Skip to content
Amila

Palahepitiya Gamage Amila · Weekly Learnings

What I approved this week that never happened

· Palahepitiya Gamage Amila

A stack of cards marked with checks beside a sealed envelope labelled Waiting to leave. The line above reads Approved ≠ sent.

On Friday I went back through everything I'd said yes to during the week. I expected to find it done. A lot of it wasn't.

Some context. I run a few companies, and AI assistants do much of the first-pass work for me. They draft emails to people we'd like to work with, research companies, and put documents together. Most of my part is reading what they made and deciding whether it's good enough to go.

What I found on Friday was a pile of things I'd approved that were sitting exactly where I left them. Outreach emails I'd okayed that nobody had sent. A research task I'd scheduled overnight that had stopped partway through because it needed me to click a permission button, and I wasn't there. A set of documents that was supposedly ready to forward and hadn't been forwarded.

In each case I'd felt the job was finished the moment I said yes. Mostly it hadn't started.

The part that interested me most was our outreach. We weren't getting replies, and my instinct was that the system needed fixing: better leads, a different tool, maybe paying for more data. When I looked at what was actually going on, it turned out to be two problems at once, and they needed opposite fixes.

One was that the places we look for leads had gone quiet. The tool was working; it just wasn't finding much worth contacting. More approvals wouldn't have helped with that.

The other was that plenty of emails had been written and were waiting on a yes from a person. They hadn't gone out, so obviously nobody was replying to them. More leads wouldn't have helped with that.

From the outside both look like "we sent outreach and heard nothing." If I'd treated them as one problem, I'd probably have paid for more leads while the real hold-up was a queue of unsent emails.

Why does approving feel like finishing? I think partly because the decision is the hard, thinking part, and once it's made your brain files the task as done. And partly because the numbers I watch reward it. I could see how many emails had been drafted and how many research reports had come back. Those looked fine. Nothing told me how many had actually reached anyone.

So the change I'm making is small. When I look back at the week, I don't ask what I approved. I ask what actually went out. It's a less flattering list, but it's the one that matches what happened.

Questions this note answers

Why does approving feel like finishing?

Partly because the decision is the hard, thinking part, and once it's made your brain files the task as done. And partly because the numbers I watch reward it. I could see how many emails had been drafted and how many research reports had come back. Those looked fine. Nothing told me how many had actually reached anyone.

What were the two outreach problems?

One was that the places we look for leads had gone quiet. The tool was working; it just wasn't finding much worth contacting. More approvals wouldn't have helped with that. The other was that plenty of emails had been written and were waiting on a yes from a person. They hadn't gone out, so obviously nobody was replying to them. More leads wouldn't have helped with that.

What do I ask when I look back at the week?

I don't ask what I approved. I ask what actually went out. It's a less flattering list, but it's the one that matches what happened.