A Marketer's Journey:
Embracing AI in the SaaS World
For non-technical marketers wondering how to actually use AI without feeling overwhelmed — permission to go slow and focus on what moves the needle.
Meet Anna Nadeina
Anna is the Head of Growth at SaaS Group, an investment company that acquires and grows B2B SaaS businesses. She's also the host of SaaS Unbound, a podcast with nearly 300 episodes featuring founders who've scaled from zero to exit.
Like many of us, Anna doesn't have a technical background. She's a marketer through and through — she even left marketing briefly to become a chef before coming back. Her journey with AI started not with ChatGPT, but with Midjourney, when she wanted to create unique branded visuals for SaaS Group's content.
What makes Anna's perspective valuable? She's been watching dozens of SaaS teams discover and adopt AI tools, while navigating her own journey as a non-technical operator. Her core message cuts through the noise: you don't need to master everything — give yourself permission to go slow.
The AI Tools That Actually Stuck
Anna's approach to AI tools is refreshingly practical: use what works, ignore the rest. After trying countless tools, here's what made it into her daily workflow:
"I would never go back to just manually searching on the internet. Obviously, that helps a lot. The other one is editing the podcast. Riverside FM... it's got amazing AI features. You know, it offers you the shorts that you can later leverage for YouTube."
For Podcast Production: Riverside + Descript
Anna calls this combination a game-changer for non-video-editors. Riverside handles recording and auto-generates shorts for YouTube. Descript lets you edit video by editing the transcript — no timeline scrubbing required.
For Quick MVPs: Lovable
When Anna needs to visualize an idea or create a lead magnet, she turns to Lovable. It's become her go-to for showing stakeholders what something could look like — without waiting for developers.
"It's amazing for me how easy it is to just go to Lovable and do a quick project. You know, this is how an alternative page on the website could look like. This is how an invitation to an event could look like. It gives you so much freedom to just create that MVP and just show what you mean."
For Content Synthesis: NotebookLM
Google's NotebookLM became Anna's secret weapon for processing large amounts of content. She fed it all the transcripts from an M&A course SaaS Group spent a year creating, and it generated an incredible overview video with narration.
For Daily Work: ChatGPT Projects
Anna uses ChatGPT's Projects feature to organize different contexts — one project for SaaS Group marketing, another for research. She uploads documents, past wins, and constraints to give the AI the context it needs.
"I have a SaaS Group Marketing Assistant that helps me brainstorm. I have a SaaS Group Researcher that helps me research. It's different set of specifications and asks that I have there. And then it's just also easier to maintain."
The Intern, Not the Friend
One of the most practical insights from our conversation was how Anna approaches prompting. She doesn't treat AI as a magic oracle — she treats it like an intern who needs clear instructions.
"Hopefully we all learned that it needs context, needs a lot of it. The more, the clearer it is, the better. My hack became going to documentation and reading how exactly I should do my prompt in order for the model to understand me."
Her approach is surprisingly simple: she keeps just 7 prompts in a Google Doc that's always open in a browser tab. That's it. No complex systems, no prompt management tools — just what actually works.
When building a new ChatGPT project, Anna uploads:
- Documents showing what a "win" looked like before
- Examples of good content that represent the voice she wants
- Constraints and where this idea is going (podcast, pitch, blog)
- Specific metrics — "I want 10 founder events, not 8, not 9"
"If you were going to London on a trip, you would probably tell your assistant when you're going, who you're going with, how many days, what airlines you prefer, what area of the city you would like to stay in. If you're vegan, so they don't suggest you steak houses. Basically the same thing works for any LLM."
Side note: Anna mentioned she works "in tabs, not in apps" — keeping Slack, Google Docs, and everything else in browser tabs. When her developer boyfriend suggested closing tabs might fix a website issue, she responded: "That is not allowed."
