This transcript was produced by robots and left as-is. Accuracy and elegance are not guaranteed.

Last episode, made you a promise. If logic isn’t sufficient for alignment, the obvious next question is, then what is? This episode answers that. Not philosophy, mechanics.

Influence is a system. It has inputs. And most engineers have been trying to optimize the output or alignment without understanding what actually generates it. That’s like trying to fix a process output without looking at the inputs. You can tune the output parameters all you want, but if the inputs are wrong, you’re just wasting your time.

You can’t brute force alignment with logic because logic only addresses one of the three inputs. The other two stay broken and you’re left wondering why nothing sticks. So let’s look at all three.

Input one is the signal before noise. The default engineering move is to present the conclusion then justify it if someone pushes back. That’s basically the scientific method. Hypothesis first, then evidence. It feels rigorous because it is rigorous, but influence works in reverse. Context has to come before the idea, because if you lead with the conclusion, the other person’s brain is already reacting before you get to the reasoning. They’re processing your answer while you’re explaining the logic behind it. The context never lands because they’re not listening anymore, they’re responding.

Now, this isn’t about softening the message. It’s not about being diplomatic or easing into things. It’s about sequencing. Without the right frame first, even a sound argument arrives as a position to resist.

Here’s a bit of a composite story I see play out constantly. Simplified, but the pattern is identical every time. Picture a controls engineer leading a project kickoff. The plant shutdown is scheduled. There’s a tight window. The execution plan matters. She opens the meeting like this. We’re going to remap all of the IO as soon as the plant goes down, then download to the controller. This minimizes impact on the plant. Technically sound, logically justified. She’s already done the analysis. She’s right.

The room nods. Nobody argues. The meeting ends in 12 minutes. Three days into the shutdown, the instrument tech flagged something. One of the remapped tags conflicts with an existing tag still live in another part of the plant. He knew about it. He just assumed it had been accounted for. The tech assumed it wasn’t his problem to raise because the plan was already set. That’s not a technical failure. That’s a sequencing failure.

Now, run the same meeting with the order of operations inverted. Before we lock the execution plan, I wanna walk through what we’re dealing with. Tight shutdown window means the sequence matters. Here’s what I’m thinking and why, but I want to pressure test it before we commit. What am I missing? Same engineer, same plan, same room. The instrument tech mentions the tag conflict before the plan is locked. It gets folded in and the shutdown runs cleaner.

The difference isn’t the quality of the analysis. It’s the order it was delivered in and whether the room had a cheap way to push back before the conclusion landed, which brings us to input two, the cost of dissent.

You’ve probably heard this framed as psychological safety. I find that framing mostly useless because it’s not actionable. Here’s the mechanical version. How expensive have you made it for someone to push back? Not whether they feel comfortable, but whether disagreement actually costs them something, reputation, standing, your visible reaction when someone challenges you. When the cost is high, people nod. They don’t argue, nothing moves. You have compliance, not commitment.

The goal is not to eliminate disagreement. It’s to make disagreement cheap enough that people actually do it. One concrete way to do this is to say it out loud before anyone has to risk it themselves. I want to hear what’s wrong with this before we decide. That one sentence changes the cost structure of the room. You’ve made it not only safe, but expected to get pushback. You’ve given people cover.

When the cost of dissent is low enough, people engage. They buy in. When they buy in, things actually move, which is the third input, clear movement. Influence without direction isn’t influence. It’s a conversation. Every influential exchange ends with someone knowing what happens next. A decision, a test, a check-in date. Not necessarily closure. Movement. If people leave the room unclear on what changed, it means nothing changed.

Go back to the first meeting example. The first version ends with nods and silence. No one knows what the next step is because the engineer didn’t name one. The plan exists in her head and it stays there. The second version ends with a specific next step. Let’s flag any tag conflicts by the end of day Thursday, and I’ll incorporate them into the final plan before we go into the shutdown. Now there’s a direction, something to act on. The meeting did its work. Same information, different output. The difference is system design, the order of operations, the cost of dissent, and a clear next step.

So before your next meeting where you need to gain alignment, write down three things. What context does this person or group need before I present the idea? What would make it easy for them to disagree with me? What movement do I need out of this conversation? That’s your prep. Five minutes or less. It changes what happens in the room.

Now the hardest place to apply all three of these is a conversation where someone isn’t performing. Those conversations get avoided for months, sometimes years. That’s where we’re going next.