Killing your favorite feature is the most senior thing you'll do

Every builder has one. The feature you fought for, designed twice, demoed with a little flourish, and secretly consider the best thing in the product. And here’s the uncomfortable pattern: that exact feature is the one you are least equipped to evaluate. Not because you’re careless — because you built it. The more effort and imagination you pour into a thing, the more you value it, regardless of what it’s worth to anyone else. So the feature that gets your fiercest defense is frequently the one running on the thinnest actual usage, and your certainty about it is not evidence of its value. It’s evidence of your investment.

This is why cutting your own favorite feature is a senior move rather than a sad one. Junior instinct is to measure seniority by what you can build — the ambitious feature, the clever system, the thing you can point to. But a 0→1 product isn’t limited by how much you can add; it’s limited by how much attention, surface area, and narrative clarity you can protect. Every feature is rent: it has to be maintained, explained, kept working as everything around it changes, and — most expensively — it takes up a slot in the story a new user builds about what your product is for. A product that does one thing unmistakably well beats a product that does that same thing plus three half-loved extras, because the extras don’t just fail to help — they blur the answer to “what is this?” The darling you can’t cut is often the exact thing smudging that answer.

The practical skill is separating two questions that feel like one: do I love this? and does this earn its complexity? They get fused because the love is loud and the accounting is quiet, so the work is to make the accounting louder. Go find the darling’s real usage — not “some people like it,” but how many reach it, how often, and whether the product’s core promise survives without it. Look at what it costs on the other side: the maintenance, the extra branch in every future decision, the paragraph it forces into your onboarding, the way it makes the product harder to describe in one line. When a feature sits in that top-left quadrant — high on your attachment, low on everything else — the senior act is to cut it while telling the truth about why: not “nobody used it” (a comfortable lie that blames the users) but “I loved it and it wasn’t earning its place.”

Two honest cautions, because “kill your darlings” curdles fast into a slogan. First, usage counts are not the whole verdict. A feature can be lightly used and still load-bearing — the thing a small but critical segment stays for, the capability that signals trust to a buyer, the quiet workhorse in the bottom-right that nobody praises and everybody relies on. Cut by raw numbers alone and you’ll eventually amputate a limb because it wasn’t waving. The judgment is precisely in telling a darling (high attachment, genuinely low value) from a load-bearing wall (low applause, high value) — and that judgment is the senior skill, not the cutting itself. Second, taken too far, “cut everything you love” becomes an excuse for timid, characterless products — because sometimes the darling is the differentiator, the strange opinionated thing that makes people choose you. The goal isn’t to become someone who kills features for sport. It’s to become someone whose love for a feature and whose decision about it are two separate faculties — able to adore a thing and still cut it, or to defend a thing on the evidence rather than the affection. That separation, more than any feature you’ll ever ship, is what senior looks like.