Hacker News new | comments | ask | show | jobs | submit login
Show HN: Tinder Revenue Estimate (manicoti.com)
137 points by ebellity 10 months ago | hide | past | web | favorite | 105 comments

Feedback: consider ditching the .00 on the numbers and spelling out "million"/"billion." It's a lot easier to quickly read "$36.2 million" than "$36,200,000.00"

At least numbers looks humongous at first glance with it. :)

Go the federal budget route, round down until you get to "in thousands" or "in millions" for everything. Then it's still easy to compare $12 to $36,000 at a glance without staring at a table full of 000,000.00 :)

It's about information entropy, you want to maximize the information each character has. using 1000 digits to show what you could in 500 and a header saying "(in thousands)" maximizes the entropy

Or better yet, $36.2M

Here's one I always wonder about:

$36.2M; $36.2m; $36.2mm; ?

K, M, B, T, Q

thousand, million, billion, trillion, quadrillion. I'm not a fan of M and MM and MMM, it's much easier to use one letter that is different from the others.

Well. A Billion is ambiguous. In many countries, it's called a Milliard. See: https://en.wikipedia.org/wiki/Long_and_short_scales

I don't think the fact that there are multiple words that represent the same thing means that one is ambiguous. A "billion" is always 10^9, so it's not ambiguous, no?

A "billion" is 10^9 in the short scale and 10^12 in the long scale.


It looks like the Long scale is falling into disuse in English speaking contexts from that article. I don't ever think I've ever seen the word "billion" in English refer to 1e+12 .

No, it's the other way around. A billion means one thing in english speaking countries and another in most of the rest of the world. For a frenchman, it's indeed ambiguous -- there, a billion (english) would be a milliard (french), whereas a billion (french) would be a trillion in english. I imagine that is the reason why it's not seen as an abbreviation, in contexts where you want to be accurate.

I think MM is the common way to denote million in finance.

I believe it comes from one thousand (roman numeral M) thousands (roman numeral M) or MM.

Can't upvote this enough. I get the urge to just stop reading whenever I come across "MM". Just use "m" or "M" for millions, for heaven's sake.

It's a finance thing. M = roman numeral for 1,000.

Will i18n ever know peace?

I think this depends on the individual. I always prefer the version with commas instead of words. It's easier to copy and paste for one.

Agreed. For sometime I was scratching my head trying to understand the exact figures. The standard format is "in thousands".

Also, right-aligning numbers that readers want to compare is generally a good idea.

I agree to drop the ".00", but I prefer the number $36,200,000.

Make it configurable!

Right, thanks

Congrats on shipping! One suggestion: I'd make the URLs prettier e.g: /companies/<id>/stripe/bmodels, /companies/<id>/stripe/submodels/1, /companies/<id>/stripe/submodels/2, etc.

That way I can delete the last part of a URL a return to the /companies/<id>/stripe/ home page. It's also easier to know what the URLs are about before opening them.

agreed, thanks for the feedback !

IAC has a chart showing total revenue of Match group, which includes tinder. Page 6


BTW Love the quote "Match Group proved definitively in 2017 that Tinder is not just a moment, nor just a solid business reflecting an enduring commercial testament to the power of love. "

TIL that IAC owns Tinder, Vimeo, Dictionary.com, Investopedia, Ask, and a bunch of other stuff.

Their stock has tripled in 2 years, also.

"Commercial testament to the power of love".

Any idea what author was trying to say??

I think what he was trying to say is "Dollar dollar bill, y'all"

actually now I reread it I'm not sure either. The rest of the paragraph is:

"Tinder gave rise to a massive expansion of Match’s market and became one of the largest mobile platforms globally, capable of a profound and positive impact on the world. Nearly every week we read news stories around the world of people celebrating weddings for couples that met on one of our properties"

Which doesn't make it sound quite as sleazy.

He's saying: "The success of Tinder shows that monetizing love is a profitable model."

But its not. Tinder relies on many months of your membership in other to make a buck. If you find love, you stop being paid member.

I say to the contrary: the financial success of Tinder proves that online dating doesn't mean quickly finding true love.

"love" ...right.

I have friends who got married after meeting on tinder. I don't think they spent very long on the platform though, they never paid for it either.

I don't really get it but it seems like people are getting married who met on Tinder at this point.

What a time to be alive

You mean what a time to be single?

Really cool idea. It would be awesome if you could crowdsource the data for any companies through employees - some big companies don't have much transparency on revenue numbers, but employees in various positions could piece that information together through this service. Maybe even the rare private company that offers their revenue information willingly.

Reliable data on revenue of private companies would be very useful to employees/investors in determining the actual value of the company's equity.

