I don't even see it as increasing KPIs or sales. Unless it's sales as in selling the idea to the leadership.
A lot of it is trying to make it look visually and superficially impressive trying to apply the common hammers like whitespace, very little visible content at once as to not "overwhelm" the user and other rules to anything, and then trying to pretend whatever that is going to increase the sales by doing non-sensical user research, only listening to what proves their initial vision and then later trying to cherry-pick success metrics.
A/B test the new Settings panel:
If User spends more time in the Settings page -> We just increased engagement. They must really like the new Settings page and find that pleasurable to use. That's a win.
User spends less time in the Settings page -> We made it quicker to find and tweak the correct settings. That's amazing.
Either way, any change you make you can find wins. You could even keep going back and forth between two versions.
Also another common flaw I notice with some redesigns or things in general is that when the features, e.g. config settings were initially released, there was very thorough understanding and need for that feature and whoever put it there knew the technical implications and why it is needed, and they put it there in such a way that it's reasonably easy to use and makes sense with other things.
However when redesigning, you are going to be exposed to all those features and nuances at once. And you won't have a good understanding of everything at all and all the little details and implications. And you might want to start with some sort of feel, framework and design to handle all of that. But then you are going to approach those features from an aspect where you have to take the features and make them adapt to the design as opposed to considering what makes sense specifically for that feature. And you won't have good understanding of the features, so either you are reducing the scope a lot or also just implement features incorrectly or in a poor way, not working well together with some other features, etc.
I was specifically talking about where I work with that criticism about KPI and sales. I have no idea what might be driving Microsoft's design philosophy.
Yeah, I wanted to specify that in my experience everyone does want increased sales and have results with KPIs, but many times these things actually won't do it. Maybe they will improve some KPIs, but still won't have results where they would bring in more sales/profits. E.g. KPIs are cherrypicked, gamed in some way, they are flawed, increasing the KPIs comes at cost of something else, so sales are actually unaffected or they even have a negative effect.
Or at least leadership definitely and obviously wants it, but product and design under leadership wants to just convince leadership that it's going to do that. But frequently it isn't or it's very difficult to analyze that data and it's easy to cherry pick flawed metrics retroactively to try to convince leadership that it actually did some improvements. It's difficult to question and verify any sort of metrics without launching a whole investigation into it yourself, which rarely someone would have an incentive to do so.
It could be that at your workplace maybe these things actually do improve KPIs and sales, I would actually be more happy with that compared to if there are meaningless changes or changes that make things worse that don't improve sales at all.
My problem with KPIs and sales is that I've seen the people I work with actively make decisions that make the experience worse for our users because it increases the amount and/or value of sales we make, verified through A/B testing. Meanwhile ignoring or deprioritising legitimate issues because it can be very difficult to "prove" that it would increase KPIs.
A lot of it is trying to make it look visually and superficially impressive trying to apply the common hammers like whitespace, very little visible content at once as to not "overwhelm" the user and other rules to anything, and then trying to pretend whatever that is going to increase the sales by doing non-sensical user research, only listening to what proves their initial vision and then later trying to cherry-pick success metrics.
A/B test the new Settings panel:
If User spends more time in the Settings page -> We just increased engagement. They must really like the new Settings page and find that pleasurable to use. That's a win.
User spends less time in the Settings page -> We made it quicker to find and tweak the correct settings. That's amazing.
Either way, any change you make you can find wins. You could even keep going back and forth between two versions.
Also another common flaw I notice with some redesigns or things in general is that when the features, e.g. config settings were initially released, there was very thorough understanding and need for that feature and whoever put it there knew the technical implications and why it is needed, and they put it there in such a way that it's reasonably easy to use and makes sense with other things.
However when redesigning, you are going to be exposed to all those features and nuances at once. And you won't have a good understanding of everything at all and all the little details and implications. And you might want to start with some sort of feel, framework and design to handle all of that. But then you are going to approach those features from an aspect where you have to take the features and make them adapt to the design as opposed to considering what makes sense specifically for that feature. And you won't have good understanding of the features, so either you are reducing the scope a lot or also just implement features incorrectly or in a poor way, not working well together with some other features, etc.