There's a version of marketing that runs entirely on vibes. The campaign felt good. The content looked great. The team was proud of it. And when someone asks how it performed, the answer is somewhere between "really well we think" and a pivot to a metric that sounds relevant but doesn't actually answer the question.

I've been in that position. It's uncomfortable in a specific way: not because you did bad work, but because you can't tell whether you did good work either. And you can't tell because nobody agreed on what good looked like before you started.

That's the actual problem with not having KPIs. It's not that you can't report on performance. It's that you have no way to learn anything.

What KPIs Actually Do

Every campaign becomes a standalone event that either felt successful or didn't, with no real thread connecting what you tried, what happened, and what you'd do differently next time. You're flying without instruments and calling it creative freedom.

KPIs aren't about proving your worth to a stakeholder, though they do that too. They're about having a conversation with yourself about what you're actually trying to achieve before you start making things. That conversation changes the work. When you know you're measuring follower growth you make different content than when you're measuring saves and shares. When you know you're measuring click-through to a landing page you think differently about your CTA than when you're measuring brand sentiment. The metric focuses the thinking in a way that nothing else does.

KPIs Protect You

The other thing KPIs do is protect you. When a campaign underperforms, you want to be in a position to understand why: was it the creative, the targeting, the timing, the offer, so you can fix the right thing. Without a clear metric you agreed on upfront, every post-mortem becomes a guessing session where everyone has a different theory and nobody can be proven right or wrong. You end up making changes based on whoever is most confident in the room, which is not a great way to improve.

The practical version of this is simple. Before any campaign, agree on one number that will tell you whether it worked. Not five numbers. One.

The one that most directly connects to what you were trying to do. Everything else is context. That one number is the verdict.

It can be around the revenue metric or not. Depending on the objective it might be reach, or saves, or time on page, or DMs, or conversions. What matters is that it was chosen intentionally before the work started, not reverse-engineered from whatever happened to go up after it launched.

Marketers who track their own work closely get better faster than the ones who don't. Not because the numbers tell you everything, but because they give you something to argue with. And arguing with data, even imperfect data, is how you sharpen your instincts over time.