For a couple of years I blamed the beans. The coffee at home was fine, drinkable, but never as good as the café down the road, and I worked through a small fortune of fancy single-origin bags trying to fix it. Different roasters, lighter roasts, darker roasts, beans flown in from places I'd have to look up. Marginal. Always marginal.
It was the grinder. It was always the grinder. I'd been using a cheap blade thing that doesn't grind so much as assault the beans, producing a mix of dust and boulders, so half the coffee over-extracts into bitterness while the other half barely extracts at all. No bean survives that. You're not tasting Ethiopia, you're tasting an inconsistent particle distribution.
So I bought a proper burr grinder, and here's where the rabbit hole opens: once you can hold grind size roughly constant, every other variable suddenly matters and becomes worth fiddling with. Dose by weight. Water temperature. The grind setting itself, two clicks finer and the whole cup changes. I now own a set of scales that measure to a tenth of a gram, for coffee, and I have made peace with the person that makes me.
The annoying lesson, which is of course really a lesson about everything else, is that I'd spent ages optimising the variable I could buy (the beans) while ignoring the one that actually gated the result (consistency). The expensive input can't rescue a sloppy process. I'd have saved money and better mornings by spending on the grinder first and the beans last. There's a tidy little parable about premature optimisation in there, but it's early and the coffee, finally, is good.