Every time I see a post about dating apps and how unfair they are or how they are scams just looking to make maximum money, I think of my own dating app idea. I have many ways to avoid side projects and one of these is to look at the tech Id use to implement an idea, rather than writing actual code.

On that note, if somebody is reading this I should say that I dont intend on implementing this idea for a bunch of reasons. This is all just an exercise in thinking through how I may implement it. Mainly so I can get it out of my head and carry on procrastinating over ideas I may one day implement (the irony). There are multiple reasons I wouldnt create a new dating site, of which I will possibly get into later.

TLDR: Dating apps bad. We set a few rules for our free (pretend) app on a £50pm budget. These rules are summarised at the very bottom of this post. In the next part we look at implementation.

So my major idea behind the dating site is that its free. This removes an incentive to game the system for cash. For some reason, as Ive thought of this idea I have always had a monthly budget of £50 in my head. Im not sure on why, I think its because at some point I decided that spending £50 a month overall on side projects is a decent investment for a hobby. Plus, with such a small monthly cost for this dating site, its sustainable without requiring payment from users.

The other idea behind the app would be to take donations. That way the £50 monthly budget can aim at supporting say 10,000 users and then any donations would allow expansion. Its all theory anyway, but lets say for this theory that we have a budget of £50 a month and have to support 10,000 users to get this off the ground.

On being just a web app, Ive always thought this may well have an impact on security. Although I may be wrong as most of the older dating sites started as just websites and Im pretty sure Tinder is available on the web. Anyway, I think while there is a larger security risk with just a web app, it would reduce maintenance costs as I have no mobile app skills and dont really intend on getting any. Plus I would mainly just aim this app at mobile/tablet sizes anyway, reducing CSS/design fatigue. I am aware how not having an app reduces the appeal for users though.

So location is probably one of the main parts of a dating app. How do you find people near you without their location. My solution to this part would be that a user puts in the areas around them that they want to search in. Which I know sounds silly. I cant say I have a great solution for location yet at this point. But equally its something that could be added later if needed.

Making this just a web app and with no calls to any location APIs would make it cheaper. But also the lack of location makes it more difficult. I should say aswell that if I was releasing this app it would start in a small area, say a part of London. Then expand from there.

Further to the 10,000 user aim, the idea behind this design would be that after reaching 10,000 users, in theory enough donations would help support expansion. Being realistic most apps dont manage 1000 users anyway, but lets just aim for 10,000 anyway in all of this.

Okay so with the small monthly budget, the app would need limitations. For example with that budget we cant afford to just allow a user to continously swipe through profiles.

I believe the first issue is solved by only accepting donations. A common problem with apps too is dormant accounts. I obviously want to protect my ego, but I have seen multiple different accounts for the same person. I believe this is when a user has left the app and then created a new account and the app has kept both accounts. Again theres an incentive for the app to keep dormant accounts - it displays they have loads of users on the app and they can be used to make a user think they arent getting matches, to persuade them to pay money.

We dont have any incentive to keep dormant accounts, because it costs money. So we infact have a small incentive to get people off the app. And this is the main 'issue' with the dating app business, your customers are looking for a partner and when they find one, you no longer need their product. So they need to keep people out or relationships, while allowing them to date. And thats one of the main reasons we are where we are with dating apps.

Continuing this, being able to endlessly look through your options is not healthy for anybody. Having what looks like unlimited options will have a user constantly comparing and makes someone think they're always missing out on something better. So our app will only be usable midday to midnight and only show three potential matches per day.

Again, this is cost. If a user doesnt use the app, then those three people do not get refreshed until they accept or reject them. This saves on our side as we dont have to just keep pushing more and more data to a user. Meanwhile, we encourage the user to engage with the app to actually get anything out of it. This is the thing with some of these constraints, they also setup an alternative experience for the user who hopefully in some way benefits. I also think removing any unused accounts after a period of time is useful for both us and users, more on that later.

Having a midday to midnight limit also allows us a huge window to do calculations and prepare data, again helping with costs. There is this app called Thursday that apparently only works on Thursdays, so you can only message matches etc on a Thursday. Its adding these restrictions that can make something more appealing from the norm.