Yeah I'm sure a company like Buffer could be willing to share that information (like they already do for their revenue on the Baremetrics "open startups" page)

I agree employees (and past employees) could offer valuable information, although they're probably legally bound by some confidentiality agreements...

However I think when you're in a market it's easier to estimate how your competitors are doing, even though you don't have any privileged info.

Nice work and grats on getting to the elusive "done" stage :)

Income tax is a big piece missing from the expense model. I'm guessing it's easy to add that line item, but it even if so it would be cool to have it computed automatically since its mostly just a function of {domicile, entity_type, pretax_rev}.

Thanks, and yes it's elusive :) Sure, will need to add this. Can't avoid taxes!

Cool! I think evidence-backed modeling is a really interesting idea.

As an unsolicited suggestion, I wonder if there's a way to emphasize numbers that are backed by evidence (links and maybe commentary?). Right now from a visual and informational standpoint all the numbers look like they're equally valid when in reality some of them could be based on reports published by Tinder and some could be just totally made up.

If you scroll down it has links and comments about where all the numbers came from.

Yeah, but none of that information is surfaced in the high-level report. If I'm reading this I'd like to have an idea of each number's validity.

Definitely something I'm thinking about. Maybe even some kind of "confidence interval" of confidence ratio could be added next to an estimate.

I think this number is a massive underestimate of their annual top-line. Based on the cost of a subscription I'd bet it's in the $1B+ range.

Agreed. Having worked at a dating app that had a significantly smaller user base, the revenue numbers are definitely low balled.

Not necessarily. Dating apps like Tinder that are focused on hookups have much lower paid subscriber numbers relative to apps focused on relationships.

Is there a variable in particular you think is too low ? I've taken the number of subscribers (3M) from Match Group report. But I've taken the assumption that there's high churn and they don't all pay for 12 months. Let me know what you think, it can be updated :)

The subscription price varies substantially according to age and region. I also suspect that they make a lot of money from "whales", i.e. people with the most expensive subscriptions who also buy their other item packs (superlikes etc.)

I spent a ton of money on boosts, and have a gold membership, easily $200+. It was worth it. Truth is for busy men with a bunch of shit to do these enhancements completely improve the experience to the point that they are practically mandatory. Any time I’m on a serious Tinder spree I’ll buy up big packs of boost. Boost gets you laid.

>Any time I’m on a serious Tinder spree I’ll buy up big packs of boost. Boost gets you laid.

Agreed. I don't know how they did it, but they finally made a dating product that works for women AND men. Match Group is going to the moon IMO, and I say that as someone who worked for their #1 competitor.

wow, interesting

True, there might be people spending a lot more than the average cost of a basic subscription. I don't know the product well enough to evaluate if having Tinder Gold is enough or if some users also buy one-off app purchases in addition to that.

Yea I spend $100 or more on Boosts a month in my market, NYC.

thanks, I will try to update that ARPU number

This is an interesting way of decomposing a business model in general, not just documenting existing companies. I could see something like this chipping away at the Excel models that we all made in business school. Reusable inputs, explicit sourcing of assumptions, creatively visualizing how to arrive at the outputs, easier to run what-if analysis.

Not sure who the market would be as a paid service, but would be being a great teaching tool.

Yep, you just read into my mind ! Also think it could be great for education, in the information age it's still not easy to understand how businesses work exactly for most people.

I did an internship for a mobile games studio. They had really detailed estimates of competitor revenues because they were able to combine:

1) their own daily revenue from titles

2) publicly listed studios' quarterly revenue figures

3) the daily rankings of 1 and 2 in the two app stores

to extrapolate every rivals' revenue.

I'm pretty sure tinder grosses more than $1 m / day (excluding ads)



Thanks, someone said pretty much the same thing to me but couldn't share the data. I'll increase the ARPU number again to reflect that, sounds like they're extremely profitable...

I like the site's intended purpose, but I agree with others that design needs improvement. Also, perhaps you can source some data from public data (i.e annual report, appannie).

thanks for the feedback. It's funny now that you mention AppAnnie I just remembered an old project I did in 2010 called AppsMetrix trying to estimate the number of app downloads based on what data was available (number of ratings, reviews, some apps who communicated about their downloads numbers etc). Somehow I keep coming back to working on the same kind of stuff!

Cool stuff.

I've noticed that some times you source numbers from news articles, which you reference. How do you plan to keep the numbers updated?

EDIT: In fact, how have you done it so far? Manually, I imagine?

Yes I've done that manually. In my head I would love for it to be something even more collaborative 'Wikipedia-style' and people would help keep the numbers up to date. Of course that could only work for popular companies if there is enough interest about their business.

For smaller businesses it could be cool to have templates like "pizzeria business plan" that people could "fork" and input their own numbers. Perhaps reuse some variables across the board such as "cost of tomatoes" instead of everyone doing their thing separately on Excel.

