4 ms·
We start with complexity and simplify. We never get to simple if we don't try to wrangle chaos and complexity, so I disagree with the “It's been tried before an
by moooshupork 5y ago
We start with complexity and simplify. We never get to simple if we don't try to wrangle chaos and complexity, so I disagree with the “It's been tried before and failed” defeatism. Pareto is one of my pet peeves because, no matter how you slice it, it can be used to argue against all advanced scenarios.
1. Just because it's failed in the past doesn't mean it's going to fail in the future.
2. Context is utmost.
3. We need to stop assuming that people are stupid and start believing the complexity is okay sometimes. I believe a tyranny of the majority is no more helpful in software than it is in democracy, so I don't buy into Pereto. That one belongs on the trash sheep along with UML, Agile, and other systems and theories of software development. IMO, of course. But there's a time and place for complexity. Audience matters. I think I'm preaching to the choir here.
- axegon_ 5y agoI think you're misinterpreting my message. People aren't inherently stupid, but they are extremely susceptive to go down the shortest route without thinking twice (even if it's not the optimal in other regards: distance vs speed vs complexity). As I said, I've seen this at very large scale and there's a clear pattern. Make a survey with 3 questions with no reward and another one with some reward and hand them out to 20k people. You would get roughly 2000 responses roughly 1000 for each group. Give them a fourth question without modifying any of the other values and the responses drop by half, regardless of whether you offer them a reward twice as big in return. The hours of mindless scrolling through social media appears as a one continuous event but if you think about it, the context switching takes place several times per scroll. This is what makes ads so effective. Amazon, Google play or any other marketplace is the same story(take a closer look at Amazon-it's just as crammed with ads). There are a few ways to monetize large user bases: sell them items, which despite effective, will only convert a tiny amount of your users into paying users, not to mention high paying users. Sure, it's a good tactic but all other users are just a water of traffic at the bare minimum(often storage, cpu time, hell, even electricity if you host everything on bare metal in your own data center). As a result, Internet meet ads. I mean there are other ways which are much uglier to monetize your users so I'd rather not go there. As I said, I'm absolutely sure a lot of very intelligent people have reached similar conclusions and that's precisely why the Google play store behaves the way it does. And I'm willing to bet they've tried other tactics as well.