Founding a Company at 25: What Building Impremis Taught Me About Risk

Everyone told me I was taking a risk.

Leave a stable career path, start an agency with no institutional backing, and compete against established shops with years of client relationships and operational infrastructure. The conventional framing was that I was gambling, that I was young enough to afford a failure, and that the downside was manageable because I had nothing to lose.

That framing was wrong on almost every level.

What I actually learned building Impremis from the ground up is that the popular conception of entrepreneurial risk is almost entirely backwards. The decisions that feel risky rarely are. The ones that feel safe are often the most dangerous. And the thing most young founders fear — failure — is not the real risk at all.

Here's what I actually learned, and why I think about risk differently now than I did at 25.

Image brief: Two parallel timelines — Fast feedback (low real risk) showing decision → 72hr result; Delayed feedback (high real risk) showing decision → 12mo compounding consequence. alt: "Fast vs. delayed feedback decisions." caption: "Risk isn't variance. It's the distance between a decision and its consequences."

The risk nobody talks about

When people describe the risks of starting a company, they talk about financial risk. Running out of money. Not making payroll. Losing clients. These are real and they deserve serious planning. I'm not dismissing them.

But the risk that actually kept me up at night — and the one I think is most underappreciated by people who haven't done this — is the risk of building something that doesn't reflect what you actually believe.

In the early days of Impremis, I made decisions based on what I thought clients wanted to hear, what I thought would make us sound credible, and what I thought a "real agency" was supposed to look like. I positioned us broadly because I was afraid that specialization would limit us. I agreed to scope that was outside our core competency because I was afraid to say no to revenue. I hired for familiarity rather than capability because I was afraid of bringing on people I couldn't fully evaluate.

None of those decisions felt risky in the moment. They all felt like the safe, reasonable, practical choice.

Every single one of them caused problems that cost time, money, and in some cases, client relationships to fix.

The real risk in founding a company is not that it fails. It's that it succeeds as something you didn't intend to build.

What risk actually looks like when you're operating

The textbook definition of entrepreneurial risk is variance in outcomes. High risk means high variance. Low risk means predictable, stable outcomes.

That definition is useful in an investment portfolio. It's almost useless for an operator making daily decisions in a growing business.

What I discovered: risk, from an operator's perspective, is better understood as the distance between a decision and its consequences.

Some decisions produce feedback quickly. You launch a campaign, you see the results in 72 hours, you adjust. The feedback loop is tight. Even if the decision was wrong, the cost of being wrong is low and the correction is fast.

Other decisions produce feedback slowly. You hire the wrong person, and it takes six months to recognize the full cost of that mistake. You agree to a client relationship with misaligned expectations, and it takes a year before the relationship deteriorates to the point of forcing a resolution. You choose a positioning strategy that limits your growth trajectory, and you don't see the ceiling until you've been running into it for eighteen months.

The truly dangerous decisions in building a business are not the ones with high-variance outcomes. They're the ones with delayed feedback loops and compounding consequences.

Three decisions that taught me the most

Deciding who we were not

In the first year of Impremis, we pitched and won work across categories that had nothing to do with performance marketing. We were doing general consulting, content strategy engagements, and one-off projects that filled revenue but muddied our identity.

The decision to stop was uncomfortable. Turning down revenue when you're early-stage feels genuinely dangerous. It felt like choosing to be smaller.

What actually happened was the opposite. Narrowing our focus sharpened our positioning, made our sales conversations easier, improved our work quality because we were operating in our area of genuine expertise, and made it possible to hire people who were specifically great at what we actually did.

The risk of saying yes to everything was that we would never become excellent at anything.

Firing a client

The first time I ended a client relationship, I lost sleep over it for two weeks before the conversation.

The client was not a bad person. They were a poor fit. The expectations were misaligned from the beginning. The account required disproportionate management time relative to its revenue. And the work environment it created for my team was genuinely corrosive to morale and focus.

The financial risk of ending the relationship felt enormous. The math was simple and terrifying: this is X percent of our monthly revenue, and we're going to walk away from it.

What I didn't fully account for was the hidden cost of keeping it. The senior team member was increasingly disengaged. The client conversations that occupied mental bandwidth should have gone toward accounts we could actually impact. The institutional energy being drained by a relationship we all knew was not working.

When we ended the relationship, the team's output across all remaining accounts improved noticeably within 30 days. The revenue gap closed faster than I expected because we were no longer distracted.

