There is a well-worn piece of product advice that says a recommendation engine should be decisive. Give people one answer. Choice is paralysis, options are friction, just tell them what to do.
It is good advice in most places and it is wrong for getting dressed. We learned that the slow way, watching what people actually did with a single daily outfit: they tapped "show me another." Then again. Then again. Eight taps deep, having seen eight outfits one at a time, with nothing to compare and no memory of what the third one looked like.
A Single Recommendation Turns Choosing Into Guessing
Picking an outfit is not a yes-or-no judgment. It is a comparison. You are weighing formality against comfort against mood against how much you feel like being looked at today, and none of those have absolute thresholds. They only resolve against an alternative.
Handed one outfit, you cannot do that. You can only ask "is this acceptable?" — a much worse question, and one that tends to get answered "not quite" for reasons you can't articulate. So you ask for another, and now you are comparing a picture on screen against a memory of a picture, which is a comparison nobody wins.
Sequential single options don't reduce choice. They make the same choice harder, one screen at a time.
Why "Show Me Another" Keeps Showing You the Same Thing
There is a second failure hiding inside the one-at-a-time pattern, and it is the one that actually drives people away. Ask for an alternative and you frequently get a near-copy: the same trousers, a marginally different white shirt.
This is not laziness, it is arithmetic. If a system ranks your wardrobe against a request and hands you the best match, then the next-best match is by definition the thing most similar to the one you just rejected. The more confident the ranking, the more the alternatives converge. You rejected the top result and got handed its sibling.
Fixing that takes an explicit rule, not a better ranking. The alternatives have to be built knowing about each other, with a constraint that forces them apart.
Three, Planned Against Each Other
So Clad9 answers a request with three complete looks at once, and the constraint that makes them worth looking at is simple: no two may share the same top, bottom, or dress. Not three variations on a theme. Three different ideas, side by side, where the comparison you were going to make anyway can actually happen — in one glance, against pictures, rather than against your memory.
There is a pleasant efficiency to this, too. The expensive part of generating an outfit is the thinking, not the outfit, and one piece of thinking that plans three looks together costs far less than three separate rounds of it. Comparing is both the better experience and the cheaper one, which is a rarer alignment than it sounds.
Then Change the One Thing That's Wrong
Comparison solves most of it. It doesn't solve the very common case where a look is right except for one piece — the outfit is good, the boots are wrong for the weather.
Regenerating the whole outfit to fix that is the wrong move, and it is what most tools make you do. You lose the thing you liked in order to change the thing you didn't. Swapping a single piece keeps the rest of the look exactly as it was and re-picks only that slot, against everything else you are already wearing — so the replacement has to work with the outfit you have, not with an abstract idea of a request.
Asking Again Shouldn't Undo Your Answer
One more piece of this is worth spelling out, because it is the bit that is easy to get subtly wrong.
If you ask for work outfits, see three, and then rephrase — "office day, I'm presenting" — you are asking the same question a second time. You should not get back the shirt you just passed on. But if you ask for something genuinely different later that day — date night, after a morning of work looks — you should get your whole wardrobe back. The clothes you were merely shown in the morning are not spoken for. You didn't wear them.
The distinction is between what was offered and what was worn. Only actual wear should reserve a garment across the rest of the week. Everything else frees up the moment you move on to a different question — otherwise three requests in a day would lock you out of most of your own closet over clothes that never left the hanger.
The Underlying Principle
Decisiveness is a real virtue in software. But the thing to be decisive about is the shortlist, not the answer. Do the hard work of narrowing a hundred garments to three coherent, genuinely distinct outfits — then let the person do the one part they are better at than any system, which is knowing how they want to feel today.