When Vibe Coding Hits Its Limits
Anna's biggest vibe coding project was an M&A data room built in Lovable. For context: when SaaS companies sell, they need a secure place to store sensitive financials and documents. Usually this is just Google Drive.
Anna wanted something better — with progress tracking, checklists, collaboration features, and benchmarks. She spent significant time building it, debugging countless issues, and got it working.
"It took quite some time and it was so much debugging. It was absolutely crazy. But at the same time, when Lovable says 'okay, this is not working, do you want me to configure it?' And you're like, yes, do that. And everything broke at some point, and then everything got back together."
But here's the lesson: they couldn't use it in production. For a data room handling sensitive M&A documents, security isn't optional. And with AI-generated code, nobody knows exactly what's happening on the backend.
"Each and every developer said, 'It's great, looks amazing, and it even works. But nobody knows what's on the backend. And if something breaks, if something really breaks, you'll have to find a very expensive developer to dig through all the thousands of lines of code that are there that you know zilch about.'"
Anna's conclusion: use Lovable for landing pages, lead magnets, and internal tools — not for security-critical applications. For her personal project, she's building the landing page in Lovable but using a real developer for the actual product.
The right question isn't "can I build this with AI?" It's "should I build this with AI?"
Permission to Go Slow
Here's where Anna's perspective diverges from the AI hype. When I asked about feeling behind — even Andrej Karpathy admits to feeling this way — Anna had a different take.
"I have a job and I want to do that well. Most of what I do is relationship based, brand awareness based. You don't write a multi-million dollar check after a cold email that was written by AI. There is no way."
SaaS Group started their AI initiative by telling the team "we're already behind" — and it backfired. Nobody understood who was ahead, why, or what they should actually do. It just created stress.
"If you're just trying to figure out what else I can automate and what else I can do better or faster with AI, it's going to be what gives you a burnout. If we give ourselves a little bit of breathing room, it will actually create more benefits."
Anna's prescription? A tool cleanse:
"I would recommend to step back and look at the tools that you're using and do a bit of a cleanse. Only keep those that you need, only keep those that give you value already and double down on those. Strip off everything else, all the noise."
At one point, Anna's thousand browser tabs became two thousand tabs as she subscribed to every new AI tool. The cleanse was necessary for her sanity — and probably for her laptop's performance.
The Human Edge
Anna shared a story about someone doing AI-powered ghost writing who deliberately adds spelling mistakes to posts. Why? Because people see mistakes and think it's human. Anna's response: "Let me make my own mistakes."
Her view on what remains irreplaceable:
"People will still crave humans. We want to substitute you as this little hamster trying to edit, distribute, write posts, write blog posts, make sure your transcripts are right. But taking away the human completely? I don't know."
For Anna, the path forward is clear: use AI for the "hamster work" (editing, distribution, research, repetitive tasks) while doubling down on the uniquely human elements — building relationships, creating authentic brand experiences, and being present in ways that AI cannot replicate.
The goal isn't to automate everything. It's to automate enough that you have time for the work that matters.
About This Conversation
Anna leads growth at SaaS Group, an investment company that acquires and grows B2B SaaS businesses. She's also the host of SaaS Unbound, a podcast with nearly 300 episodes featuring founders who've scaled from zero to exit. A marketer through and through, Anna brings a refreshingly practical perspective on AI adoption.
Ready to Go Deeper?
Builder Lab
·Wed, Feb 18 · 18:00 CETFree weekly online session where we work on real projects together. Bring what you're building and get live feedback.
Builder Forge
·In 59 days · Wed, Apr 16-week cohort program to build and ship your first product. Weekly workshops, accountability, and hands-on guidance.
Growth Lab
·Wed, Feb 11 · 18:00 CETFree weekly online session where we work on funnels and positioning together. Bring your challenges, get live feedback.
Growth Forge
·In 17 days · Wed, Feb 186-week cohort program to build predictable, scalable revenue. Weekly workshops, accountability, and hands-on guidance.