How this creative studio uses AI to do better, not more

_{⏩ Forward this email to: someone on a lean creative team}

👋 Hi, I’m Amanda. I’m a fractional head of growth with a decade of experience working with rapidly-growing B2C startups. I work with lean teams in a few ways:

✅ Fractional: You have an expertise gap on your team you need to fill with someone who can flex into both strategy and execution, for an open-ended time frame.

✅ Advisory: You’re either just starting out or need to validate your existing approach, and need someone with growth expertise to guide you.

✅ Project-based: You have a specific scope and deliverable in mind, and need someone experienced to execute quickly and cleanly.

If you’d like to explore working together, start here to tell me a bit about what you’re looking for. If you’re not quite ready but have growth questions you want to chat about, you can book office hours here.

This week, I spoke with Matthew Gattozzi, founder of Goodo Studios, a creative studio focused on creating ads for consumer brands. His approach is to connect creative strategy and production, using AI only where it makes sense for the business, to create ads that convert viewers into profit for brands.

We talked about:

• 🏆 When AI is the best tool for the job, versus when it’s not

• 🎨 How to think about using AI for creative, from an honest standpoint

• 🎥 How his team uses tools like ElevenLabs and Adobe Premiere in their development process

• ✏️ How a physical pen and paper (really!) are foundational to your AI tech stack

One of my favorite parts about this conversation is Matt’s clear framework for assessing when to use AI. Spoiler alert: “Because everyone else is” or “It just feels like I should” isn’t a good reason — instead, Matt asks a simple question: is this going to help get us the best possible outcome, more efficiently?

Our conversation, edited for clarity, is below.

Amanda Berg
Could you share a little bit about who you are, what you do, and what your journey into incorporating AI more into your day to day has looked like?

Matthew Gattozzi
I’m the founder of Goodo Studios, a content and creative studio that works with consumer brands on converting viewers into profit. Our whole thesis is that we want creative strategy and production living in one place and working together as closely as possible, to create the most effective content for consumer brands. The old way is having a big agency overseeing strategy, and a separate production company doing the creative work. And those two incentives are very different. 

The rise of social media opened up the ability for a creative person like myself to make a living with a camera, and that business like food trucks needed Instagram content just as much as massive brands like Nike.

I realized that I needed to bridge that gap between marketing and production. The pandemic and the rise of TikTok supercharged the need for creative for consumer brands, because these things changed where advertising is consumed — primarily in a video format.

Where AI comes in is that we operate in a performance landscape for our clients, so anytime we can leverage AI to get deeper insights to make the ads more effective, the better. For us, the use case for AI is more so in insights than it is for production.

What I care about is, how can we make the best ads? And if that’s using a cinema camera or using AI to generate an image, I don't care. I just care about outcomes for our clients, and I will use whatever tools we have at our disposal. I think a camera still works just fine, but I think if that changes, I'm okay with changing that, too.

Amanda
What did a typical day look like for you and your team before you started integrating AI into your workflow? When did you start making this shift?

Matthew
One of the biggest lifts that goes into producing these ads is the strategy behind them.

When we initially take on a client, we're doing everything from researching the product, their competitors, looking at reviews, to really trying to understand the psychographics of the consumers. There's a ton of research that has to get done when we first start, and as recently as two years ago, all of that was manual.

In between the steps of doing the research and documenting it all, all that info would live in a strategist’s head. The initial searching, building of tables, reviewing competitive ads, etcetera was done by someone and then synthesized into a doc. So that takes time and really limits the number of clients one person could take on.

Then, you’ve got this rapid feedback loop of data from Meta and Youtube ads, so every week we’re coming up with new ad iterations. We're trying to find new headlines, new statics, and testing on a very deep, detailed level. And in writing that brief each time, you’re starting from a blank page. So there was a lot of mental and manual load there.

Separately, because we own the production and editing process, there are situations where a creator would send video back to us and they’ve misread a line, or a brand name hasn’t been pronounced right. So we’d need to go back and forth, rebook talent, and get video re-recorded. You can see how all of this process limits what we could accomplish in a day.

Then around two years ago, we started using ChatGPT to analyze a CSV of an ecom site to pull out key themes, instead of reading the whole site manually. And slowly but surely, as these models have improved, we can use AI to search for things like competitors and UVPs. This just means you can get up to speed that much faster.

We were using Claude and some other tools to help us with writing and research last year, but it wasn't until 2025 where we were clear with everyone on the team that we are AI-first, because the tools are finally where we need them to be to give us an advantage. Two years ago there was still some hesitancy, but now it feels like everyone is on board.

