The first idea came from a friend.

The problem was simple:

Having a meal plan is one thing.

Turning that plan into the groceries you need — and actually getting them to your door — is another.

So I started building MealFlow.

Meal plan → customized grocery list → groceries delivered to your door.
Adjust ingredients.

Remove things you don't like.

Build the shopping list.

Then send it to a delivery service like Instacart or Amazon for Whole Foods.

But MealFlow taught me something bigger than the app itself.

I wasn't programming it.

I was describing the result I wanted.

And the platform was figuring out much of the how.

It created screens.

It made design choices.

It handled pieces of the database and structure I hadn't even told it to create.

The first version wasn't perfect.

Far from it.

But within roughly 30 minutes, I had something that felt maybe 60% of the way toward an actual app.

That caught my attention.

Because normally I would have been thinking about fields, databases, structure, code and all the mechanics underneath it.

Instead, I had to learn a completely different skill:
​
How do I clearly explain what I want the app to do?
​

But MealFlow also taught me something else:

simple ideas can become complex very quickly.

The more I worked on it, the more moving parts started to appear.

There were ingredient preferences.

Shopping lists.

Delivery connections.

Services like Instacart and Amazon for Whole Foods.

What started as a straightforward idea kept expanding.

And the more I explored it, the more I realized MealFlow was going to take more time to fully work the way I wanted.

Eventually, I hit a wall.

So instead of forcing it, I made a different move.

The second idea went in a completely different direction.

What if you could hear someone speak and think:

Not copy their voice.

Not imitate them.

But study their delivery.

Their pacing.

Their pauses.

Their emphasis.

Their energy.

That became SpeakLike.

You provide a sample of yourself speaking and a sample of a speaker you admire.

SpeakLike compares the two and gives you five specific things you can practice to improve your delivery.

By this point, I was getting better at communicating with the platform.

The first app taught me what not to say.

The second taught me how much more effective I could be when I described the result instead of trying to micromanage how the software should create it.

And somewhere around the 13-hour mark, I realized something.

I was supposed to be building one app.

I already had two.

Neither was completely finished.

And the clock was still running.

At that point, I could have spent the rest of the challenge trying to force one of those apps across the finish line.

Instead, I made another decision.

Simplify.

I wanted to prove that I could get one app working from beginning to end.

That decision led to app number three.

I went back to a problem I already understood.

Funnels.

More specifically:

The original idea for FunnelScan was bigger.

I wanted it to scan a ClickFunnels page automatically and identify technical and conversion problems.

That was too much for the time I had left.

So I pulled it back.

What is the simplest useful version?

Enter a few numbers from your funnel:

Visitors.

Leads.

Bookings.

Attended calls.

Checkouts.

Sales.

FunnelScan compares the stages, identifies where the biggest measurable leak appears to be and estimates what improving that area could potentially be worth.

I also added questions about tracking.

Because a diagnosis isn't worth much if the underlying numbers can't be trusted.

And this time something different happened.

It worked from beginning to end.

I tested it.

I tried different numbers.

I tried breaking it.

I used AI to help me think through test cases and edge cases.

And after more than 20 hours of building, experimenting, getting stuck, changing direction and trying again…

FunnelScan became the app I submitted for the challenge.

Not because I planned it that way.

Quite the opposite.

I kept running into walls.

So I kept learning.

MealFlow showed me how quickly an idea could become something visible.

SpeakLike helped me get better at communicating what I wanted.

FunnelScan taught me the value of stripping an idea down until I could actually complete it.

And somewhere during those 20+ hours, the challenge stopped being about building an app.

It became a lesson in iteration.

Try something.

See what happens.

Learn.

Change it.

Try again.

I have spent a lot of my life believing I needed to understand something before I could start.

Study it.

Learn it.

Think through every possibility.

Then act.

This challenge flipped that around.

And that may have been more valuable than any of the three apps.

For years, programming had been one of those things I never believed was really mine.

I had taken programming courses before.

Let's just say programming and I had a complicated history.

So the idea that I could sit down, speak into a microphone and describe what I wanted an application to do…

and then watch it begin to appear on the screen…

was a little surreal.

The biggest limitation wasn't whether I could write the code.

The limitation became whether I could clearly explain the result I wanted.

That is a completely different problem.

And it made me start wondering:

What else could be built this way?

Because if I could go from being challenged to build one app… to creating three different apps in 24 hours…

what could you build with an idea of your own?

Before this challenge, building an app felt like something that required programmers, developers, databases, code, and a whole lot of time.

That is not what this experience felt like.

With OverSkill, I was able to describe what I wanted the app to do, make changes as I went, test ideas, and keep iterating without writing the code myself.

That does not mean the app magically comes out perfect.

It doesn't.

You still have to think.

You still have to make decisions.

You still have to test it.

And probably most important, you have to learn how to clearly explain what you want the app to do.

But that's a very different challenge than having to learn how to program the entire thing yourself.

And after going through these 24 hours, I can tell you this:

If you've ever had one of those:

ideas…

you may want to try building it yourself.

Describe the idea. Build the first version. Then start improving it.

But maybe you don't want to build it yourself.

Maybe you just have the idea.

That's where this gets even more interesting…

Then:

Maybe there is something you do every week that drives you crazy.

Maybe there is a spreadsheet you are tired of maintaining.

Maybe there is a small problem in your business that shouldn't require five different tools to solve.

Maybe it's something completely ridiculous.

I don't know.

That's the point.

The next app might start the exact same way these three did:

With somebody saying…

"I wish there was an app that…"

Tell me the problem.

It doesn't need to be technical.

You don't need to know how the app would work.

Just tell me:

What do you wish was easier?