#45 The End of Loneliness
Dating apps are dead, long live dating apps.
Hiiii. We’re so back. I forgot to celebrate last week, but September marks the first anniversary of Oblique Forecasting!
A bit about me: I’m Esther Moisy-Kirschbaum, and I live in Paris after ten years spent abroad. I currently serve as the Head of Content for Magma, uncovering upcoming business trends. OF serves as my more personal take on the next it-business, and even though editing Magma helped me become more rigorous when it comes to sourcing and methodology, this Substack can lean a little more vibes-based.
Now, a bit about you:
OF is read across 75 countries (!), with roughly half of my readership based in France.
In one year, I went from 160 to nearly 1400 subscribers. I’m not being recognised in the street for it yet, but maybe one day. Many many of you come from recommendations—don’t gatekeep OF!!—; my top three sources are personal finance media Snowball, brand strategy newsletter Post-Culture by Sibling Studio, and design newsletter Dreamspaces. Seems like we all contain multitudes then :)
My most popular post is the one below, detailing my personal interest in health and longevity, an update on an older post:
Warning: this letter is a bit image-heavy for your inbox. Make sure to open it in your browser! Enjoy xoxo Esther 💋
What I’m divesting from
Dating apps… again.
Nearly a year ago, I published this piece diving into some obvious issues within the current dating app landscape hellscape:
And here’s a deep-dive I wish I had written about Hinge, one of the current key players, owned by Match Group:
Some notable changes in 2025:
One of the newcomers I highlighted—and the first dating app I became aware of that leveraged AI—, Gigi, has pivoted entirely from dating to… something to do with your Social Capital score?? Unclear what purpose it serves.
Bumble is abuzz with Swiped, the biopic of its founder, Whitney Wolfe, out this month. Meanwhile, here’s the first sentence of a WSJ article about her and her next AI dating app idea: “Whitney Wolfe Herd has a confession: She would never have swiped right on her husband.”1 A few lines below, “Wolfe Herd, who helped build the modern dating-industrial complex as a co-founder at Tinder and a founder of Bumble, believes artificial intelligence could be just the matchmaker we need.” Does that sound familiar?
Well, everything is AI now:
And that was just human x human dating, because there’s also human x bot dating to contend with as well:
But it’s not all doom and gloom! Here are a few considerations I would take into account, should I inject money into a new dating app.
Generational adoption
Every successful platform has been catalysed by the it-generation of its time2: Facebook by college-age millennials in the late 2000s, Tinder by the same millennials in the early 2010s. But they’re now grown, some of their cohort are married with children, and they find swiping passé. Gen Z users, who grew up in a post-#MeToo and post-COVID world, don’t relate to the dynamics either, meaning no one likes the process nor the (lack of) results! If the incumbents were built by and for millennials, the next wave needs to be built by Gen Z for Gen Z. That’s the only way to create something sticky enough that older demographics (and in turn, younger, when Gen Alpha grows up) will eventually follow. I’d back a team of Gen Z founders building an app that reflects their social codes. Less gamified, more consent-forward, and more fluid across gender/relationship identities. I would bet on a Gen Z-led team that’s culturally fluent and willing to throw out the swipe deck entirely (!!).The algorithm trap
Hinge claims it’s “designed to be deleted,” but the real experience feels like being trapped in algorithmic jail. Anecdotal evidence: many people around me report that their long-term partners were among their first matches. Reading between the lines: they got lucky before the system learned how to lock them into the attention economy. Tinder and Bumble’s Elo system and Hinge’s Gale-Shapley stable marriage model optimise for utilitarian balance, not meaningful connection. Translation: you’re matched with someone “good enough,” but rarely with the person you’d actually be most excited to meet. The stickiness comes from frustration, not fulfilment.
