Up the assesment
Don't make implementation the sole focus of system development.
In every leadership role I've held — marketing director, chief marketing officer, head of strategy — I made the same drastic mistake. I went for the elusive, beautiful, easy and easy-to-sell free scope. I paid people to resolve patterns I caught. I'd bring in an expert to fix a problem, and then hand them my own half-formed understanding of it as the brief. I'd acknowledged I needed outside help, but kept all the diagnosis for myself. The surgeon got invited in and told to start cutting and work it out as we went. If I understood the problem well enough to scope it, why didn't I just fix it myself? And if I didn't understand it that well, why was I the one writing the brief?
The thing I'd missed is that having someone who's solved this kind of problem several times before is valuable for the diagnosis, not just the cutting. That's the part I was skipping.
It's worth saying what's changed, because it makes the mistake more expensive than it used to be. In 2026 good code, implementation and media still aren't cheap — but the timeline to value is a fraction of what it historically was. You can move from decision to working thing very fast now. Which is great, right up until you point all that speed at the wrong target. When shooting is quick, the only thing that still matters is aim. And aim is the part I kept in my domain.
Now I'm very fortunate that it worked out (mostly). But certainly that working out costed a lot more than it needed to, created a lot of grey hairs, missed a lot of family walks and a mammoth amount of luck.
It's an easy part to leave out, because any single problem looks finished on its own. The report's wrong, fix the report. The handoff keeps dropping things, add a step. Each one has an obvious answer and the answer works, for that one thing. The trouble is the problem in front of you is usually not a one-off. It's one place where some larger shape in the business happens to poke through, and that same shape is causing other problems you haven't joined up to it yet. Fix the bit that poked through and the shape's still there. It comes back up somewhere else and looks like a completely different issue.
The self assessment isn't a warm-up before the real work. It's a chunk of real work, and not an easy one. You have to get inside how the place actually runs, which is never how it gets described in a meeting. You have to follow a problem past the person who reported it to whoever's downstream. The users who will (maybe) get an SOP document that changes 30% of their life now. You have to not take the bait of the first thing you can obviously fix. None of that happens in the gap between spotting a problem and pitching a solution. There's no room in there for it. It certainly didn't happen when I was the one holding the brief.