We open your competitors' iOS apps for you and record every feature they add or remove,even if the App Store stays silent
Each change comes with a date and the screenshot that proves it.
We record
An agent walks the competitor app, 20 minutes, 400 to 900 screens
Machines read
Computer vision and language models name what is on every screen
You get the diff
We compare passes. Every change gets a date and a frame
You name the competitors, we watch them for you
Every feature side by side, and what sits behind each paywall
| Feature | |||
|---|---|---|---|
| Water tracker | |||
| Calorie counter | |||
| Barcode scanner | |||
| Food photo recognition | |||
| Macronutrient breakdown | |||
| Onboarding quiz | |||
| Ad banner | |||
| Voice food input | |||
| Day streak | |||
| Spin-the-wheel discount | |||
| Anchor price on the paywall |
Every product feature your competitors add, gate, or remove
Each change carries a date and the frame behind it. Removals are the part nobody can rebuild later.
Click a row to see the frame and how you compare
Every cell opens the screenshot, with your app beside the competitors
Day streak
How the implementations differ
Your frame stands here, next to the competitors.

Milestones: the day streak next to badges, Rookie to Immortal.

On the profile, with a flame, next to steps and calories.

Full-screen after you log food, the week laid out below.
Follow one feature through the flow, frame by frame
Every frame that names the feature, in the order it was shot
#198 · items list
#199 · sofas
#200 · back to the plan
#201 · sofas
#202 · sofa colors
#203 · more sofas
#204 · items list
#205 · mirrors
#206 · table on the plan
#210 · outdoor
#211 · garden sheds
#212 · more outdoor
#233 · garage in 3D
#234 · garage
#235 · garage tables
#236 · items list
#237 · lighting
#238 · more lampsEveryone has a water tracker, the difference starts inside
None of it shows in the store listing

Goal 2 L, eight glasses to tap. Water from food on its own line, 862 mL.

Quick-add sizes: glass 250, bottle 500, large 750 mL. No daily goal at all.

Setup first: a 2,550 mL goal and a cup size you tune before logging a drop.
The patterns nobody has time to spot, every week
A model reads every screen and writes what repeats across competitors
Everyone has the barcode scanner, only one marks it paidAll show it on a free pass. One flags a Pro badge, the rest hide the paywall until you tap.
The leader reminds you about the subscription 10× more, but locks no more featuresA paywall marker sits on 21% of its screens against 2% at a rival, and both lock about as many.
Twenty-eight features run at every appOut of 131 tracked. That is the floor of the niche, not an edge. They split on the other 103.
Two apps are near-twins, 63 shared featuresA 52% overlap. A second shoot shows who copied whom, and how fast.
One app builds five times the retention features of anotherTen in streaks and rewards against two. The store shows none of it.
Product managers, growth managers, founders
Read what changed over one coffee on Monday. We read 8,000+ screens per competitor a quarter.
- Product ManagerWhich to copy, and which someone already shipped and removed.
- Growth ManagerWhat moved behind the paywall, and what the leader tests on the price screen
- FoundersOne place that remembers your competitors, so nobody restarts the research
Your competitors, tracked beyond features
Every App Store change on the apps you pick: releases, description, title, languages, publisher, charts. iOS for now.
| App | Free chart | Paid chart | Watching |
|---|---|---|---|
| #12up 2 | #9no change 0 | recording | |
| #3up 6 | #4up 3 | recording | |
| #24down 3 | #12down 1 | public data | |
| #8up 1 | #6up 2 | public data | |
| #16up 4 | #1up 3 | recording | |
| #33up 9 | #21up 6 | recording | |
| #18no change 0 | #10down 2 | public data |
What the store shows while we record inside
Checked every night from day one
| Date | App | Detail | What changed |
|---|---|---|---|
| Aug 14 | re-shoot queued | Release 12.2.0, changelog mentions a new reports screen | |
| Aug 13 | 184 MB → 231 MB | App size jumped | |
| Aug 11 | #4 → #1 | Top grossing, Germany | |
| Aug 10 | rewrite | Description: "calorie diary" became "AI nutrition coach" | |
| Aug 08 | +2 languages | Added Polish and Turkish | |
| Aug 06 | bug-fix streak | Three releases in a row with an empty changelog | |
| Aug 04 | 16.0 → 17.0 | Minimum iOS raised |
Ask about your competitors in the chat you already use
Connect your own AI and ask in plain words. The frames come with the answer.
How do these three build their object catalog? Show me the frames.
All three have one, built differently. magicplan lists categories with a count: 17 doors, 57 plumbing items. CamPlan opens photo tiles, 435 sofas and 2,873 storage units. Home Planner AI lists objects with brand and size, down to a 66 by 66 cm table.
Works in Claude, ChatGPT, Cursor and Codex. One read-only link, and it comes with your account.
Screenshots
"Show me the screen this feature was seen on."
The matrix in words
"What do three of my five competitors have that I do not?"
Changes and removals
"What changed since June, and what did anyone drop?"
Before you build
"How did three rivals build the streak, screen by screen?"
Publisher claims
"They announced AI coaching. Did you film it?"
On a schedule
"Every Monday at 9: what changed at my competitors last week?"
We build market memory in three layers
The store tells you a version shipped. A pass tells you what is inside it. History tells you what they tried and dropped, and that layer can't be rebuilt later.
Each layer only exists because the one under it runs every day
Questions we help product teams answer
A few examples, swipe through
What appeared, and what disappeared?
Did someone already ship it and take it out?
We are about to ship a feature. What have others already tried?
Who copied the streak freeze, and how fast?
Who rewrote their description, raised the price, or went quiet on releases?
What did competitors have six months ago, and what now?
We can open five competitors ourselves
Once a quarter, without history. We hold your competitors continuously and remember what was there in March.
Is this just screenshots?
No. An agent records a pass, then a model reads every screen, folds names into one dictionary and compares passes against each other. You see the result and the frame behind it.
Why trust what the model names
It must quote the words on the screen before it names a feature, and nothing counts on a single frame.
When do we see the first result
Store data on day one, the first breakdown in two weeks, change events from the second recording.
Can we query this from our own AI
Yes. Your account comes with a read-only MCP link for Claude, ChatGPT, Cursor or Codex, over the same data the dashboard shows, screenshots included. Your model answers, so we never see the question, only which tool it called.
How often do you record
On release. We read the store every night, and a new version puts the app back in the queue. Between releases the screens do not move, so a second pass would only cost you time.
Which platforms do you cover
The iOS App Store, in the US, German and British storefronts. Android is not covered yet, and we would rather say so than show you a thin answer.
Who decides which competitors you film
You name them. We watch the rest of the niche from the store at the same time, so the day you add a name, its releases, prices and chart history are already there.
What if you miss something
Then we say we have not seen it. A feature we never filmed reads as not measured, never as missing, and nothing counts on a single frame. When we do claim a removal, we have covered that part of the app on later passes and it was gone.
Can we see our own app in the matrix
Yes. Your app sits in the same grid as the competitors, and every comparison starts from your row.
See what changed at your competitors and decide with proof
Tell us your niche and get the first breakdown in two weeks