Product

Deconstructing Tinder and the Dynamics of Online Dating

After years of consciously avoiding the online dating game, I recently joined Tinder – mostly out of academic curiosity. I’d read a lot about how online dating works but most opinions don’t survive first contact with reality. This turned out to be partly true in my case.

This post is a mix of things I’ve learnt about product management and online dating through my limited experience on the app and some reading. Let’s begin.

Disclaimer: The dynamics of online dating discussed here are those that operate in India. I’m positive this won’t be the same for other countries owing to differences in cultural and societal attitudes towards dating.

Characterizing the Dating Market

From my understanding of markets and observations from Tinder, I think of the dating market as a highly inefficient market, one that sucks at allocation. This is a good thing, for Tinder. An inefficient market is one where the price of an asset does not accurately reflect its true value. You could argue that it’s better than having no marketplace at all but when you have marketplaces as commonplace as Tinder or Facebook they begin to affect societal attitudes which is not ideal.

This is how the market allocation on Tinder looks like from my understanding. The matches follow a power law for men and a skewed normal distribution for women.

This is not mere speculation and analysis of data from OkCupid also shows a similar trend. You can find the analysis here.

While revisiting the draft and doing some research I came across a paper on Tinder user activity which confirms that this is indeed the case (how to read this graph – e.g. 20% of male participants had a match rate of 6-10%; 30% of female participants had a match rate of 81-90%). You can find the paper here. https://arxiv.org/pdf/1607.01952.pdf

 6

Why does the market look like it does and what are the ramifications?

The market characteristics are driven primarily by the swiping strategy men use on Tinder. The most common strategy employed by men is that of casual, sometimes indiscriminate swiping. This means that women are inundated with matches. Women, consequently, are picky about swiping right and choose carefully. Although a high-level abstraction, this behavior explains why almost all women get a good number of likes whereas most men don’t get any.

Note here that this system has a positive feedback loop, which makes the inequality worse. Consider your average man on Tinder who is unable to find matches. What is the easiest way to increase matches? Swipe more. This leads to women getting even more matches and induces higher selectivity. Overall, this leads to a poor experience for all users on the platform. It is not uncommon to hear of women being harassed on Tinder which obviously sucks but it is a by-product of Tinder’s design. 

3

This built-in inequality of Tinder likes gives rise to a lot of interesting things:

  1. Arbitrage opportunity for Tinder and striving for a more efficient marketplace (More on this in the next section on Product)
  2. By tweaking system design we come up with alternative marketplaces like Bumble where women are given control over swiping strategy (new rules to avoid this particular manifestation of system behavior)

Notes on Product

Tinder’s incentive lies in exacerbating this natural inequality of matches as much as they can (promoting organic behavior) and then introduce workarounds (premium features) to reduce it. This makes for a killer money-making app. ($USD 800 million in revenues this year for Tinder alone)

How does this manifest itself in the app?

Nudging organic/default behavior – If you look at the UI of tinder you’ll notice the subtle push at swiping more and swiping quickly.

  1. Tinder wants people to swipe quickly and finish their daily quota of swipes. To enable this, they’ve added a small but powerful design feature that encourages indiscriminate swiping. It’s the ‘view bio’ button – it’s made really small and it lies in a corner of the app.
  2. In the first, image only mode you can swipe the card left or right. Once you click the bio button, you can only use the icons to swipe right or left. These are incredibly small design features but have significant impact on the user experience.
  3. Once you’ve chosen the appropriate icon to swipe left or right from bio mode you are sent back to the default card only mode which tends to shape user behavior such that you move towards indiscriminate swiping. (Note: I use the term indiscriminate swiping to indicate behavior where you don’t read the bio)

Premium features as a workaround – A multitude of features exist at all price points to entice users into buying them and fixing the match inequality. I’ll briefly touch upon a few. 