Amanda
Aside from the upleveling that AI has been able to bring to the research and editing processes for your team, are you using AI in the actual creative development and content generation process? How do you think about the balance of what should be AI generated, versus what should start with a camera?

Matthew
One thing that we’re working on that’s not quite there yet is the use of AI greenscreen technology. Right now, a green screen can fall short when the background image doesn’t match the lighting of the person or elements you’re actually filming with. Today, when we need a person talking in a living room, we have to go rent that living room space. What we’re starting to build out is the ability to create that background space with lighting that matches the person you’re filming exactly.

Of course this technology already exists, but it requires a ton of CGI investment. If you can pull this off with AI, and can generate, for example, an upscale kitchen background for a high AOV customer without having to rent that space, that opens up a ton of possibilities. The future state is having a full “green screen studio” where you’re able to film entire videos without actually having to rent a house.

We’ve not yet done AI generated people or avatars or anything like that, because I don’t think the technology is there yet. Some people may say it is with tools like Google Veo3, which is a few hundred bucks a month. But it takes a really long time to generate those videos. And if you need to make edits and re-prompt, it can take a week to get the 5 seconds of video you need, so that’s not an efficient use case compared to a person who can do that work much more quickly.

So for us, using these tools is not so much a philosophical decision, but more about whether it’s actually efficient and helps us.

Amanda
As a creative yourself, do you have a POV on where to draw the line between what AI can create versus what humans should create?

Matthew
These are just tools we can use to create. Personally, I have digital cameras, but I still love shooting film —there's something beautiful about it. And film is a tool, it gives you a different aesthetic that digital just can't get. Where the line comes in is whether you’re being honest. You can't have an AI avatar say, “I lost 100 pounds using this supplement.” But an AI avatar can absolutely help create a new way to communicate your product of your brand.

So if these tools improve to allow people to be creative who otherwise may not be, and that barrier to entry doesn't exist anymore, I think that's a beautiful thing and we should embrace that.

Amanda
You talked a little bit about using AI to help with some post-production woes, like if somebody mispronounces a brand name. Can you share more about how you’re using AI in post-production to speed up the process?

Matthew
We use Eleven Labs to help us with voiceovers. If a client comes back to us and needs us to update or change a line, or they have a new idea they want us to incorporate, we don’t have to wait to get a voiceover actor, we can just generate a new voice.

So all of a sudden the speed from idea to execution is so much faster, because we already have all the footage we need. The barrier used to be the voiceover, whether we had an actor or were doing it ourselves, but now our editors can just input a script, make sure it sounds great, then export the content into Adobe Premiere and finalize the ad.

There are other ways AI in Premiere has helped us with sound design too, where we can make sure music sounds seamless without us needing to make the decisions and doing the technical work of where to cut. Auto-captioning is also so much stronger now, as is tagging clips, so our editors can just search for what they need rather than digging through 300 different clips to find what they’re looking for.

All these features Adobe is coming out with are crazy and have saved us so much time, which means ads can be more effective, and that gives us space to do better work. Most people think the time saved means we can produce more output, but the way we look at it is that we can significantly increase the quality of our work.

Amanda
Are you able to quantify how much time do you feel like your current process has saved your team? Tell me more about how you think about using that time to be more productive, whatever that means to your team.

Matthew
It’s so hard to quantify time, but it’s probably over a hundred hours.

In terms of how we think about it, it’s not like, okay, now people only work 10 hours a week. It's that instead of doing manual work that can be automated, now they get to work on more strategic things. I believe everybody in the world will become creative, because everything else is going to be automated at one point.

It’s going to be less about “can we get that data into an Excel sheet?” and more “how do we interpret this data and what actions do we take as a result?” Opening up that higher level, higher quality thinking is a way better use of a strategist’s time than putting data into a table to present to a client.

Amanda
From an organizational level, were there any aspects of AI adoption across your team that have been messy or haven't gone as smoothly as you would have hoped?

Matthew
I think that there’s this desire to automate things, but the reality is to automate something, you have to know your process first. When the SOP isn’t clear in the first place, that’s where I see people getting tripped up.

The other thing is that when people do get excited about automation, they want to automate everything. But maybe you don’t need that fancy Slack notification in the first place. So before we look to automate things, we look to eliminate things.

Also, I acknowledge that not everyone uses AI the exact same way I use AI, so we need to create flexibility with the tools that we use and how we’re using them. The hard part is that these tools are changing so fast, and it’s challenging for your tech stack to keep up. You used to have a bit more time to compare your options when buying SaaS, and now it feels like you have to move really quickly, but also avoid continually re-building processes you’ve already invested in automating, so we have to strike that balance.

