Potential backlash to features is another reason to keep software simple. Skipping work on stuff people do not want is good business sense, but avoiding criticism through simpler software might end up being a strategy for more nefarious goals.
There are many principles in the fight against feature creep on the production side of software, but with enough demand those concerns can be overridden as the value provided to users sometimes justifies the cost. Something I did not consider as another downside to development until recently is that even with the best intentions to improve the lives of users and good technical execution, the functionality or implications of a new feature can repulse users. While some users may be pleasantly surprised by a new way software utilizes personal information, privacy-conscious users may have a harsh reaction that can spread. For some people this may make them more self-aware about the impact of feeding their data to websites.
So clearly there is a new factor to consider around feature creep: will the features creep out the users? This is particularly interesting with big companies that already have the data serving simpler and presumably agreed upon purposes. Putting aside uses of that personal data that generally do not serve to the direct benefit of the user, I think these companies also genuinely develop new features and products that are intended to at least be perceived as improving our lives while taking advantage of the low-friction access for the existing userbase. It is somewhat perverse that the undesirable acts that companies perform with data can go undetected and unscrutinized, but convenient affordances that seem a bit too convenient to be comfortable lead us to realize the true cost of the bargains we are making in the information economy.
Perhaps when software is dumb with our information, that information seems less valuable to us and not a big deal if others have it. The increasing sophistication of software makes us more cognizant of what potential purposes our data could be providing to someone else. When software is kept simple, a person is probably less likely to critically re-evaluate the initial “deal” of being the user. Incremental changes can more easily pass this analysis. So is a slow boil enough or is there an unavoidable point where the feature backlash will be too fierce? Will we demand simpler software that we can trust to respect our privacy or by then will the conveniences be too difficult to give up?
Direct and transparent communication with users during planning is important to develop software that is delightful to use while not overstepping personal boundaries. This article shows how to use user stories to center the user experience in new features: https://www.toptal.com/designers/ux/feature-creep. Beware though, if you add every user request you may run into feature creep in the traditional sense of unmaintainable bloat on an unfocused project. Discretion is critical!