Ben Ornstein

Startup ops & tennis content

FJ Podcast Producer: of The Draw: with @Gill_Gross

[Ben Ornstein @benorn7]

Almost everything an operator does can be automated. Except context.

Crazy statement on its own, I know. Give me a chance to explain.

You can put a lot into a prompt. You can describe the business, the strategy, the goals. But you can't transfer what you know about the people, at least easily. The unspoken dynamics in a client relationship. The thing your teammate said on a call once that changed how I approached the next six months of decisions… there are so many of these little things.

Tools help… but there's a layer of awareness that I genuinely don't think you can program away. Reading the room. Noticing things. Knowing the history well enough to catch what the context alone won't tell you.

That's what the operator role actually is. It's the person who holds the context that doesn't fit in a prompt.

[Ben Ornstein @benorn7]

There's a version of managing up that's just flattery. You tell your founder what they want to hear, everyone feels good, and three months later you're cleaning up a mess that was totally preventable.

Then there's the version that actually matters.

My founder set up certain processes early on that I've been running and refining for a while now. At some point I knew them better than he did, just from being in the weeds. So when a decision came up that touched one of those processes, I flagged it. Totally saw what he was doing and why. But I told him my perspective from being in the thick of it.

Not a lecture. Not pushback for the sake of it. Just context he didn't have because he wasn't the one doing it every day.

That framing has changed decisions. Sometimes stopped them. Every time, he's been glad I said something.

The difference between flattery and actually managing up is pretty simple. One of them requires you to actually know what you're talking about.

[Ben Ornstein @benorn7]

Being liked is downstream of being trusted. Not the other way around.

When someone genuinely trusts you, they'll probably like you, at least professionally. They don't have to want to hang out with you outside work. But they'll let you make a call without hovering. That's what trust actually earns you.

The ability to operate with independence. Make decisions without sign-off on every single one. The founder isn't in your business constantly because they know you're not going to do something that surprises them in a bad way.

Being liked is nice. Being well-liked but not trusted means you're fun at the team offsite and nobody gives you any real responsibility.

In a startup, trusted is the one that matters. Work backwards from there.

[Ben Ornstein @benorn7]

At a small startup, credibility doesn't come from your title. It comes from being the person who shows up before anyone had to ask.

Titles are fluid. Org charts shift. The person who builds real credibility across the company is the one who made themselves known before the moment of need.

What's worked for me: over-communication, especially in remote. When I start working alongside a new teammate, I don’t wait to be introduced. I hop on a 15 to 30 minute call, tell them what I focus on, where I can help, and that I’d always available. That's it. Not a big move. But it's a move most people don't make.

In an operator role you have visibility across the whole business. You can help directly or route to the right person. That's actually a valuable thing. Make people aware of it before they need it.

The ones who build credibility fast aren't the ones with the clearest job title. They're the ones who show up first.

[Ben Ornstein @benorn7]

I've said no to my boss. More than once.

Not in a dramatic way. Just a direct conversation. I can't take this on right now and here's why. This belongs with someone else for these specific reasons. I disagree with this direction and here's what I'm seeing from where I sit.

It has to be logical. You can have some emotion in it, and sometimes that's unavoidable. But the case has to stand on its own. Being at capacity is a real reason. A task genuinely belonging to someone else is a real reason. Disagreeing with the direction is a real reason. Vague discomfort is not.

What I've found every time is that founders don't want to wait until the wheels are already coming off. When you flag something early, the usual response isn't frustration. It's relief.

Small teams can't afford to have someone silently drowning. The conversation is almost always worth having.

[Ben Ornstein @benorn7]

Writing things down is the foundation of delegation.

Before I hand a process to anyone, I need to know it front to back. Every step. Every edge case. Every question the person I'm delegating to might ask and why things are done the way they're done. Because if I can't answer those questions, that's a failure on my part, not theirs.

SOPs don't have to be formal. They don't have to be beautiful. But they have to exist. A Notion doc. A Loom. A voice note transcribed and cleaned up. Something.

The alternative is a process that lives entirely in your head, which means it's not actually a process. It runs as long as you're the one running it. That doesn't scale, even at a company of five.

Documentation isn't busywork. It's how you create capacity.

[Ben Ornstein @benorn7]

Speed is the obvious one. Quality is the reason I actually care.

I use Claude as a thought partner. I'll drop context on a situation, get a response, push back on what it missed, and work toward an answer I'd never have reached as fast by staring at a Notion doc.

The back-and-forth is the whole thing. I'll point out what it missed. It adjusts. We work through it. By the end, the answer has been stress-tested from a few different angles.

For a remote operator especially, that matters. No one to grab and talk through a decision with. Claude handles that in a way I didn't expect when I first started using it.

[Ben Ornstein @benorn7]

Cash flow is the only number that tells you if a company is real. Everything else is noise.

I learned this working at a company with almost 20 people and almost no real money. Blockchain grants. VC funding. A full team. Looked like a real company from the outside. But cash flow was negative and everything depended on the next round coming through.

I've operated at both a venture-backed startup and a bootstrapped one since. Headcount isn't health. A full Slack and a busy calendar don't tell you if the company survives the month.