A breakdown of the algorithms, skippable if you don’t nerd out on that kind of stuff:
Tinder and Bumble’s algorithm used to be based on the Elo rating system, which was originally designed to rank chess players. You rise in the ranks based on how many people swipe right on you, but that was weighted based on who the swiper was. The higher rank (more desirable, more right swipes) they have, the more their swipe weights on your score. Thus, those with similar scores will see each other more often.
In the first few weeks of signing up, you will see profiles with all different rankings. Just an analogy, if you like most profiles who rank 8/10 and they like you back, you’re an 8. If you like most profiles that rank 8/10 but are only liked back by people with a 3/10 ranking, you are more likely to be rated a 3.
Hinge uses the Gale-Shapley algorithm, which is a Nobel prize-winning algorithm that finds optimal matches for the ‘Stable Marriage Problem’. It’s about pairing people who are likely to mutually like one another. The gist is that you don’t get matched with your ideal, but rather you get matched with whoever supports the result that allows everyone to get a utilitarian optimal match. Stable doesn’t mean perfect, not everyone is going to be completely satisfied with their pair, but they also wouldn’t prefer anyone else in the choice pool, since their preferred choices are also taken.
There’s an opportunity for an app that ditches ranking systems altogether in favour of transparent, user-directed matching. Think: filters you control, matchmaking modes you toggle on/off, even periodic “resets” that surface people outside your algorithmic lane. I would back founders willing to prioritise user control over stickiness.
A new definition of success
Right now, dating apps measure success by engagement time: how long you stay swiping, chatting, paying for subscriptions and extra features. But the future of dating apps won’t be built on endless engagement loops. It will be built on proof of outcome: did people actually meet, connect, date, and build relationships? The next wave could monetise not by keeping you online, but by facilitating your offline transition. That might look like integrated event discovery, partnerships with IRL spaces, or structured “first date credits” that expire if you don’t use them. The next unicorn will be the one that earns when users form relationships, not when they keep swiping. I would look for teams experimenting with offline integrations and business models aligned with human connection, not attention capture.
If you’ve made it this far, consider adding a like, a comment or sharing the post.
📈 Bullish news
Is there really no such thing as bad publicity? The controversial “Sydney Sweeney Has Great Jeans” campaign has proven to be a commercial win. CEO Jay Schottenstein credited Sweeney with fueling an uptick in “customer awareness, engagement and comparable sales.” The campaign launched just a week before Q2 closed, but still helped beat revenue expectations; it drove double-digit denim sales growth, generated 40 billion impressions, and sold out the “Sydney Jean” in one week. (WSJ)
Robinhood launches Robinhood Social, allowing users to post their trades, takes and commentary while following other users as well as avatars mirroring the trades of politicians and executives. (Axios) Can it bypass Reddit and X?
Wall Street bet on AI is too big to fail (Axios). We’ll see about that!
📉 Bearish news
Tinder x LoveShackFancy = the fragrance collab no one asked for
Jaden Smith was named Christian Louboutin’s first men’s creative director (The Fader). Why??
Sounds familiar!!! God, it’s hard being so right all the time:
Just like you need hot women to populate your nightclub to attract all the other population groups, an app needs the it-generation to adopt it, and other generations will follow!
The algorithm breakdown is so helpful, especially the part about Hinge's Gale-Shapley model optimizing for utilitarian balance rather than actual chemistry. That explains so much about why Match Group's apps feel like they're trapping you in just-okay matches. Your point about generational adoption is crucial too, every platform needs its it-generation, and you're right that Tinder was built by and for millennials who are now aged out of the swipe culture. The Gen Z observation really hits, they grew up with these apps already enshittified so they never expereinced the early promise. What's really smart here is recognizing that the next unicorn won't be built on engagement metrics but on proof of outcome. Match Group's entire buisness model depends on keeping people swiping forever, which is fundamentally misaligned with what users actually want. The offline integrations you mention could be the unlock, but it would requir Match Group to completely rebuild their revenue model around user success instead of user frustration.
Happy one year, Esther 🫶