To address ghosting, I have a couple of methods. Firstly, you can only have 5 matches at any one time. This means to get more matches, you have to explicitly unmatch somebody and possibly provide some feedback. Im not definite on numbers, maybe 5 is too many, but its a good start.

My other idea for ghosting, which again happens to be cost effective for us, is that we dont provide live messaging. While its nice to have a back and forth going, how often are you messaging back and forth quickly? If you are, you have probably moved to a proper messaging app. Instead I think we should provide a slow inbox, like email.

I genuinely believe that sometimes my messages on dating apps havent been sent properly. I know, I know, Ive probably just been ghosted. But how can I be sure. Isn't it easier on me to just believe the app has betrayed me? In fact, I should mention here that I was once messaging somebody on the Hinge app, I didnt get a reply and two days later they messaged to say 'What happened?'. We got onto Whatsapp and through screenshots saw that they had sent a message but I had not received it on the dating app. So it does happen, in fact this could lead into the fact that these dating apps want more money. Somebody thats getting ghosted might possibly pay as they are seeing less success.

Now, again Ill talk about the benefits for us on this in a moment, but first lets pretend this is beneficial for the user. You shouldnt need loads of photos for a profile - enough to show what you look like. It should really be about your personality. A friend of mine mentioned once how dating apps dont actually show off what a person is like. And its true, what does my favourite TV show or food really say about me. If anything its a good, fast way to filter somebody out. In an ideal world we wouldnt need photos, but we are all shallow lets not pretend. But anyway three should be enough to show what you look like. I think here aswell we could ask for most recent and then check metadata on the photos. Yeah people can manipulate that but still, its a simple way to ensure the three photos are relevant.

My idea was that any uploaded photos would get resized so that they maximise 15Kb. Again cant remember where this number came from. With words, they are cheap to store. So lets say 3 x 15Kb for photos and then 5Kb for text. It would be less than that for text, nobody will use the whole thing. But that gives a total of 50Kb budget per user. I thought this was good enough, thinking about it some apps probably dont have much more, as they're mainly photos and one sentence answers to random questions (Hinge). To continue this maths, 10_000 (our target users) x 50kb = 500Mb, thats easy to store anywhere right?

Another restriction I had too was account deletion. As Ive said some of the benefits with this app mean interaction (the 3 a day limit) but I think another issue with dating apps is stale accounts. Its not been outright proven, but there is strong incentive for dating apps to keep dormant accounts as mentioned. Again this is maybe me defending my ego, but when you swipe and get maybe 1 match in 100 swipes, your not telling me all of those accounts are active and being shown your profile to match back.

So thereforce I think removing any accounts that have no use after 14 days is the best way to go. This ensures we have engaged users and you know your match is going to get reviewed otherwise the user is not engaging with the app. Again too, I have to mention it, it suits us as it costs less. We can delete those accounts as they are bad for our users.

As mentioned before, there is no intention to use the person's exact location. Mainly because this isnt cheap. An example, Google now charges $5 per 1000 calls to their geolocation API. So that would take the whole budget just for the 10,000 target signups for the app.

So whats the alternative? The only thing I can think of here is when you create your profile you select where you are from and then select the areas you would like to search. This also restricts who can sign up and allows the app to grow (right?). We would then just have to maintain a list of specific places. There are issues with this though. For example, we cant just have every major city, as they dont cover every part of the country. So this list of places would need thinking about.

One possibility here may be to seperate a map into smaller sections and then allow the user to select their section and the sections they want to look at. Overall though, this isnt very user friendly. Major cities, with London broken down further maybe, would probably just be the best approach. But yes this underserves a large part of the country. My only counter to that is that this app isnt real anyway and it would probably just get started in a place like London and grow from there anyway.

Next time, I will look at the data requirements of the app and maybe some of the code that would need writing. Then I will go on to look at a more detailed implementation details (Server, Cloudflare, AWS).

Dating App theory part 1 - Rules