  1. Unlimited swipes – Swiping faster means you eventually hit the daily limit. You can buy more swipes to increase the number of probable matches but as we’ve seen this doesn’t work very well.
  2. Super likes – Stand out from the crowd. Since women are flooded with likes (I’ve seen almost all my friends having 99+ matches at which point it becomes tedious to even bother looking at them) a super like has higher visibility because it has a different queue.
  3. Boost – Tinder doesn’t show your profile to everyone you swipe right on. There’s a little more complexity involved here because tinder keeps an ELO score that determines who your profile gets shown to. You can boost your profile to bypass this filter temporarily and have a shot at profiles that are out of your league (as determined by Tinder’s ELO score).
  4. Undo swipe – Swiping indiscriminately often means you have false negatives you want to correct (missing out on good profiles feels like a loss) which can be done by buying tinder gold and having the ability to undo your swipe.
  5. Tinder gold for changing location – Once you’ve exhausted the people around you (I’ve heard of this happening in relatively smaller cities) you can swipe on profiles in other locations using tinder gold.

All these features are aimed at either increasing the potential pool of matches or pushing back against the high selectivity factor induced by indiscriminate swiping. 

Notes on system design and new marketplaces

The Tinder marketplace dynamic is a characteristic of their system design – swiping based on images, allowing everyone to swipe, not showing everyone all profiles, etc. and so is their strategy (of creating inequality and bringing in tools to fix it). The way to create an efficient marketplace is not to have more features (feedback loops are too strong) but to change the system design. And that’s exactly what other people did.

  1. Bumble – Given that most men seem to swipe indiscriminately, Bumble was introduced to shift the dominant strategy of selection from mass swiping to selective swiping by giving the choice to women. Only women can initiate conversations on Bumble and if they don’t do so within 24 hours of matching the match disappears. Since this system is not ideal, Bumble has premium features for men to send super likes and increase the 24-hour window in the hope of getting a response. Bumble would also make for an interesting case study, but I don’t understand the system as much. Fun fact: Bumble is the only dating app not owned by the conglomerate Match.com (which owns Tinder, Match, PlentyOfFish, OkCupid and now Hinge) and was started by an ex-Tinder employee.
  2. Hinge – This started off as a marketplace where connections were made only via mutual friends on Facebook. Hinge now has pivoted to a profile based matching system and is owned by Match.com. Hinge lies somewhere between Tinder and OkCupid in terms of the criterion for matching. Tinder focuses exclusively on photos, Hinge uses a mix of photos and questions whereas OkCupid uses a full questionnaire to show match percentages. Match.com truly owns the dating market and has a solid diversification strategy.
  3. OkCupid – Users on this app are required to answer a lot of questions (in different categories) and matches are made using an algorithm that maps user tastes. The system design here focuses on fewer but higher quality matches.

I think my time on Tinder has been worthwhile given that I’ve managed to write a 1500 word essay on it. I also managed to meet some really cool people I otherwise wouldn’t have. 

Do you have any Tinder insights? Feel free to share them as comments or message me directly. Happy tindering!  

Uncategorized

Thoughts on Information Dissemination and Attention Spans in Today’s World

Whether it is Donald Trump being elected as the Republican nominee, Brexit, or any other serious topic in today’s world, there is a common link that I feel exists behind such instances. In this blog post I’ll try to explain what it has got to do with psychology, biases, and more importantly, thinking.

(Critical?)Thinking

Forget about critical thinking, even thinking in today’s world is underrated. It is too easy to make some believe in a notion, given that you have the apparatus for it. People fall for all kinds of bullshit, all the time. How many of us think before making decisions, whether it is choosing a school, a major, a political party or even a referendum to leave the EU or not? In my opinion, very few of us do that. Even when few actually think, it’s not good enough. That in itself is not surprising since the law of averages works everywhere, meaning most people are average at best when it comes to thinking. Combine this with the fact that people only listen to what they like, add a pinch of ignorance, and use this with short attention spans that we have today, and you can probably explain these occurrences which appear baffling to at least a section of society.

Information Dissemination and Thinking

Here’s a rough infographic I made to highlight what I think is a major issue when it comes to thinking in today’s world. Information penetrates at different levels within the population thereby making the judgement process skewed and in most cases driven by superficial evidence which might be misleading to say the least.

