Sorted - App to help couples plan their movie night
The problem
Couples often spend a long time deciding what to watch. Each person scrolls through the streaming app, suggests a few options, and the other person rejects most of them because of mood, genre, or language. By the time they agree on something, a good part of the evening has gone, and sometimes they do not agree at all. Sorted was built to help couples finalize a movie quickly.
Before building Sorted, I spoke with about 10 couples I know across India, the US, and Canada, and the struggle to agree on a movie came up consistently.
User persona
The primary users are couples between the ages of 24 and 49, across all geographies, who have access to at least one streaming service.
Sorted also has a single mode for the same users when they want to play alone. In single mode, the user matches against one of four fictional characters, each with their own taste in movies.
Why build Sorted
Streaming services recommend movies based on a single profile, so they do not consider what two people want to watch together. Review sites and curated lists help people find good movies, but they do not help a couple agree on one.
There are several swipe-to-match apps for couples, but most are available only on iOS, and some collect identifiers that can be used to track users across other apps. Sorted is built for Android first and collects no personal data. It also includes a single mode with characters for people watching alone, shows only movies available on the streaming services in the user’s region, and reaches beyond the most popular titles so users find movies they have not already heard of.
There is also potential to expand Sorted well beyond the original use case. Over time, it could grow into a community of people with shared tastes, a movie review site, a dating site built around common interests, or eventually a social media platform.
The business case is to build a user base first and introduce a subscription model later. Monetization is not a goal for the initial release, and any future revenue model has to work within the privacy-first approach of the app.
MVP feature set
-
Both partners join a session using a shareable code and swipe through movies on their own devices. When both swipe right on the same movie, it is a match.
-
Users can filter movies by genre, language, time period, and IMDb rating, and the deck only includes movies available on the streaming services in their region.
-
Each movie card shows the name, a short description, the language, and the IMDb rating.
-
Single mode lets one person use the same swipe and match experience with a fictional character.
-
There is no login. Session data stays on the server only while the session is active and is deleted once a match happens. Matched movies are saved on the user’s device, with up to 10 movies in the watchlist.
While completing the Google Play data safety form, I found the app was still creating a persistent device ID, which went against the privacy design. I replaced it with session IDs that are deleted after use before the release. I also chose not to include any crash-reporting tool and rely on Android vitals instead, so no additional data leaves the user’s device.
I chose Android for the first release mainly to get to users faster. Android runs on about 69% of mobile devices worldwide, compared with about 31% for iOS, according to StatCounter’s June 2026 data. The United States is the main exception, where iOS has the larger share, which is one reason an iOS release is on the roadmap. As monetization is not the goal initially, reach is more important than revenue per user.
Roadmap
In the near term:
-
Add TV shows, books, and podcasts
-
Add a group mode for more than two people
-
Add an optional login so users can sync across devices
-
Release on iOS
In the long term, the plan is to build a community around Sorted, which could eventually lead to a social media platform.
Launch
Sorted went through internal testing and then fourteen days of closed testing on Google Play.
Feedback from testers led to several changes:
-
The deck only showed popular movies, so a broad filter returned the same blockbusters users already knew. I changed how the deck is built so it also includes less popular movies that still meet a quality bar.
-
Apple TV+ was missing from the streaming services, and it was added back.
-
There was no way to mark a movie as already watched, so I added a watched list that holds up to 50 movies.
-
The four characters in single mode were picking movies that overlapped too much, so their filters were tuned to make each one distinct.
Sorted has since been granted production access on Google Play, and is now live on Google Play Store.
Metrics
Sorted does not collect analytics, so counting matches needs a privacy-preserving approach, such as a simple total kept on the server with no link to any user or device. Until that is in place, I track installs, retention, and ratings through the Play Console.
Since Sorted was recently launched on the Play Store, my top priority for the first six months is to grow the audience. To track that, I will use these KPIs, all available in the Google Play Console:
-
Install Base: The number of active devices (used in the last 30 days) that have the app installed.
-
This is the primary metric, and it will decide whether I invest more in the project. Once I start charging, I need at least 300 paid users to cover the costs of running the app, assuming net revenue of $1 per user per month. Free-to-paid conversion is typically around 5% even in the best case, so I would need an install base of at least 6,000 to break even.
-
I will track this week over week, along with supporting metrics such as new installs, device information and country-level data.
-
For now, I am targeting English-speaking users in all countries and regions, although the language filter supports any language. Localisation adds cost that is not justified until I see stronger demand.
-
-
Reach: The number of unique devices that were shown the app on Google Play.
-
This tells me whether I need to do more to market the app. The higher the reach, the more likely someone is to visit the store listing and install the app.
-
I have a baseline and will track whether reach grows over time. To improve it, I will run campaigns on YouTube and Instagram, and keep improving the store listing.
-
Related metrics include store listing visitors, device installs and device acquisitions
-
-
Quality: The average Google Play rating
-
This matters less while I focus on growth, but I will watch reviews and feedback to catch anything seriously concerning. Reviews also feed the roadmap.
-
Supporting metrics include user-perceived crash rates and crash reports.
-
Public documentation: github.com/kppadukudru/Sorted-Public





