The fastest way to waste money on software is to build everything at once. The best first versions do one job, for one kind of person, really well. Here's how to decide what makes the cut.
Start with one person and one job
Picture the single person who feels the problem most. Now name the one job they need done. Not five jobs. One. That's your first version.
For a building surveyor, the job is "turn site photos and notes into a finished report". Not "manage my whole practice".
Write everything down, then cut
List every feature you can think of. Then sort each one into three piles:
- Must have: without it, the one job can't be done.
- Should have: makes it nicer, but people could manage without it for a while.
- Later: everything else.
Your first version is the "must have" pile, plus maybe one or two from "should have". It will feel too small. That's normal, and it's a good sign.
What usually waits
- Several types of user with different permissions.
- Detailed settings and customisation.
- Connections to lots of other systems.
- Both iPhone and Android apps on day one.
- Big admin dashboards and fancy charts.
What you shouldn't skip
- Security and privacy basics: safe sign-in, data stored properly, and a clear privacy policy.
- Backups, so nobody loses their work.
- An easy start, so a new user gets value in the first few minutes without a manual.
- A way to hear from users, because they'll tell you what to build next.
- A way to pay, if it's a product you're selling. Even a simple one tells you if people value it.
An example
For the surveyor's report tool, a sensible first version could be:
- Take photos room by room on site.
- Rate each defect using the surveyor's own guide.
- Produce the finished report as a PDF to check and send.
Later: a portal for clients, team accounts for larger firms, and links to booking systems. All useful, none needed to prove the idea.
A second example, sorted into piles
An accountant spends the start of every tax season chasing clients for documents: bank statements, receipts, P60s. It's done by email, from memory, and things slip.
Example · A document chaser for accountants
- Must have: a list of clients, a checklist of documents for each one, automatic email reminders until they're sent, and an upload page clients can use without creating an account.
- Should have: text message reminders, and a view of who's furthest behind.
- Later: links to practice management software, team accounts for bigger firms, a client phone app, and branding for each firm.
The first version is the top row. It could be in an accountant's hands in a few weeks, and one tax season would show whether it saves real time.
How long should a first version take?
For most of the products we see, 4 to 12 weeks. If the plan runs much longer than that, it's usually a sign that too much has crept in. Go back to the piles and move something to "later".
How to know it worked
- Real people use it for real work, week after week.
- They'd be annoyed if you took it away.
- They ask for more, or someone pays for it.
If those happen, you've earned the right to build the next piece. If they don't, you've learned that cheaply, and you can change course.
Need help deciding?
Deciding what goes into the first version is the first thing we work on together. Bring the long list, and we'll help you find the small version that proves the idea. See how our development process works, or how we kept Invoice Drop focused.