Information Penetration

I might be overestimating the number of people who actually think or I might just be wrong (in which case I would be happy to correct myself), but what I feel is clear is that thinking is a dwindling phenomenon. I discuss this further under Attention Spans.

There is just too much information and not enough time or interest among people to think about everything before taking a decision. I feel this can explain why masses are easily swayed by slogans and why politicians, celebrities, etc feel necessary to pander to the public every once in a while.

The Confirmation Bias

The essence of the confirmation bias is that you see what you want to see. Even when presented with evidence contrary to your beliefs or alternative to your thinking you choose to ignore those, sticking to what you think is correct.

Consider the ramifications this has over the events that I’ve mentioned at the start. Donald Trump supporters cannot see why he’s wrong, they can only hear sentences they want to.

While I’m no expert on the US elections or their mentality, the political polarization is there for everyone to see. Here’s an infographic on Political Polarization from Pew Research Center.

Political Polarization

This polarization I believe is in part created by the media. My dissatisfaction with the media in general is probably worth a blog post and something I might work on later. But just to connect the dots between the confirmation bias, not thinking and the effect of media on public sentiment here’s another infographic from Pew Research Center on Political Polarization and Media Habits.

Political Polarization and Media Habits

Some other interesting finds from this study (I’m summarizing them):

  • Consistent conservatives are tightly clustered around one news channel, meaning they receive most of their information from a single source thereby increasing the chance of bias creeping in. Also, they are more distrustful of news outlets in general. They are also more likely than other ideological groups to hear political opinions in line with their own views. They are also more likely to have friends who share the same views. See anything that strikes you ?
  • Consistent liberals rely on a greater number of sources for information and are in general more trusting of news outlets. They are more likely to unfriend or block people on social media because of political reasons, meaning they don’t want to hear the other side of the story?

It is clear that both sides of the political spectrum have biases which have now led to the creation of the biggest ideological gap in decades.

This article is not meant to discuss political ideologies, nor am I commenting on or comparing the two. I just happened to take this up as an example to explain my views on why these things could have happened.

Attention Spans

With the world becoming more and more “social” and the constant and unending flow of information, our attention spans have reduced to those of gnats. There is not enough time to put in the effort to understand and process information, but only enough to consume.

Here are some statistics from National Center for Biotechnology Information, U.S. National Library of Medicine and The Associated Press on Attention Spans and how they’ve reduced over time.

Attention Spans

Here’s a very click-baity article on the same: http://www.telegraph.co.uk/science/2016/03/12/humans-have-shorter-attention-span-than-goldfish-thanks-to-smart/

There’s even a rebuttal saying Millennials are accused of having shorter attention spans and that is just because they have much better stuff to do. While that is very subjective and may be true to an extent, I disagree and find that a convenient excuse to ditch critical thinking.

Just to summarize this piece: Humans are prone to biases; not many people are up for thinking; the media has a role to play in this; shorter attention spans are making the problem worse; all of this can lead to instances that are high impact and may even be disastrous.

Just jotting down a list of topics that we considered important and now are as good as forgotten:

  • NSA spying scandal
  • Panama Papers
  • Every other mass shooting that is happening
  • Russia’s annexation of Crimea

The list is probably bigger and even I might have missed many but you get the gist. Issues, even important ones that don’t stay in the limelight are easily forgotten, gone, poof.

Bonus Trivia: The newest and fastest growing social media platform Snapchat allows you to post snaps or video stories with an upper cap of, well as you might have guessed, 10 seconds. Coincidence much? I think not.

Also, if on Snapchat, you can follow me on rrahul30. 😛

Disclaimer: All views expressed are personal. As is with humans, I might also have certain biases or pieces of information that I might have missed; although I have tried to present a very fact-based opinion piece, any comments to ensure correctness are appreciated.

References:

  1. Infogr.am used to create infographics.
  2. Pew Research Center
  3. Thinking, Fast and Slow by Daniel Kahneman
  4. Predictably Irrational by Dan Ariely

P.S. If you liked this, please give this article a heart (recommend) so that it can reach more people! Thanks ☺