The secret to Tinder's success isn't that they use a swipe-based interface. It's that they solved one of the biggest pain points in dating: users never have to talk to anyone that hasn't already expressed mutual interest in them. Since the app discloses romantic interest only after both parties have reciprocated, Tinder offers a fundamentally better way of meeting new people than has existed previously. That was a game-changer for the entire dating scene, and that is why they are successful.
Small Data: Tinder-like apps are the way of the future
11–20 of 37 posts
Re: Small Data: Tinder-like apps are the way of the future
#12This is so dumb...
One paragraph they're introducing the concept of "small data", the next paragraph they're talking about how getting the swipe information for huge numbers of users helps them train their collaborative filtering algorithm...
Ultimately it's a big data hype article... GASP we can infer your preferences from your behavior... Written by someone with a Medium tagline of "Sassy futurism. Tech and words."
Re: Small Data: Tinder-like apps are the way of the future
#13One big false generalization. I mean Netflix's interface sucks for discovery, and the suggested improvement in the article is to force me to look at one suggested movie at a time? Seriously!?
Any claims for discovery efficiency is a lie. Shitty UX's "convert better" due in part to the way that they hurt their users. Netflix cannot let you discover there is nothing you want to watch in their entire catalogue because you will unsubscribe. They have to make you believe there is more that you have not found. Similar to Amazon, they cannot make search so efficient that you have the time and energy to search ot…
I think you're confusing search intent with discovery intent.
Of course Netflix isn't going to stop showing you movie options when you're trying to discover something new. If the user doesn't intentionally know what they want to watch, they need to continually see content to discover it.
And Amazon search isn't terrible because of a price-hiding scheme, it's because they have a terrible UI company-wide, from the Kindle to EC2. People typically aren't browsing Amazon to discover new products - they search and move on. Are you suggesting that Amazon search should be an internet price aggregator too?
It seems like you're playing the Halloween music early by saying these companies have terrible UX as a complex ploy to keep you engaged while not giving you what you want. The reality is that they have wide userbases which are fickle to UI changes, and it takes time and research to roll out anything new.
Re: Small Data: Tinder-like apps are the way of the future
#14This is so dumb... One paragraph they're introducing the concept of "small data", the next paragraph they're talking about how getting the swipe information for huge numbers of users helps them train their collaborative filtering algorithm... Ultimately it's a big data hype article... GASP we can infer your preferences from your behavior... Written by someone with a Medium tagline of "Sassy futurism. Tech and words."
That is not at all dumb. Doing the preference learning for a huge number of users lets you set up a prior over the preferences that includes correlations between different attributes. Then you use the population prior together with the data for the individual to get the posterior for their personal preferences, which will then take into account the correlations revealed in the population data. This is what we do with http://findmelike.com where we perform the Bayesian update over the user's preferences after every user interaction.
Re: Small Data: Tinder-like apps are the way of the future
#15This is so dumb... One paragraph they're introducing the concept of "small data", the next paragraph they're talking about how getting the swipe information for huge numbers of users helps them train their collaborative filtering algorithm... Ultimately it's a big data hype article... GASP we can infer your preferences from your behavior... Written by someone with a Medium tagline of "Sassy futurism. Tech and words."
That is not at all dumb. Doing the preference learning for a huge number of users lets you set up a prior over the preferences that includes correlations between different attributes. Then you use the population prior together with the data for the individual to get the posterior for their personal preferences, which will then take into account the correlations revealed in the population data. This is what we do with…
Ha ha... I'm not against learning preferences at all... I think you misunderstood my comment
Re: Small Data: Tinder-like apps are the way of the future
#16Re: Small Data: Tinder-like apps are the way of the future
#17Earlier quoted context omitted.
Any claims for discovery efficiency is a lie. Shitty UX's "convert better" due in part to the way that they hurt their users. Netflix cannot let you discover there is nothing you want to watch in their entire catalogue because you will unsubscribe. They have to make you believe there is more that you have not found. Similar to Amazon, they cannot make search so efficient that you have the time and energy to search ot…
> Netflix cannot let you discover... They have to make you believe there is more that you have not found. That's an insightful observation and explains perfectly why Netflix does not provide a simple straight list of all titles that you can scroll through. I'll bet that the Netflix engineers have wanted to add such an obvious feature, but someone in marketing yelled, "Don't!" If you ask the average non-techie user ho…
7,750 is a shockingly low number for a service like Netflix, but I guess it reflects the fact that most movies aren't really that good and you'd probably only be interested in watching a maximum of 10k titles over your lifetime, maybe 15,000 hours / 2 years of content
Re: Small Data: Tinder-like apps are the way of the future
#18Earlier quoted context omitted.
Any claims for discovery efficiency is a lie. Shitty UX's "convert better" due in part to the way that they hurt their users. Netflix cannot let you discover there is nothing you want to watch in their entire catalogue because you will unsubscribe. They have to make you believe there is more that you have not found. Similar to Amazon, they cannot make search so efficient that you have the time and energy to search ot…
I think you're confusing search intent with discovery intent. Of course Netflix isn't going to stop showing you movie options when you're trying to discover something new. If the user doesn't intentionally know what they want to watch, they need to continually see content to discover it. And Amazon search isn't terrible because of a price-hiding scheme, it's because they have a terrible UI company-wide, from the Kind…
Can you explain the difference between search intent and discovery intent
Re: Small Data: Tinder-like apps are the way of the future
#19Earlier quoted context omitted.
I think you're confusing search intent with discovery intent. Of course Netflix isn't going to stop showing you movie options when you're trying to discover something new. If the user doesn't intentionally know what they want to watch, they need to continually see content to discover it. And Amazon search isn't terrible because of a price-hiding scheme, it's because they have a terrible UI company-wide, from the Kind…
Can you explain the difference between search intent and discovery intent
I want some shoes, I shall filter the shoes to match my preferences.
I feel like shopping, I wonder what's on sale.
Re: Small Data: Tinder-like apps are the way of the future
#20Earlier quoted context omitted.
Any claims for discovery efficiency is a lie. Shitty UX's "convert better" due in part to the way that they hurt their users. Netflix cannot let you discover there is nothing you want to watch in their entire catalogue because you will unsubscribe. They have to make you believe there is more that you have not found. Similar to Amazon, they cannot make search so efficient that you have the time and energy to search ot…
I think you're confusing search intent with discovery intent. Of course Netflix isn't going to stop showing you movie options when you're trying to discover something new. If the user doesn't intentionally know what they want to watch, they need to continually see content to discover it. And Amazon search isn't terrible because of a price-hiding scheme, it's because they have a terrible UI company-wide, from the Kind…
>> And Amazon search isn't terrible because...
Amazon may not be intentionally malicious. When you make decisions only on a metric like sale conversions, you will be optimising other factors you are not even aware of. Nothing necessitates that the best UX and happiest customers will be the most profitable.
For example, I recall a story about Amazon trialling infinite scroll. It had positive customer feedback and users viewed more products compared to pagination. However, sale conversions dropped so obviously it was never rolled out. Was it a better browsing experience? Possibly, but that isn't the metric Amazon is optimising for so it just doesn't matter. I expect that there is a threshold where any time saved for a user searching Amazon is not spent buying more products from Amazon but spent on competitor sites.