You should allow requests to add or alter data, mostly like nomadlist.com do (example [1]). The owner could discuss and accept or reject the request. I think it would enrich the plan much quicker. Congrats for the launch!


Thanks, I agree, maybe like a "pull request" that can be accepted / validated by the owner

Exactly, or even "clone" the plan. Seems cool, git applied to tabulated data.

This interface generalizes nicely to estimating any Fermi problem, not just business models. I love it!

Thanks, agreed it could be extended to other types of models and estimates !

I really like what you're doing here. Two things that I would like as a user: -ability to get updates (pref email) when new companies are added or companies I follow are updated -timeline for the financials. are these annual revenues? monthly? etc.

Nice work on shipping!

Thanks ! Indeed, that's actually something I should have added before posting here (to collect a few emails for future updates).

Timeline is also a big area of improvement. It's good to be able to see a snapshot at a moment in time, but I'll have to develop ways to make it more clear which it is, and perhaps to extend that and make projections for months/years to come.

Update : I've now added a top banner to collect emails for people interested in company news

The actual revenue may be in Match Group or IAC’s Quarterly or Annual financial filings on sec.gov or Morningstar. It’s also possible the revenue was teased out in one of their earnings call transcripts as well.

From what I've read they just gave a few hints such as the number of subscribers for Tinder Plus or the overall increase in revenue for Match Group, but not a lot more than that.

This is a cool idea! I'm not sure why but all the submodel pages say "There is an error in your formula." next to the current formula even though they seem to be the correct calculations.

I've fixed that, thanks!

I love the idea here, but I don't find the report very easy to read.

yeah I'm not completely happy about how it looks either, will probably have to work with someone better on design/UX for that. if you have ideas on how to improve or feedback on the parts that were not clear don't hesitate

in my mind that looks something like a spreadsheet but more structured, say like the old


There is a large precedent of people writing financial reports and projects so I would look at accounting literature or talk to an accountant (possibly in a non-professional relationship) or business analyst about it before talking to a UX person.

I think I'd like to see an overview calculation at the top (which I think I see) and then see it broken down hierarchically below. I think that might be what you are trying to do, but it doesn't come across clearly.

Another issue is that this is described as a revenue report but I see stuff in there about salaries, desk costs, etc. which is not revenue but expenses. A real financial report would probably consider both and should be completely clear about what is what.

Thanks you're right I've updated to "Revenue & Expenses Estimate" on the page title.

Indeed there are a lot of norms about how these reports are presented in the financial world, but for the common people it's often complex to understand. I'd like to make a design that could help a 20yo waitress understand business better, more than for accountants !

Have you built capability for nested sub-models? It doesn't look like it's structured that way, but I can imagine that being a desired feature as a model grows in complexity.

At the moment there's only one level of submodel, but yes this can definitely be an interesting addition. Wanted to keep things simple at first. Maybe this could be an advanced feature.

I find it interesting that the site-generated revenue almost covers what they pay in advertising to others. I wonder if that's an intentional strategy.

frankly that's a weak part of the model, I have no idea how much they spend on marketing

Ah. Thanks for the clarification.

Employees + Salaries seems a bit higher than what I would expect. Does tinder employ 500 people? Last I heard it was around 200.

You can see the source for that 500 estimate below in the model (or if you click on the variable name). 200 was in October 2016 and they planned to double in the next 12-18 months.

Suggestion: be able to specify ranges for variable estimates, and give ranges for the answers to calculations.

Good one to think about, thanks

So RapGenius for Financial Statements?

Spot on, a big inspiration for me ! I really like RapGenius and even looked at some of their UI while I was developing this.

Either this or "GitHub for business plans" (but many features missing for that, to collaborate, fork, go back to a moment in time...)

I like it very much, but you should add some caching as the submodels are loading very slowly.

thanks will do, and just temporarily upgraded to a better dyno on Heroku... I was on their small Hobby things, mostly good for side projects and it crashed!

Any chance of Tinder IPO someday? TNDR


MTCH gives exposure to Tinder???

MTCH owns tinder. MTCH's parent company is IAC

Yea I just looked it up. Picked up some shares in MTCH, should have looked into this earlier.

This is really interesting. Thank you!

Thanks :)

Where are you getting these models?

I made it, you can make one too !

Now do Bumble.

This should be a Show HN, as it is showcasing your (interesting) service, no?

Well to be honest you might be right. I didn't think it would get so many upvotes, just wanted to throw an example out there to see the reaction and make sure the site doesn't crash. It's good content though, I did some genuine research on Tinder for that example, I'm not just pointing to the homepage :)

We've added “Show HN” to the title. Please let us know if you'd rather not!

it's fine, can't back down now anyway :)

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact