The Three Decisions You Don't Delegate

Most founders I talk to think the risk in hiring an outside partner is handing over too much control. That's backwards. The risk is keeping it.

I'll explain what I mean, but first: a number. Ninety days. That's roughly how long it takes a new agency to learn your company well enough to do useful work. Pricing positioning, who actually buys, where the product breaks, why your last hire didn't work out, what the board cares about this quarter. None of it transfers in a kickoff deck. You pay for the relearning every time you switch agencies, and every time you bring in a new vendor for a new function. Call it the re-explanation tax.

An embedded partner never pays it. Not because we're smarter, but because we're inside. The information stays with us, compounds, and gets used the next quarter without a kickoff.

Here's what that looks like in practice.

A seed-stage SaaS founder booked a call last year asking for a logo. Standard ask, or so it seemed. Six weeks in, we were managing his 12-person development team in India. The work expanded that fast because the underlying problem wasn't the logo. It was that nobody internal had time to coordinate a product roadmap across a twelve-hour time difference, write the positioning that would let sales close, and keep the brand consistent through three simultaneous launches. The logo was the symptom he could name. The operating layer was what he actually needed.

This is the part founders miss when they're sorting between an agency, a fractional CMO, and a first marketing hire. The work in a seed-stage company doesn't sit in clean lanes. Positioning bleeds into product. Hiring bleeds into brand. Vendor management bleeds into runway. Hand each piece to a different specialist and you become the integration layer. You take the 2 a.m. India call yourself. You translate between the brand agency and the dev shop. You become the bottleneck you hired three vendors to avoid.

Everyone treats decision rights as the dangerous thing to give up. I think withholding them is what makes a vendor useless. A vendor without decision rights can only execute what you've already decided, which means you have to make every decision, which means the company moves at the speed of one human's attention. Founders who try to run it this way ship slower, not faster. I watched a CEO insist on personally approving every piece of copy, every engineering email, every contact point sent to investors and affiliates. Her product launch slipped three quarters. Not because she was a bad CEO, not because her team was incompetent; critical movement slipped through the cracks because the team was relegated to waiting on her to catch up to her overwhelming work load.

The fix isn't to hand the whole company over. It's to be deliberate about what you actually hold.

Before you hire anyone, write down the three decisions you will not delegate. Not five, not ten. Three. For most founders I've worked with they look something like: who joins the executive team, how much equity goes out, and which customers we say no to. Yours might be different. The point is that the list is written down, it's finite, and everything not on it is fair game for an embedded partner. Pricing? Delegable. Hiring below executive? Delegable. Product positioning? Delegable. Vendor selection? Delegable. The launch date? Probably delegable, though most founders flinch at that one.

This isn't a permission slip to disappear. It's the opposite. It frees you to be sharp on the three decisions that actually compound, and stop being the rate-limiter on the other thirty.

I don't pitch a deliverable when founders book a call. By the time someone shows up they've already decided they want an operator. My job is sorting out whether we're the right one. The conversation that follows is short. We talk about what they hold, what they're willing to hand off, and where the seams are between them. If the three-decision list is already clear in their head, the work starts fast. If it isn't, that's the first thing we build together.

Write the list before you take the next call.