is there are a lot of things you can add here that are gonna make the AI do a better job. I give it a role as I did in the past, and you’re gonna have access to a sample prompt to check out how I set things up. You give it the goal. So what’s the end result that you’re trying to have?

You [00:13:00] give it the constraints. So what should it do or should it not do? And this can be technical, keep the whole text to between 800-1000 words or something like that. Or it can be more specific write something that if a person who knows Vassy well reads, they’re gonna think that she wrote it without the help of AI or something like that, then I tell it about the external assets.

All of these knowledge blocks that we’ve created, the AI needs to know, what these things are and how to use them. So I give it a little bit of a reference guide, and then I provide the step by step. So here’s step one, step two, and step three of that process that we’re gonna be using again and again.

And when it comes to that step by step, this is the final third mindset that I wanted to talk to you about today. And yesterday at our speaker’s dinner, there was a common theme that came up, which was about slowness and slow living and being mindful and being in the moment [00:14:00] and doing things without rushing through them.

And I think when it comes to us working with ai, we can really learn from this guy. This guy is called Carl Petrini and he’s the founder of the Slow Food Movement. He first was an activist who protested setting up a McDonald’s in the mecca of Slow Food, which is Italy in Rome. And then this whole thing grew into a lot of other stuff, and it came to birth the Slow philosophy in general.

And I think this is a nice description of that also applies to our work with ai. So Carl Lenore, who wrote the book In Praise of Slow, says this, that the slow philosophy is all about the revolution against the notion that faster is always better. It’s not about doing everything at the snail space.

It’s about seeking to do everything at the right speed, doing it as well as possible instead of as fast as possible. [00:15:00] And I think that really applies to our work with ai. A lot of people think that when you work with ai, the goal is to save time. I think that’s not the point at all. I think the point is for us to create better content by always having at our side a trusted sidekick, a wingman as an handly, content marketing giant calls it, who can help us hone our ideas better.

And in between having one person on my team, a cheerleader who always says, that’s amazing, you did it so well. And someone who’s more skeptical and says, maybe we should rethink this. There’s a different way of doing it. I definitely choose the latter. I don’t know what you guys think, but that’s my choice.

I see a lot of people nodding. So I’m on the right spot. And this is the sort of short example, like I was struggling with an idea for my newsletter. I rattled off in my Claude [00:16:00] projects in audio form, like what I was thinking about and then said. You know what, like what we’ve done up till now, that’s not the best.

Like it’s a bit generic, it’s a bit vague, it’s one-on-one. People know this, so here’s what I’m actually, what I actually think you are seeing. And then it says, but I need your help. So we ended up doing a lot more back and forth to actually finally come to an idea that was cool and interesting. And you can read that newsletter edition on my website when you subscribe, of course, see that like always a call to action.

So thank you so much. I hope I made justice of the time that I was allotted, and I hope you like this. On this link you see a lot of additional materials. One sort of conference hack that I have, take a picture of this because when you’re browsing through your photos later, you’re gonna remember to actually open that link rather than just having it in your browser and sitting there.[00:17:00] 

And you see some of my favorite prompts that I reuse time and time again, a couple of videos with other workshops on AI that are a lot more hands-on than philosophical than this one. And also a special offer for my courses, which are currently discounted because I’m celebrating three years of actually launching my first solo course.

Thank you. And you’re actually on top of the discount. You’re gonna get a two week AI writing bootcamp that I’m gonna do in January. So this is something I haven’t announced anywhere yet. You are the first people to hear about this. I think it’s gonna be tons of fun. But I’ll let you figure it out for yourselves.

So thank you so much.