[philmcc]

Not sure if you care about validation. If not skip to 2.

1. Anybody want it?

First I pitch the rough idea to a few paying customers. If they are lukewarm on it I set it aside.

If they are warm on it I set it aside.

If they say they want it I set it aside.

When some time goes by and they check in and say “whatever happened with (feature) I consider it validated.”

2. Be bad at it

Give myself permission to make the absolute worst version of it.

That frequently involves writing a one page max description of it, and then asking myself “okay how much of this could I cut before it was literally doing nothing.” But if I’m feeling bored I just start coding.

This version has no error checking. No tests. No capacity for large requests. No logging. Nothing. It just prays beyond all reason that the user doesn’t do anything silly that breaks it.

3.beta test the MEV

I just made it up. It’s not a real acronym. most execrable version. It fits because it does feel like something I excreted. I hate it. I’m embarrassed. My impostor syndrome is no longer a syndrome it is just accurate identification and classification.

I give it to those users who were interested.

I watch them on hot jar.

4. Tea leaves

This is the hard part. If they are enthused great. But if not You kinda have to sense whether or not their dissatisfaction is because it’s an MEP.

If you’ve done it right, however, and found something they super want they will actually struggle through it to get that thing done. Sometimes the MEP isn’t far off from an MVP. Sometimes their satisfaction is enough to convince you not to do the rest of the stuff you thought was necessary in your on paper Valhalla take.

5. Users of Software

I often forget that your average software user does not have our relationship with software. They don’t see the amazing version in your head so the crap version feels like magic.

You’d be surprised how hard people will work with a subpar system with the hope that it’ll work out for them, both from optimism but also from a lack of viable alternatives.

Or maybe you’re not that surprised because you’ve voted in America zing I went political at the end.

[AlchemistCamp]

Do you ever code something for yourself, or are there always customers involved?

[philmcc]

Oh sure, all the time. But that's always a bit wild-west, since I'm more or less doing it for fun, then my approach is always "Well what do I want to do, emotionally" so it's hard to advise on fun projects. It'd be like "What's fun for you to do?"

I guess if I had any kind of advice for that kind of thing it might be this: I've realised that even in my fun projects, there's a 10% crappy annoying tedious thing that -very- frequently stops me from finishing the fun projects.

A potentially bit good of advice would be to do that thing first, use all the "new fresh code base" enthusiasm to BURN through the code you're gonna hate so that it's just fun from then on.

[Rastonbury]

If you're referring to a product feature the dev wants versus a feature the customer wants, I don't think there will ever be a case when the former takes priority