Amanda
Everyone using AI in different ways is a great point. Most people are thinking about AI as this defined thing that exists in a box, when in reality AI is starting to have the ubiquity of “the internet.” And in the same way that how I use the internet is not going to be the exact same as the way any other person on earth uses the internet, the same can be said of AI. It sounds like that presents new problems organizationally to figure out what your AI tech stack looks like, at least in the current state with the number of tools out there.  

Even with the number of tools we have at our disposal now, I’m curious whether there have been any specific work streams or processes that you've tried to apply AI to that just haven't been a good fit for any reason.

Matthew
In terms of ad creation, I think there are a lot of tools that overpromise right now. Even if you have the perfect prompt and the output was exactly what you wanted, you end up with text baked onto an image. What if I want to change that text? I don’t have a file where I’m able do that, so there are some things that just don’t completely work yet. Sure there are ways around this, but you have to diverge from your existing workflow or add steps, which is inefficient.

I also think there’s a lack of good qualitative analysis from performance data, which is what really moves the needle. An AI agent can tell you which ad is your top performer, but they can’t watch the video so they don’t know why. They don’t know why you made this particular ad in the first place. So agents that can identify metrics for you aren’t super helpful, because we can just identify those metrics ourselves.

Amanda
How have you handled adoption of AI tools at your org such that the human elements of creative work, and focusing on the ability to do that deeper work, are still respected?

Matthew
Because our whole approach is to tie strategy and production so closely together, our strategists are the ones directing the shoot. That’s what differentiates us, and because they’re doing all the ideation, directing, and creation, they know that’s not going away in the near term. If your creative strategists are essentially just sitting behind a computer, writing and coordinating and reviewing content, a lot of that can be automated. For us, that automation is a good thing, because then we can focus on creating the best possible content and making it stand out and feel human.

Amanda
If you were advising a marketing team of one or a creative team of one, which is not super uncommon at small startups, is there a particular AI tech stack or starting point that you would recommend to them?

Matthew
Claude is the best writer, so I would start there. You could toggle between Claude, Perplexity, Grok, and ChatGPT for more searches and queries, it just depends how much you want to spend to get the best of everything. Where I’d really start is by getting used to using AI in everything you do. Next time you need to respond to an email, put that info into AI and get used to prompting it the right way.

I wouldn’t worry too much about automations and building out these crazy workflows at the start, because it’s so much more important to get comfortable with just using AI as a tool. Feedback is instant, so you can learn how to improve your prompting quickly.

And when you are ready to start automating, go back to pen and paper or post its and write out your process and your SOPs. You’ll either figure out what you can eliminate, or you’ll figure out what you can get ChatGPT or Claude to do — and there’s no cost to pen and paper plus an LLM being your tech stack.

In my case, I need to know what makes a great ad. So what I would do is spend some time creating a document that says, this is what a great static ad is, this is what a great video ad is. If you spend that time upfront to define what success looks like, the more time you’ll end up saving. If you get a bad output from AI, it’s because your input isn’t great. Focusing your time there instead of trying to be at the cutting edge of the latest automation will be much more valuable.

Amanda
Any closing thoughts that you want to share?

Matthew
Ultimately, the most important question to answer is: why use AI for any given task? If it’s FOMO or it’s just to be able to say you have this AI agentic workflow, you’ve probably missed the point for your business, which is to focus on driving the most revenue and profit in the most efficient way possible. Consider your main goals, ask whether AI allows you to do that, and only use it if the answer is yes.

I recently had a photoshoot where I was able to tell my photographer and model in real time how to adjust, and had tons of great content for the client to review. It would have taken so much longer for me to prompt all of that, the packaging wouldn’t have looked right, and that’s not what would have been best for our client. A real photo is actually the best output possible right now, so it’s not worth the tradeoff.

As a leader, I want to encourage innovation and encourage people to use these tools, but I also need to encourage that principled thinking so we don’t feel like we’re missing anything when we choose not to use AI. It’s all changing so quickly, and as long as we’re continuing to grow our business and our clients’ businesses, that’s what matters.

💡 Are you a marketer using (or figuring out how to use) AI in your work? I’d love to chat with you for the newsletter. Just hit reply and give me a 1 - 2 sentence summary of your current relationship to AI.

Was this message forwarded to you? Need a little Growth Therapy? Sign up below 👇

Questions? Comments? Topic requests? Just hit reply ↩