The TikTok question that turned into my startup

Key takeaways

• Founders who can't hire a product marketer yet are usually short on one of two things: knowledge or time.
• AI removed the barrier that kept me from building: I could find out whether an idea held up without raising money first.
• I'll share the real numbers in this series as they come in, good or bad.

On this page⌄

1. Who gets left out
2. Why I finally built something
3. What I wanted AI to do
4. How I'll know if it's working
5. The marble

A note before we start: I draft these articles with an AI writing assistant and edit them myself.

At the end of 2025, a founder found me on LinkedIn and invited me to one of WeWork's regular social events. Once people heard I'd spent my career in product marketing, I ended up talking with a few founders who all wanted the same thing from me: the most effective way to launch their product. And they wanted it right then and there.

I asked them a question back. "OK, now that you've all figured out how to build a product, how would you build a product to rival TikTok right now? Go."

Each of them started somewhere different. One went straight to what the backend would need. Another started on the front end. Then they started asking me questions. Instead of cutting in, I let the conversation run and slid my point in where it fit.

Launching a product works the same way. There was still so much I didn't know about their products that I couldn't give any of them a straight answer on how to get there. I could give them a blueprint, or even a paint-by-numbers, but they'd still have to fill in some of the details to make the launch fit them. No two launches are the same, because no two products are the same, and neither are the parts that make them up, including the people. All of that has to be taken into account, and we weren't going to figure it out at a two-hour social event.

No matter how many years of experience a product marketer has, they can't hand you a go-to-market plan without answering a lot of questions first. My GTM process works a lot like a choose-your-own-adventure book. The decisions you make send you down different paths. The more questions you can answer, and the more clearly you can answer them, the better your ending. Until we have those answers, neither of us knows which path gets you the awesome sword and the pegasus… I mean, more attention and more signups on day one.

Who gets left out

That evening reminded me of something I've known for years but had only been able to act on in small ways. Most founders don't know much about launches or marketing, and they shouldn't have to. Companies that need marketing should have people who do it for them. Not all of them can, though, and it usually comes down to stage.

That leaves the founders who can't hire a product marketer yet to launch on their own, with a small marketing team or none at all. I can tell you how that launch is likely to go. Not well.

There are too many moving parts in a go-to-market to say "here are the two most important things you can do." That list doesn't exist. Marketing is hard, and you can't just pluck someone off the street to "do the marketing." The founders and small teams I meet in this spot are usually short on one of two things.

The first is knowledge. Solopreneurs, founders and marketers thrust into their first marketing role often don't know where to start, how much time marketing takes or how much of it comes down to showing up and doing the work, week after week.

The second is time. Even a team that knows what to do has to keep up with industry news and draft blog posts that fit the season and the company's goals. Someone also has to train new teammates and outside vendors. After all that, there are still a dozen other things that needed to get done yesterday.

Why I finally built something

I've always wanted to build my own startup, and I'm constantly thinking up ideas. What held me back were the usual questions. How sure was I about my ideas? Where would the funding come from? Where would I find investors, and who would manage them? I let those questions become a limitation. I never should have, but we all have our weak spots.

What took away most of that barrier was AI. Once tools like ChatGPT and Claude were widely available, one person could do a lot more on their own. I didn't have to raise money and hire a team just to find out whether an idea held up. I could build the first version myself.

I jotted down about 20 ideas, and a few of them still interest me. To pick one, I used two tests. Does it meet a need people can already see in the market? And is it something I know well or am eager to learn? Product marketing, the thing I've dedicated my career to, passed both easily.

I'd been circling this one for a long time. My consulting firm, The PMM Playbook, grew out of the process I'd used throughout my career. I already knew I wanted to build something around that process, something that would bring it to founders and solopreneurs and automate as much of it as possible. That night at WeWork showed me what it could be: the idea that became Smarkly. My gears were turning, and I stayed up late into the night jotting down ideas and sketching out how it might come together.

For the next four months, I helped that founder's small team launch their product while I worked on Smarkly. Working inside their team, I kept finding parts of the work that AI could make more efficient. Having access to WeWork helped, too. I kept meeting entrepreneurs who fit my ideal customer profile, and they were open to telling me where their launches got hard. In return, I answered their launch and marketing questions wherever I could.

What I wanted AI to do

Those conversations made it clear what I wanted to build: a launch automation tool. Launches come in different sizes, but the work inside them is roughly the same.

The limit was customer research. You can make assumptions about a lot of things, but only real conversations tell you whether those assumptions are right. Much of what makes a product marketer effective is the ability to ask questions, then ask the follow-up that opens up insights the other person wouldn't have shared otherwise. That's hard to automate.

So I tried to automate as much of the rest of my process as I could. Then I looked at what else could come off a busy product marketer's list, and a lot of it overlapped with the marketing team's work. It was the same work I described earlier: the news monitoring, the blog drafts and the training that eat a team's week.

To me, that's the promise of AI. It clears away the noise so people can focus on what matters. A marketer who isn't stuck reading industry reports or drafting and planning social posts has time for the strategic side: figuring out what's working, digging into campaign results and deciding what's next.

The Smarkly I have today looks nothing like what I sketched that night. It grew from a launch tool into a marketing co-pilot. It drafts content for you, suggests relevant topics and organizes your team on one shared calendar. That frees a product marketer to focus on leading a team through a successful launch. It can help a marketing manager put together the content for their blog and emails, so they can focus on how well their marketing is performing and build new campaigns from those insights.

I believe I've turned my vision into an app, and I also know there's a lot more to do to make it the right app.

How I'll know if it's working

I try to stay optimistic. You have a chance until the bell rings or the clock hits zero.

The businessman in me says my clock is year one. If there are no positive growth signs or revenue by then, that tells me something. The marketer and builder in me pushes back, because there could always be one more change that flips the switch. That moment rarely comes, though, and sometimes you have to call it quits. So here's my deal with you: I'll share the real numbers in this series as they come in, good or bad, and you can judge along with me.

The marble

One of the hardest parts of being a founder is being patient. Your startup is a big Rube Goldberg machine. Sometimes the boot has to kick the ball into the basket, and the basket has to drop before the race cars take off. I'm only at the beginning. I've barely let go of the marble that starts the whole thing.

Next time: what I learned building Smarkly with AI as my engineer, mistakes included.

Share this post

LinkedInXEmailCopy link