Cutting Version One of Your App: Building a Real MVP Feature List
Short answer: Write one sentence describing the job your app does for one type of user, then sort every feature on your wish list into must, should, could and won't for now. A feature is a must only if the core job fails without it. For many of the remaining items, a manual workaround such as a spreadsheet, a phone call or an email does the job for the first few months. Set two or three measurable success signals for version one, and put everything else into a dated version-two backlog so nothing is lost and nobody feels ignored.
Why version one keeps growing
A wish list grows because every feature has a story behind it. One customer asked for it, a competitor has it, or it would be "quick to add". Each addition looks small, but together they double the build time and push back the moment you learn whether people actually want the app. The aim of a minimum viable product is to learn that as early and cheaply as possible.
Public sector teams use the same idea when they test ideas in an alpha phase before building a full service; the GOV.UK service manual on alpha is a clear description of that approach.
Step 1: Write the job-to-be-done statement
Use this format:
When [situation], [type of user] wants to [action], so they can [outcome].
A worked example for a dog grooming salon planning its own app:
When their dog needs a groom, regular customers want to book the next free slot with their usual groomer in under a minute, so they can stop phoning the salon during working hours.
Every feature is now judged against that sentence. If it does not help regular customers book quickly, it is not part of version one.
Step 2: Sort the wish list with MoSCoW
MoSCoW stands for Must have, Should have, Could have and Won't have (this time). Here is the salon's wish list sorted:
| Feature | Category | Reason |
|---|---|---|
| See available slots for chosen groomer | Must | Core of the job |
| Book and receive confirmation | Must | Core of the job |
| Save dog profile (name, breed, notes) | Must | Needed to book correctly |
| Login with email | Must | Ties bookings to a customer |
| Reminder notification the day before | Should | Reduces no-shows, but SMS from the booking system can cover it |
| Cancel or move a booking in the app | Should | Customers can phone for now |
| Pay a deposit in the app | Could | Valuable, but adds payment provider setup |
| Loyalty stamps | Could | Nice, but not needed to prove booking demand |
| Shop for shampoo and treats | Won't | A different business case |
| Photo gallery of grooms | Won't | Instagram already does this |
| Chat with groomer | Won't | Phone and WhatsApp exist |
Be strict with the Must column. A useful test is to ask: "If this were missing, would a customer give up on booking?" If the honest answer is no, it moves down.
Step 3: Replace features with manual workarounds
Many "should" and "could" features can be covered by people, not code, while you learn:
- Cancellations: a "Need to change? Call us" button opens the phone dialler.
- Deposits: the salon takes payment at the counter, as now, and watches no-show rates.
- Reminders: the existing booking calendar sends an email or SMS.
- Reporting: bookings export to a spreadsheet once a week instead of an in-app dashboard.
- Loyalty: staff stamp a paper card or note it in the customer record.
If a workaround becomes painful because demand is high, that is excellent news, and it tells you exactly which feature to build next.
Step 4: Set success metrics before launch
Decide what success looks like for version one, with numbers you can actually measure. For the salon, for example:
- At least 40% of repeat bookings come through the app within three months.
- Phone calls for booking fall noticeably compared with the same period before launch.
- Fewer than one in ten app bookings needs a staff correction.
These are example targets. Pick yours based on your current numbers. Without them, every opinion about whether the app "works" is equally valid, and you will be tempted to add features instead of learning.
Step 5: Park the rest in a version-two backlog
A backlog is not a graveyard. Keep it in a shared document with:
- The feature
- Who asked for it and why
- The evidence that would move it up, for example "if more than 15% of customers call to reschedule"
- A rough size: small, medium, large
Review it once a month after launch, alongside your metrics. Features that real users keep asking for move into the next release; ideas nobody mentions again quietly fall away.
A note on app store review
Cutting too far has its own risk. Apple's App Store Review Guidelines include a minimum functionality rule (section 4.2), and apps that are little more than a wrapped website or offer very limited value may be rejected. A booking app with real accounts, live availability and confirmations is a genuine app; a single page linking to your website is not. Keep version one small, but make sure it does its one job properly.
FAQ
How many features should an MVP have?
There is no fixed number. It should have exactly what the core job needs to work end to end, and nothing more.
What if a big customer insists on a feature?
Ask what problem it solves for them and whether a manual workaround would do for now. If the feature is truly critical to that customer and they matter commercially, it may belong in version one.
Is MoSCoW the only method?
No. Some teams score features by value and effort. MoSCoW is simply quick and easy to explain to everyone involved.
When should version two start?
Once you have a few weeks of real usage data against your success metrics, not on a fixed date.
If you are trying to trim a long wish list into a first release you can actually ship, share it with us and we will help you shape version one.