01Research
Is anyone actually asking for this?
Store demand, search signals and competitive density are measured before a project is approved.
Stops here when demand is thin, or strong apps already serve it well.
Studio · Louisville & Seoul
Small apps, built on evidence.
8bit is the mobile app studio of VRAVIO Inc. A pipeline reads Google Play every day, scores what it finds, and forecasts the outcome — before a single line of code is written.
One job per app. No feature bloat.
Store scan
Charts · keyword suggestions · ranking movements · review histograms
Candidate scoring
| Candidate | Demand | Supply gap | Score |
|---|---|---|---|
| PDF tools | Wide | 86 | |
| QR scanner | Medium | 74 | |
| Unit convert | Medium | 68 | |
| Step counter | Narrow | 61 |
Every score carries its reasoning, so a low one can be argued with.
Outcome forecast
- Optimistic · P90 21%
- Base · P50 14%
- Pessimistic · P10 9%
The pessimistic case is the one that has to clear.
Illustrative — not live data
What the pipeline reads, every day
327
Apps tracked
35,151
Chart snapshots
15
Categories watched
Public Google Play signal only — charts, suggestions and store metrics. No private data, and never another developer's assets.
01Research
Demand is a live number, not a slide from last quarter.
An automated batch reads Google Play every day — charts, search suggestions, ranking movements, competitor install brackets, ratings and review histograms. Every record carries a reliability grade, and low-grade data is never allowed to calibrate a forecast.
Daily collection
Pulled on a schedule, not when somebody remembers to look.
Reliability grading
Every record is tagged. Weak data is barred from calibration.
Teardown
Leading apps taken apart — what works, and what reviews complain about over and over.
The gap is the spec
The need nobody serves becomes the specification.
Daily collection
- 06:00 Top charts · free & grossing A
- 06:04 Search suggestions · seed keywords A
- 06:11 Ranking movement · 7-day delta A
- 06:19 Install brackets · category peers B
- 06:26 Ratings & review histograms A
Grade B and below never calibrates a forecast.
02Simulate
The pessimistic case is the one that has to clear.
Before development starts, a cohort model projects installs, retention and unit economics, then runs a Monte Carlo across the uncertain inputs. We read the pessimistic case first. Projects that only work when everything goes right are dropped.
Cohort model
Installs and retention curves projected month by month.
Monte Carlo
Uncertain inputs sampled thousands of times, not fixed to one guess.
Unit economics
Revenue per user against the cost of building and keeping it alive.
Kill criteria
Written before the run, so the result cannot be negotiated afterwards.
Forecast run
- P10 – P90 band Confidence
- P50 median 14%
- Pessimistic · P10 clears the bar 9%
Day-30 retention. The bar is set before the run, not after it.
03Build & measure
Narrow on purpose — then measured against the model.
The first version is deliberately small: original design, our own assets, and at least one thing the category does not already do. After launch, retention, stability and reviews are compared against the forecast, and the difference goes back into the model.
Scope is a decision
Not an accident, and not everything that fits in the sprint.
Original work
Our own name, icon, artwork and store listing. Every time.
Post-launch review
Retention, crash rate and reviews on a schedule — fix or retire.
It compounds
Every launch makes the next forecast less of a guess.
Release checklist
- Our own name, icon, artwork and store listing
- One measurable improvement on the category
- Privacy disclosure written before submission
- Support contact reachable from inside the app
- Post-launch review scheduled, not promised
Every item is a blocker. None of them is a nice-to-have.
The gate
Four stages. Each one can stop the project.
Nothing advances because somebody liked it. These are the points where a project dies, and what kills it.
02Simulate
Does it pay for itself when things go badly?
Installs, retention and unit economics are modelled across pessimistic, base and optimistic cases.
Stops here when the pessimistic case does not pay for the work.
03Build
Can it stay one app, doing one job?
A deliberately narrow first version, with at least one measurable improvement on the category.
Stops here when the scope cannot stay narrow, or the one clear improvement disappears.
04Measure
Did reality agree with the model?
Retention, stability and feedback are compared against the forecast on a schedule.
Stops here when the real numbers miss the model and there is no honest way to close the gap.
Standards
Built to be legitimate
Publishing many apps only works if every one of them is legitimate. These are the rules we hold ourselves to — the same four, on every app, with no exceptions for a good week.
Original work
Our apps use our own name, icon, artwork and store listing. We study categories and user needs — never another developer's assets, branding or copy.
Privacy first
We collect the minimum needed to run and improve the app. We do not sell personal data, and we say exactly what is collected in our privacy policy.
Honest monetization
Ads are clearly labelled and kept out of critical interactions. No misleading buttons, no forced clicks, no surprise subscriptions.
Real support
Every app links to a person who answers. Support requests get a reply within two business days.
Apps
What we publish
Our first apps are in preparation.
We are finishing research and the first builds. As soon as an app is live on Google Play it appears on this page automatically, with its store link, category and support contact.
Want to hear when the first one ships? Tell us.Questions
Before you ask
How long does a reply take?
Within two business days. Louisville and Seoul cover different time zones, so a message sent overnight is usually read the same day.
I was charged and want a refund.
Purchases in our apps are processed by Google Play, not by us. Refunds are requested through Google Play — open play.google.com/store/account, find the order, and choose Report a problem. If Google declines and you believe the charge was our error, send us a message and we will look into it.
How do I delete my data?
See the data deletion page. It explains exactly what we hold, what gets deleted, and how long it takes.
Can I report a security issue?
Yes, and we would be grateful. Use the contact form, choose "Report a problem", and please give us a reasonable window to fix the issue before disclosing it publicly.
Contact
Working on something, or need help with one of our apps?
We are a small team and we read everything ourselves. A person replies within two business days.