The risk of keeping a bad client is not visible in the revenue line. It lives in everything else.

Hiring someone better than me

At some point in building Impremis, I had to bring on people who were more technically skilled in specific areas than I was.

That sounds obvious. Most founders will tell you hiring people smarter than themselves is the goal. In practice, it is deeply uncomfortable when you're doing it for the first time.

Hiring someone who knows more about media buying than you do, who has a more sophisticated creative instinct, or who manages client relationships with more natural ease, creates a specific kind of founder anxiety. The fear is not that they'll do the job poorly. It's that their excellence will reveal the limits of your own.

Working through that fear was one of the most important professional developments I experienced in building this company. The business only gets better when the people inside it are better than the person who started it. (How I hire for the role now reflects that.)

How founding a company restructured my relationship with uncertainty

Before Impremis, I had a normal person's relationship with uncertainty. I wanted to know outcomes in advance. I wanted plans to work as designed. I was uncomfortable when the path forward was unclear.

Running a business cured most of that.

Not because I became more comfortable with chaos, but because I learned that uncertainty is the medium you operate in, not the condition you escape from. The question is never whether things will be uncertain. The question is whether you've built the judgment to navigate uncertainty well.

That judgment comes from the volume of decisions over time. From making a call, watching what happens, and updating your model accordingly. From being wrong often enough and recovering from it quickly enough that you stop treating individual failures as existential events.

The founders I know who seem most capable and most calm under pressure are not the ones who have avoided failure. They're the ones who have been through enough of it that they have a realistic internal model of what failure actually costs and how recoverable most of it actually is.

What I would tell someone starting at 25 today

Not advice. Observations.

The window you have right now to take on concentrated risk is real, and it is not because failure is cost-free. It is because the asymmetry between what you can build and what you stand to lose is genuinely different at 25 than it will be at 35 with a mortgage, dependents, and a higher baseline lifestyle.

The market for performance marketing skills is large, global, and growing. If Impremis had failed in year one, I would have been a 26-year-old with hands-on operator experience running paid media at scale. That's not a bad starting point for what comes next.

The real window is not about having nothing to lose. It's about the cost of being wrong being lower than it will ever be again, while the potential upside is as large as it will ever be.

Do not waste that asymmetry on something that does not genuinely interest you. The difficulty of building a company from scratch requires a level of intrinsic motivation that external rewards alone will not sustain through the hard periods.

Build something you would work on for free for the first six months if you had to. Because you probably will.

FAQ

What's the biggest risk most young founders ignore? Delayed-feedback decisions. The wrong hire. The wrong client. The wrong positioning. The bill comes due 6–18 months later.

Should you take outside investment when you're young? Only if the capital genuinely accelerates a clear path that you couldn't reach otherwise. Capital obscures bad decisions. Bootstrapping forces clarity.

How do you know when to fire a client? When the cost of keeping them — measured in team morale, distracted attention, and management time — clearly exceeds the revenue. The math always says stay. The reality usually says go.

Was 25 too young to start an agency? For me, no. The window of low downside / high upside is real. But it's not about age — it's about the asymmetry of your specific situation.

Closing

Risk does not have an objective value. It is a perception shaped by the information you have, the models you carry, and the experiences you have learned from.

Every time I made a decision at Impremis that felt risky and then watched the consequences unfold, I updated my model. Some things I feared were genuinely dangerous. Many things I feared were not. And several things I did not think to fear turned out to be the actual problems.

That process of calibration is the real work of being a founder. Not eliminating risk. Not avoiding uncertainty. Getting better, iteration by iteration, at understanding what actually matters and what doesn't.

You build that calibration by doing the work, making the decisions, and staying in the game long enough for the feedback to arrive.

That's the whole thing. Stay in the game.

Keep reading

Pieces I've written on related topics that pair well with this one:

• What Scaling Past $1M/Mo on Meta Taught Me About the Algorithm — Lessons from scaling Meta ad spend past $1M/month — creative structure, algorithm behavior, attribution at scale,
• $1,000,000 in Learnings from a 30-Year-Old Entrepreneur — What building a dog food subscription business taught me about partnerships, money, and reality.
• First-Party Data as a Competitive Moat: How DTC Brands Are Building Audience Infrastructure — First-party data is the moat most DTC brands aren't building.