Tinder Feed

How Tinder Created the “Feed”

A hypothetical explanation/internal conversation.

No highlights, no onboarding… All of a sudden, out of nowhere Tinder dropped its “Feed” on users. It seems to me that it’s a Tinder launch strategy: dropping features on users without any explanation. Like that time they introduced Stickers and it was actually more buggy on Android than when Android released its first OS.

As a Product Manager I am always curious to understand which problems new features or designs solve for, which prioritization frameworks were used and why certain features were launched to begin with. Especially if I can’t find a good explanation, no matter how hard I try.

Late at night I try to replace myself in the PM’s shoes and imagine the conversations that happened prior launch. Somewhere between going from problem to launching a feature. And that’s exactly what this post is about.

The hypothetical explanation and conversation at Tinder how the feed came about…

The problem

Sean Rad: “Can someone explain to me why we’ve been around for 5 years and we’re still “just” a hook-up app? We didn’t expand in other verticals or didn’t launch new products. WHY? What have we been doing all that time?”

Product Manager Shaun: “We kind of were busy monetizing our platform through at least 5 experiments that…”

Sean Rad: “Why has Bumble expanded to finding friends and mentors and all we’re seem to be doing is staring at our sinking retention rates. Hell, they are sinking faster than the Titanic did.” 

PM Shaun: “We have a hypothesis that people drop off because they find a match and no longer need…”

Rad: “That’s great Shaun, but we need to beat Bumble. We didn’t dispute sexual harassment in court for months and lose a sack of money to then ALSO lose our users to Bumble as well. Fix our sinking retention rates.”

Shaun scratches his balding head, takes a deep breath, and walks back to his desk.

The ideation

A few days later PM Shaun gathers his cross-functional team in a sweaty meeting room looking over Santa Monica beach.

PM Shaun: “Team, I have an exciting new challenge. This is going to change everything. We need to get more users hooked so they don’t drop-off and by the way.. we have no limitations. Considering that we have to ship it in less than 8 weeks is not réally a limitation but more something we just need to work with. Aight? Let the ideation begin!”

Engineer Paul:“No limitations? How about we start with completely engineering our architecture so we can properly scale? Remember that time when we launched stickers and it’s still buggy? Architectural issues. And btw: let’s -for once and for all- rewrite our Python code leftovers in Node.”

Engineer Rudy: “Plus one”

Shaun: “Guys, let’s hold our engineering horses. I heard your suggestions, I know about the back-end work we need to do, and I’ve slotted that on the roadmap in 2022. Let’s focus on the user here.”

Designer Eddie: “Can we ideate around the user experience, pals, instead of jumping to solutions? How can we create something that will fit the daily lives of one of our personas? Let’s storyboard this.”

Eddie:“Let’s imagine Johnny again, our 25y old hot guy that swipes in the morning to find a date for tonight…”

The whole team looks kind of uncomfortable.

Eddie: “Does anyone remember our persona Johnny and the problems he has?”

The whole team minus PM Shaun: “Not really”

Eddie:“Le sigh. Alright. So the main problem of Johnny is that he is a last minute planner. He is a good looking dude with no issues scoring dates, but he plans last minute.” 

Data scientist McCarthy: “Which is an issue because we know that it takes more than 12h to arrange a date with newly matched girls that score 7/10 on the attractiveness scale based on the output of the algorithm we’ve built. No denying, the p-value was 0.01.”

Shaun: “So if we want to help Johnny, we cannot count on matching with new chicks. We need to find a way to remind Johnny about the old matches he has.” 

Eddie: “How might we help Johnny find talking points to re-engage again with past matches, knowing he has <6h to arrange a date?”

Engineer Paul: “That’s basically creating a feed. I never understood why we didn’t show a feed. Facebook does it. Instagram does it. Everyone does it except for us. It seems obvious.”

Engineer Rudy: “Plus one”

Shaun starts dreaming about the all the possible ways this can increase retention and gets excited.

Shaun: Guys, this is brilliant. With a feed we can push every change of any matched profile and inform our users about it. Hell we send them push notifications for every change their matches made!!

Designer Eddie: “What about the user experience? Does Johnny want to be notified on his home screen at 3 AM that Heidi has updated her profile with a picture of a goat?”

Data scientist McCarthy: “Technically we’re still working on recognizing animals in profile pictures, but we’re not there yet since we wanted to focus on recognizing nude first.”

Shaun: “Thanks for this wonderful update McCarthy!”
“Alright. Push notifications are out. I’m sure we can find a way in-app to make users aware of the feed. Eddie, you can roll with that?”

Designer Eddie: “Sure thing.”

The design

12 weeks later after numerous rounds of user research, wire framing, user-testing and 4 post it blocks used, Eddie & Shaun proudly presents a design to the team that includes a scrolling list of updates that matches made to their profiles.

Designer Eddie: “Team. We have a first draft of the feed here and we would love your feedback and engineering implications. Let’s start with the feedback.”

Engineer Paul: “Yo guys. This is all fine and dandy, but imagine Johnny notices that Heidi posted this wonderful picture of her goat and Johnny is now madly in love with Heidi and wants to ask her out. How will Johnny get in touch?”

There’s a long, awkward silence. As if the Titanic all of a sudden erected from the bottom of the ocean.

Eddie: “Euhm.. yes… right… hmmm.”

Shaun: “Let’s not overcomplicate this. We had to ship this 4 weeks ago. We just add a box underneath each update where they can type in a message. Right, Eddie?”

Eddie: “Actually.. The correct design pattern would be to add it in a drop down on the right hand side.”

Shaun:“Even better. Two places to get in touch with a match. Brilliant.”
“McCarty, wake up. You provide instrumentation requirements to measure whether this is increasing retention and while you’re at it, can you try figuring out what the trends are in the messages people write from the feed?”

Data Scientist McCarty:“Sounds like a fun project. Is this more important than recognizing animals in pictures?”

Shaun: “No it’s not. After all we can correlate message content to picture changes. Wouldn’t that be useful?”

Engineer Rudy: “Plus one”

Shaun: “Eddie, can you update the designs to include the messaging features? ”

Shaun: “Alright. That’s settled then. Back to our computers!”

The result:

6 months later, hugely behind schedule:

Picture courtesy to Tinder Blog

The disclaimer

I have no connections at Tinder, so this post is purely hypothetical. It’s based on my own imagination at 3 AM seriously wondering why the feed was launched and which user/business goals it meets.

If you work in tech and this was recognizable in any way, or it worked your humor muscles, feel free to leave a comment and satisfy my 3 AM brain connections.


Leave a Reply

Your email address will not be published. Required fields are marked *