EN

DEREKSOLUTIONS

Who would you like to talk to?

Choose who you would like to talk to and call them directly.

WhatsApp
Menu

Custom SaaS development: what to include in the first version and what can wait

Published on

Anyone considering custom SaaS development usually arrives with a long wish list: a client portal, billing, reports, a mobile app, integrations, roles, notifications, an AI assistant… Each item makes sense, but cramming all of it into the first version is the most common way for a project to drag on, get more expensive and launch when nobody is waiting for it anymore. In Mallorca and the rest of the Balearic Islands, where so many businesses live by the season, launching late can mean losing a whole year.

This post isn't about whether a custom SaaS is right for you. It's about the step after that decision: how to decide what goes into the first version and what waits.

What a first version is (and isn't)

A first version of a SaaS is not a mock-up or a polished demo. It's a product that real users rely on to solve a real problem, even if it does fewer things than it eventually will. It needs to meet three conditions:

  • It completes one specific job end to end. Not half a process, but a whole one, even if narrow.
  • It's secure and reliable. Your customers' data can't be exposed just because "it's only version one".
  • It lets you learn. It should capture just enough to show what customers use, where they get stuck and what they ask for.

What it isn't: a feature list trimmed at random to fit the budget. Cutting without criteria leaves you with a product that does a bit of everything and nothing well.

Step 1: write the core job in one sentence

Before talking about screens, write one sentence describing the job your SaaS does for the person paying for it. For example: "It lets a maintenance company schedule its technicians' visits and lets the end customer see what was done on each visit."

If it doesn't fit in one sentence, you're probably describing two products. Pick one to start with.

Step 2: map the smallest complete journey

With that sentence in hand, list the steps a user takes to get that result, from signing up to getting value. In the example:

  1. The company creates its account and adds its technicians.
  2. It adds its customers and their sites.
  3. It schedules visits on a calendar.
  4. The technician sees the day's visits on their phone and logs the work done.
  5. The end customer receives a visit summary.

That's the skeleton. Anything outside those steps is a candidate to wait.

Step 3: sort every feature into four buckets

Must-have

The journey breaks without it. In the example: account sign-up, visit calendar, mobile view for technicians, customer summary.

Must-have, but can be manual

It's needed, but a person can handle it at first. Importing customers from a spreadsheet can be done by your team for the first users, and subscription billing can run outside the product for a few months.

Clear improvement

Adds value, but the journey works without it: advanced reports, complex filters, per-company branding, accounting software integration.

Bet

Features you don't know anyone will use: an AI assistant that writes the summaries, a technician marketplace, a separate native app.

Version one should include only the first bucket and, at most, a cheap piece of the third. The second bucket is covered by manual processes. The fourth gets validated before it gets built.

What you can't postpone

Some parts of a SaaS look "technical" and easy to push back, but cost far more when added late:

  • Data separation between customers (multi-tenancy). If each company must only see its own data, that belongs in the design from day one.
  • Basic roles and permissions. At least admin versus user. Fine-grained permissions can wait; the foundation can't.
  • Backups and recovery. Tested, not just switched on.
  • A minimal activity log. Who changed what and when, so you can resolve incidents and customer questions.
  • Languages, if your users need them. In the Balearics it's common to work in Spanish, English and sometimes German or Catalan. Preparing for several languages from the start is simple; retrofitting it isn't.
  • Data protection compliance. Where data is hosted, who can access it and how it's deleted when a customer leaves.

Where AI fits in version one

It's tempting to put AI everywhere, and sometimes it belongs there from day one. The useful question is: is AI part of the core job, or an improvement?

  • If your SaaS exists to classify documents, answer enquiries or extract data, AI is a must-have and goes into version one, with human review and a way to correct its mistakes.
  • If AI only saves a few minutes on a secondary task, such as drafting the visit summary, it's usually better to launch without it, see how users actually write those summaries, and add it later with real data.

You can read more about how we approach these pieces on our AI and automation in Mallorca page.

Tie-breakers when you're unsure

  • Has a potential customer said they wouldn't use the product without it? If that comes from a real conversation, move it up. If it's an assumption, move it down.
  • Could it be done by hand for three months? If yes, wait.
  • Would adding it later force you to rebuild something? If yes, consider including at least the foundation.
  • How much does it delay launch? If it pushes you past a key moment, such as the start of the island season, it probably isn't worth it.

A hypothetical example in Mallorca

Picture a pool maintenance company based in Llucmajor that built an internal tool to organise its technicians and now wants to turn it into a SaaS for other pool companies across Mallorca, Ibiza and Menorca.

Their initial list has fifteen features. After applying the steps above:

  • Version one: multi-company accounts, technicians, customers and pools, a visit calendar, a mobile visit sheet with photos and an emailed summary for the owner, in Spanish and English.
  • Manual at first: customer import and subscription billing.
  • Phase two: per-technician reports, chemical and stock tracking, accounting integration.
  • Validate before building: an AI assistant that suggests tasks based on each pool's history.

The result is a smaller product that can be in the hands of the first companies before peak season, and that grows based on what those companies really use.

How to tell whether version one works

Before launch, decide what you'll watch in the first weeks:

  • How many companies finish sign-up and schedule their first visit.
  • At which step the others drop off.
  • What users ask for most often by email or phone.

With those answers, the phase-two list almost sorts itself.

Thinking about a custom SaaS?

If you have a product idea, or an internal tool you'd like to turn into a SaaS in Mallorca or the Balearics, we can help you define the first version and build it. See how we work on custom software development in Palma de Mallorca, or get in touch at derek@dereksolutions.com or +34 680 283 973.

DEREKSOLUTIONS

Privacy & cookies

Essential preferences

We save your choice in your browser for 180 days so we do not ask again. It is not used to identify you.

Analytics (Google Analytics and Hotjar)

Google Analytics counts visits, page views and where visitors come from (Google Ireland Ltd). Hotjar shows heatmaps and anonymous session recordings, without recording form fields (Hotjar Ltd, Malta, EU). Only enabled if you accept.

Current status: disabled

External links

WhatsApp, Google and project websites have their own policies when you open them.

You can change your choice at any time with the shield button at the bottom left.

WHATSAPP

Who would you like to message?

Choose who you would like to talk to and we will open WhatsApp with your message ready.