Every MarTech team has a spreadsheet that holds everything together. Nobody documents or talks about it. And when it breaks, no amount of AI can save you.

There's usually one question that, if answered honestly, answers the rest of the questions. Finding it one thing, answering is another.

When the only good outcome is an uplift, the test backlog slowly fills up with safe, small bets that don't matter for the bottom line.

There's a better question: did the experiment give you enough information to make a decision? A test that confirms something doesn't work is just as useful as one that finds a winner.

Most teams measure experimentation by win rate. How many tests produced an uplift. If the number's low, there's a perception that experimenting is not the right thing to do, and people quietly stop proposing bold ideas.

You can't micromanage people and expect initiative. Define what good looks like. Share the constraints. Then step back. The first outputs won't be perfect. That's the investment of building capability in people. Most managers aren't willing to invest it.

The quality of what AI gives you is a direct reflection of what you demand from it.

When you stop expecting people to be a version of you, you start seeing what they actually bring. And usually, it's more than you thought.

Most confusion in business isn't caused by hard problems. It's caused by inherited ones that nobody thought to re-examine.

Funny thing is, when the real number finally calculated, it rarely changes the logic. In my experience, not having the data is rarely the bottleneck but lack of strategic thinking is.

Put in a placeholder. Call it X. "If LTV is above X, we do this. Below, we do that." Now you have a plan for either outcome.

When there's no immediate answer to a data question in meetings like "what's the LTV of our users?", a missing data point becomes an excuse to pause all the thinking around it. Like the absence of one number made every other thought not relevant.

But you have to be willing to be a beginner again. And that's hard when your whole identity is built on having the answers.

So they find reasons: "It hallucinates." "It doesn't understand our business." All reasonable but equally missing the point. What made you a good manager was never the knowledge itself. It was knowing what to do with it. AI doesn't replace that part. It actually makes it more valuable.

Middle managers built their careers on knowing things other people didn't. AI closes that gap fast, and that's terrifying when your authority depends on being the smartest person in the room.

The tech is not bad, obviously. It’s not my point here. I’m a heavy AI user and a productivity geek in my daily life. But I genuinely believe that we're outsourcing the part of the work that actually creates real understanding of what matters.

We keep optimising for speed, nowadays even more with AI. We have transcription tools, AI summaries, and automated reports generated in a few seconds. However, each of these automations removes a critical step that used to require us to think critically.

People who take notes by hand tend to remember more than those who copy and paste. The friction is the point.

You can't capture everything, so you have to decide what matters. This friction makes you really understand that just copying doesn't.

Every roadmap conversation has a version of this: if we had more money, more time, more headcount, more <insert your fav excuse here>, we could do XYZ.

“Wishlist” thinking is accountability avoidance. If the blocker is always something you don't control, you never own the result.

Every year, MarTech leaders face the same question: how much of the stack are we *actually* using?

A tool’s value doesn’t grow because you use more of it, it grows when the part you use changes an outcome you care about.

The tool isn’t the strategy. It’s meant to serve one.

Ask yourself: when did I last change my mind because of data? If you can't remember, you're not data-driven either.

Real data-driven culture means changing your mind when the data contradicts what you want. Most teams can’t afford that.

So we keep the language and optimise the data selection instead.

Teams decide first, then look for the data that supports the decision. The deciding happens on intuition, internal politics, or constraints no one can say out loud. The data comes after, because "my gut says so" doesn’t get budget.

The phrase ‘data-driven’ is one of the most misused terms in business, as it describes the wish, not the reality.

Shipping isn’t the finish line. Learning is.

The most valuable part of learning isn’t the conclusion, it’s the struggle that rewires your brain to get there.

The goal isn’t to be right the first time.
The goal is to be wrong quickly, learn rapidly, and adjust immediately.

Most teams focus on shipping more.
Smart teams focus on learning more.
The best teams use constraints to force faster learning cycles.

Getting ahead isn’t about doing more. It’s about caring less about the things that won’t move you forward. Strategic neglect is a superpower.