The Screening Question I Added Before Starting Anything New

A few weeks ago I wrote down a rule I was testing: before starting anything new, write one sentence saying where a stranger who wants it would actually be looking. If I couldn’t answer that honestly, I didn’t get to start building. It was a reaction to noticing that I keep reaching for a new build instead of fixing the distribution problem on something I already have. Writing the rule down was the easy part. Actually running my own ideas through it turned out to need more than one sentence.

Why the sentence alone kept slipping

The first time I tried to apply it, I wrote a sentence that sounded right without actually answering the question. Something like “people looking for this kind of tool would find it through search.” That’s technically an answer. It’s also the same vague hope that didn’t hold up for the n8n templates or the boilerplates, both of which also had a checkout page a stranger could theoretically search their way into. A one-sentence rule is easy to satisfy with a one-sentence excuse. I needed something that forced a more specific answer than “search” or “word of mouth,” which are the two answers that let almost any idea pass.

What I turned it into

Three questions now, asked in order, before I let myself open a new project folder:

Is there an existing place people already go looking for this kind of thing, by name? Not “the internet” or “search,” but an actual directory, marketplace, or community, the way the Obsidian plugin directory or Etsy’s own search exists for what I already sell there.

If the answer is no, am I willing to name, specifically, what I’d do differently this time to build that channel myself, before I start on the product? Not “market it better,” an actual first step I could point to in a week.

If I can’t answer either question with something concrete, does the idea survive being shelved for two weeks while I work on distribution for something I’ve already shipped instead?

That third question is the one doing the real work. Most ideas don’t survive it, and that’s the point. If an idea can wait two weeks without me losing interest in it, it’s probably worth doing properly later. If I lose interest the moment I can’t start immediately, that tells me more about my own pull toward novelty than about the idea itself.

Running it against what I’ve already built

Honestly checking my last few products against this, the plugins pass cleanly: the Obsidian community was the channel before I wrote a line of code. The Etsy shop passes too, for the same reason the listing SEO work was worth doing. The boilerplates and the n8n templates both fail question one outright, which I already knew from writing about taking them down. What’s new is that they’d have failed question two as well. I never had a specific first step for building a channel for either, just a general sense that I should “market them more.”

Where I can still talk myself past it

I’m not going to pretend this closes the loop. A checklist run by the same person who keeps failing the underlying habit is still vulnerable to being answered generously on a day I really want to start building. I can write a plausible-sounding first step for question two and never actually do it, which would make the checklist worse than useless, since it would let me feel like I’d done the diligence I skipped. The honest test isn’t whether I can answer the three questions. It’s whether I come back in two weeks and check whether the specific step I named for question two actually happened.

The other end of this

I already have a checklist for when to kill a product, built after the fact from two retirements. This is the checklist that should have run before either of those products existed, and didn’t. Having both doesn’t guarantee I use them well. It at least means the next product I ship, or don’t, will have gone through a specific question instead of a feeling I trusted because the last idea felt exciting too.