top of page

MVP Software Development: Why Getting What You Asked For Isn't The Same As Getting What You Needed - Part 1

3 days ago
6 min read

The bottom line upfront: MVP software development, or building a Minimum Lovable Product (MLP), takes more than a good brief. Software can match the brief exactly and still miss what you actually needed. The fix isn’t to build more; it's getting an early version in front of real users before you commit additional budget.


We sat down with Faith Rees, Founder of SixPivot, to talk about a common experience for first-time founders commissioning custom software. Her message: a team can deliver exactly what was asked for and still leave the founder with the wrong product. Here’s what she shared and what she advises founders to do before building more.


Faith Rees is standing beside a whiteboard and listening to team members during a SixPivot Summit

Why does software that matches the brief still miss the mark?

A founder said recently, after her product had launched: “I got what I asked for, but not what I needed.”


This story is common among non-technical, first-time founders. The founders we typically work with know their business and subject matter deeply, but what they can’t know before launch is how other people will use what they’re building. When you build from your own experience alone, it’s easy to assume what suits you will suit everyone, because you built it to solve the very problem you needed solved. Put the product in other people’s hands, and they typically engage with it very differently than you planned.


What should founders do before making big bets on software development?

AI has made the first steps far more accessible and, frankly, cheap. Here’s where we tell founders to start.


Start building out what you want in AI. Gone are the days of extensive design briefs, Miro boards, and slide decks. Whether you choose Claude Design or Lovable, start building what you want. AI vibe coding tools now let you spend a couple of months building a rough version and learning what you need. Work out what an MVP or MLP needs to look like for you.

 

If you’d rather not use AI, that’s fine, but you still need to spend the time thinking through and mapping out what your product should look like. Otherwise, you’ll burn money and time doing it with a software development company. To be frank, if you don’t have the time to do this, you shouldn’t be building anything yet. The same goes for design and overall look and feel. Software developers aren’t designers; that’s a different skill set. Collect examples of sites or software you like to save cycles on look and feel.

 

If you’re unsure where to start, talk to a software consultancy. Some will spend some time with you upfront for free, and others will charge. A good consultancy will spend an hour understanding what you are trying to achieve, where you are at in the process and what help you need.  They’ll also tell you whether you can build it yourself or whether you’ll need a software development partner at some point.


Faith Rees of SixPivot, who advises founders on MVP software development, smiling as she talks with senior developer, Gary Jackson, at the 2026 conference

What will a good developer ask you?

Founders are often so keen, and so confident they know what they want and how it should be done, that they just want to get going.


A word of caution: a good software developer will ask you a thousand questions. It will drive you insane, and you won’t have all the answers. That’s ok, but the fewer answers you have, the longer the process takes. They’re asking to make sure you’ve thought through the workflows before they spend your time and money building things that haven’t been thought through. This will happen with or without the prototyping mentioned previously.


We generally see two types of founders. The first is overly detailed, has mapped everything out and won’t listen to experience on how to build something better, especially an MVP. The second just wants it built and expects us to figure it out.  The best outcome sits between the two.


Why does choosing the right software partner matter?

Sometimes we will tell you what you don’t want to hear, and that’s our job.

For example, many founders come to me and say, “I want an app”. We ask why. In some cases, your product should absolutely be an app. But we build a lot of B2C software, and for the most part, products aren’t, and should not be, an app. An app may sit alongside the product, but it’s often not the core. It’s a fight we often lose. The app gets built, and the founder discovers their target market or product isn’t suited to an app experience, and they’ve burnt their money.


When should you put your product in front of real users?

Early, and in phases:


  1. Ask people what they think. Gauge early feedback. Stop worrying so much that someone will steal your idea. As the saying goes, 'ideas are cheap, execution is everything', and that's where the value is.

  2. Build an early prototype, then show it. With AI tools, you can do this with a click. If you can't build it yourself, you could hire a software development company to spend a day ot two building something for you. Then put it in front of people.

  3. Include people who will be critical, not just those you know will love it. Accept that the first version will never be perfect or exactly what you imagined. Explain that it's a test and feedback pilot, not the final product, and people will usually be forgiving. They generally love helping and feeling like they've helped you create something.


If you can’t show it to someone at this stage, you’ll burn a lot of time and money before you feel you can.  I can almost guarantee you’ll still end up changing it and spending more time and money. Fail fast and early.


Faith Rees is speaking into a microphone to seated attendees at the SixPivot conference, sharing advice on MVP software development

Why is polishing in private so risky?

A product built for another year before anyone sees it might be loved or rejected. The opportunity may pass, or a competitor could beat you to the plate. If you wait for perfect, your costs will just keep climbing before anyone has used what you're selling.

 

We’ve recently been working with a founder who has a day job and is building a product to solve a problem in their existing business, which is the type of founder we often work with. They did the first part of prototyping well, including early target market surveys on whether the product would be useful, but not how it should work.


We’ve now spent three months, including around six weeks of paid development. The MVP is done and meets all their initial criteria, and more. But they refuse to put it in front of anyone because they want changes, want more features. We’ve provided the quote and recommended they pause further development until they put what they have in people’s hands, so the next phase delivers what users need. That’s an ongoing argument.


A few years ago, we spent 12 months building something for a B2B business (that had attempted and failed to build it before. Throughout the process, we encouraged them to bring in users and potential customers from outside their business. They refused: ”We are customer 0; we know what is needed.” Three months after launch, the MD said to me: “You told me to put this in front of potential customers early, and we thought we knew everything. Every customer we have put it in front of uses a different workflow than us. You can say I told you so now”.


We still work with this client today, but the changes they needed to make in the early release days could have been avoided, and we could have delivered within the original budget.


How is AI changing MVP software development?

AI makes the early part cheaper and more accessible. A clickable, working prototype is now days of work, not months, so more ideas get tested before real money is committed. You can also show potential investors something tangible rather than an idea.


But AI hasn’t changed the most difficult part. Great software still takes time to build well, and we’re now being asked to review, integrate and secure projects built entirely with AI tools. AI has made it easy to get something working. The challenge now is ensuring it’s the right thing.


Planning a build but not sure what to test first? Talk to the SixPivot team.


Frequently Asked Questions


Should I build an AI prototype before engaging a software consultancy?

Often, yes. A prototype helps you work out what you need and gives a consultancy something concrete to react to. Pair it with expert advice so you’re asking the right questions.


How do I know when to stop building and get feedback?

When you have a prototype people can use. If you’re still adding features without anyone outside your team having tried it, it’s time to stop and test.


What happens when an AI-built prototype needs to become a real product?

Purely AI-built software usually needs review for security, structure and maintainability before real users can depend on it.


Could you use third-party products?

Depending on what you're building and where the real IP sits, the answer varies. Not everything needs to be custom-built, or at least not at first.


Will I need an ongoing developer, and what can I do myself?

This is a big question, and you need to consider it upfront. Software is not built once and done. It is important to understand what you can do yourself, what you’ll need a developer for, and whether certain things should be built into your product to help you do more without one. Many people, founders and software companies alike, don’t have this conversation early enough.

bg1.webp

SIXPIVOT BLOG

OUR INSIGHTS

smooth_6.jpg

Got a project for us?

1800 6 PIVOT

SixPivot Summit 2023-150.jpg

© 2023 All Rights Reserved by SixPivot Pty Ltd. 

ABN 59 606 416 693

​

Website Design OLYA BLACK

bottom of page