F2F #52: Your MVP has too many features

I’ve been building products for years. As a consultant, as a developer, and sometimes as the idiot who agreed to build a monster "MVP" for a friend or a relative.

Here’s the problem: people don’t understand what MVP means. As strange as it sounds, in 2025 most people still don't understand that the M stands for "Minimum".

In fact, in 2021, I wrote a tweetstorm about this and every bit of it still rings true (although I am not too keen on the Lean Startup take, tbh).

It’s Minimum Viable Product. And if you're subscribed to this newsletter, I know that you know this, but if you're not part of the solution, you're part of the problem: help me clear this confusion when you meet someone who asks for too many features for their MVP, as an advisor, an investor, an employee or a friend.

Friends don't let friends build mastodontic MVPs.

Minimum doesn’t mean "minimum price". It means "minimum functionality". But clients hear "cheap" and immediately start asking for champagne on a beer budget. Our job is to set the record straight. We're in this together.

Because an MVP shouldn't cost $100k. And it can’t have 15 features, the same reason why you can't have a great chef cooking an open buffet in a low cost hostel: it just doesn't make sense. The cost of feature is cost up by the pound. Put in too many of them into the scale and it'll sink to the deepest pits of the ocean.

Just because you want to invest twenty grand in an MVP, that doesn't buy you a bottomless list of features. In fact, most companies selling fixed bid MVPs come with hidden charges, extras here and there and really basic functionality that will get you only this far, if you're lucky. That initial 2-5k will turn into 15 or 20 easily over the months. I know it because that's when their customers come to us to fix that mess and see if we can salvage anything out of that rubbish.

If you’re building more than 2 or 3 core functionalities (ideally just one), it’s not an MVP; it’s a bloated product that’ll collapse under its own weight.

And the hidden cost? Maintainability. Every nice-to-have is another way for your system to break, another week doing back and forths with the developers, more stuff to test, another potential security liability, etc etc etc. Skip testing if you want, but every line of code will come back to bite you right in the ass someday.

Stuff that doesn’t belong in your MVP

Here’s the junk I see crammed into MVPs all the time:

• Delete your account, recover password, complex profiles.
  • All of the above have to be done manually over email until it becomes extremely painful. After all, if your users want to delete their profile on your launch day, you're doing a terrible job. And if they have to recover their password on their one, they do have some serious problems on their end.
• Fancy WYSIWYG editors.
  • The first MarsBased logo was designed with Microsoft Paint. Take that!
• Reporting systems, contact forms, social logins, two-factor auth.
  • Most of these things don't make sense until you have a minimum amount of users.
• Invoice generation, multiple payments, multi-language, multi-currency.
  • Your startup has only one user, Joseph, and it's your imaginary friend. He can pay you with imaginary coins, which require no integrations!
• Bespoke design for every screen size.
  • A lot of successful companies started being iOS or Android only for many many years. If Instagram could do it, so can you.
• File uploads with custom storage.
  • If you want to syphon off your investment to Amazon, that's great, but doesn't sound like the wisest of decisions to me.
• Admin dashboards with complex roles.
  • Your company has got exactly zero employees. Do it all with the superadmin role for the time being.
• Feeds, timelines, likes, comments, reviews.
  • On a more serious note here: these are actually counterproductive if they look empty. They signal that you don't have real users. These have to be implemented when you've got an engaged userbase that's large enough to make these metrics matter.
• Workflows that you could just do manually.
  • Self explanatory.
• Favorites, lists, advanced search, endless filters.
  • No one uses advanced functionality when they're still trying to figure out even the most basic features. Stick to doing one thing and doing it right.
• Notifications, chats, A/B testing.
  • Notifications from new apps are usually clunky and invasive. Chats are hard to make them work as intended (Wallapop's chat is still a catastrophe 11 years after their launch - Airbnb's sucks big time too).
• "Social" anything: following, befriending, sharing.
  • We're moving past the "social" hype. Don't waste time here.

This isn’t an MVP. This is a black hole, and you should avoid black holes lest you be sucked into one.

Stop playing in the wrong league

Want perfect UX? That costs millions. Need flawless geopositioning? Not even Niantic nailed it with Pokémon Go. And don't get me started on Apple Maps!

And yet founders compare themselves to Instagram or Facebook. Wrong game. Those guys have been around for over a decade, with bottomless pockets and armies of engineers.

Play in your own league. Build the smallest thing you can actually test with real people and don't burn friendships asking for hugely unrealistic things.

The only metric that matters

The reason I have written this post is because time and again we keep being asked to develop MVPs at MarsBased. Over 90% of these requests end up being turned down because we know they won't see the light of day. Sure, we will get paid for our work, but the project will die before its release, so it'll turn into another churned client, and that is bad business for us.

I can count many examples off the top of my head where we had to develop overbloated platforms despite our advice, with credit systems, virtual currencies, real-time document editing, custom invoicing systems, smart contracts and all kinds of crazy requests. With a few of them we decided to stop working for them because we felt it was pointless to develop something that was going nowhere fast.

Your startup won’t live or die because your MVP had the right feature set.

It’ll live or die because you got clients.

Everything else is noise.