Without cash flow, it can be over overnight.

You hear this. You don't believe it until you watch it happen.

[Ben Ornstein @benorn7]

Every remote startup says communication is everything. Almost none of them mean the same thing by it.

I've watched how responsibility actually moves on remote teams. It goes to whoever over-communicates. Whoever shows up to real conversations, thinks out loud, and adds genuine value before the decision is made.

That's the actual culture. The pattern of who earns trust and what they did to get there.

Joining a remote startup? Ask who gets handed more responsibility over time. The answer tells you everything about what the company actually values.

[Ben Ornstein @benorn7]

Test every tool. Don't stop testing.

I'm building in one tool right now and already know there's probably a better one I haven't tried yet. That's not a problem to fix. That's the job.

You spend time in something, learn its limits, notice what it can't do. Then you find the next thing. Then you do it again. This is what keeps you sharp.

The willingness to look at what you're currently using and ask what the downside is, what's next, what would make it obsolete... then adjust before you have to. That's operating in 2026.

[Ben Ornstein @benorn7]

The skill that's paid off most in the last year isn't one I planned to build.

It's curiosity. Specifically... the willingness to just get in the weeds fast.

Think of a mouse in a maze. You try a path. It doesn't work. You back out, try another. You're not the most efficient worker this way. But you learn the maze faster than the person still planning their route.

I've always done this. Tried things before fully thinking them through. Wasted time and money occasionally because of it. But the upside is real. I find actual answers faster than I would if I waited until I was sure before moving.

That curiosity to just dig in... it's the operator skill of the next few years.

[Ben Ornstein @benorn7]

Most of my path was deciding to stay in the room when things got uncomfortable instead of finding a reason to leave. That's the honest version.

The polished version sounds intentional. Back office banker turned startup operator, learned all the right skills, figured out the tools, landed in the right place.

That's not how it went.

I left corporate without a plan. I had no idea what back office even meant when I was there. I left because I wanted autonomy and I was bored. No vision, no strategy.

The podcasting thing wasn't a calculated skill-build. I just liked doing it while I was still at the bank. The first startup founder that ended up hiring me did so partly because of it. I did not see that coming.

The rest got tidied up in retrospect.

[Ben Ornstein @benorn7]

Everyone talks about startups as the risky option. That's backwards.

At a big company, you're trading learning for stability. The stability feels real until it isn't. The layoff numbers in 2026 are proof. ~100k jobs at companies that were supposed to be safe.

At a startup, the risk is visible. Revenue, runway, market. You can see it. At a big company the risk hides behind org charts and performance reviews until one quarter it doesn't.

Big companies don't teach you to figure things out. They teach you to work within a system. That's a fine skill until the system stops working.

Startup experience is the only kind that actually compounds. The mess, the ambiguity, the no-playbook moments. That's what builds the thing that can't be cut.

[Ben Ornstein @benorn7]

Hot take: Non-technical people aren't actually behind the engineers on AI tools.

I have no technical background. No coding. No CS degree. I've learned how to use AI tools by just… doing it. No course, no mentor, no documentation worth reading.

Am I doing things as quickly as a technical engineer would? No chance.

But does my context as an operator give me the unique advantage to use these tools and push things forward independently? I sure think so.

What I found is that most of these tools aren't gated by technical ability. They're gated by willingness to figure it out when it's messy and nothing works yet.

Engineers have an advantage when the bottleneck is code. When the bottleneck is judgment, taste, and knowing what you're actually trying to build, that's a different problem entirely.

Most AI tools are being built for operators, not engineers. The people falling behind aren't the non-technical ones. They're the ones waiting for a playbook that isn't coming.

[Ben Ornstein @benorn7]

~230 different tech layoffs in 2026... already. 91K+ people's jobs erased.

Most of them didn't see it coming. They were in "safe" roles at stable companies.

I thought the institution was the security. The brand, the size, the structure. I left once I saw it was never actually something I could have an impact on.

What changed was realizing the security was never the job. It was whether I could figure things out when no one handed me a playbook.

That's the only thing that transfers.

[Ben Ornstein @benorn7]

The people waiting for the playbook will always be behind the people who write it themselves.

I have learned everything that matters by doing it before I knew how. That is not a hustle take. It is just how ambiguous environments work.

[Ben Ornstein @benorn7]

Emotional regulation is a career skill. Nobody puts it in the job requirements.

I have watched technically sharp people get sidelined because they could not stay even keel. And seen people with fewer credentials become indispensable because they could.

It compounds. Treat it like a skill.

[Ben Ornstein @benorn7]

Everyone wants to be the founder. Nobody talks about the person who makes the founder work.

The operator role is not support. It is infrastructure. Most companies do not figure that out until it is missing.

[Ben Ornstein @benorn7]

Most meetings are just people avoiding the discomfort of writing things down.

Async is not a style preference. It is a discipline. Writing forces clarity in a way a 30-minute call never does.

[Ben Ornstein @benorn7]

The skill everyone told me was just a hobby ended up being why I got hired.

When I interviewed for my first startup job, the hiring manager asked about my podcast. Not my Morgan Stanley background. My podcast.

What you build independently signals capability in ways your resume cannot.