Hacker News Reader

Comment by feanaro | Hacker News Reader

I must say your dismissal of this is rather annoying. Multiple users are telling you they are facing a real -- and completely obvious and predictable -- problem due to the feature not being available. In a comment section regarding research showing it would be an improvement to recommendations.

Increased UI clutter due to a single extra icon is not a good nor sufficient argument for not including it.

View on HN

My dismissal is based on a many-million digit customer base and online real world data. Sorry if a few hackernews users feel they want a dislike button, most users don't and if you give to them most users don't use it. I don't mind disagreement but dismissing real world data from this industry is sort of silly to counter with 'because I want it you're wrong'.

I'm not even saying you're wrong, I'm saying your argument for not including it for (at the very least) a sizeable minority is weak and somewhat silly.

What is a sizable minority aside from an oxymoron?

A sizeable minority is something that is still visible on a pie chart.

A minority is anything that is not a majority (this is the definition I'm using). So a < 50% group can hopefully still be viewed as non-negligible, even though it is not the majority.

I disagree that your reasoning is entirely sound. It's a bit of an elitist point of view I suppose; but fundamentally I don't think all customers are created equally in terms of importance. One could even in this day and age call some of the more important ones "influencer" or something similar.

While the general customer base might not be served by this feature, what if a subset of the customer base enjoys this feature to an extraordinary extent? Maybe it fixes a lot of problems they have with the service; and negates a lot of complaints they otherwise would have posted.

Would it still not be the correct choice to "clutter the UI" with this?

I don't think it's as easy of a choice like you're trying to make it.

But I could also be jaded by the fact that Spotify for years has trying to pull me into listening to all kinds of odd European folk music that I have no interest in what-so-ever.

For a long time we called this sort of user a "power user". People who are not necessarily technical but who love your application and use it extensively will appreciate power user functionality.

It's probably fine to not have anything to appeal to power users, but those power users probably will be the sorts to be influencers to talk up your application to everyone they know.

It’s interesting isn’t it, why did we stop talking about power users and catering to their needs?

Because the masses are now large enough that you don't need to have an actually useful application. You just need to make it kind-of work and for it to become popular.

I think for technically minded or abnormal users a dislike button would be valuable, but that's a minority of the population and I understand why a company would choose to focus on the larger market to the detriment of the enthusiasts. Props for taking the heat and trying to present this perspective.

I personally had these same complaints and acknowledged that I'm unusual and my needs are best served by my own hosted solution. I'd encourage people who are unhappy with major streaming services to roll their own if they truly feel strongly about it- I did and I'm happier for it.

How could people dismiss real world data that you didn't share? You only shared your interpretation of it and people pointed to lots of ways that this interpretation could be wrong. What data is this? Did you conduct a survey asking people if they “want” a dislike button? Or did you just measure usage frequency of the dislike button and decided that people didn't want it because it wasn't used as frequently as the like button? How could low usage frequency mean users don't “want” it?

This totally ignores effects from the long tail, black swans, power users, and criticality dynamics.

Downvotes are way more "potent" than upvotes. If I'm listening to music that I like, I'd say I probably log over 100x upvotes than downvotes. But when I downvote, I usually want to nuke the site from orbit.

What is the ratio? What ratio of users use downvote? What ratio is that compared to users that interact in any way? How are you controlling for sampling bias? Berkson's paradox? How much utility are the users that downvote obtaining, vs the disutility of being unable to?

Reply on news.ycombinator.com