Part 2: Lean Startup Methodology - MVP Testing and Rapid Market Validation Strategies

Today we talk about failure.

And why we love it.

Let's learn how best to fail.

Failure is key to learning.

And the key to failure is to make each failure small.

There are two ways to do this. The wrong way and the right way.

The wrong way to fail is how most businesses operate.

They spend time and money:

• preparing the perfect plan
• building the perfect team
• getting an kick-ass logo designed All before actually, you know, opening the doors to customers.

What generally happens is a flop.

At no point did the business bother asking what the customer actually wanted. Or, better, asking them to make a purchase.

Instead the business followed the Hollywood Blockbuster model of "build it, do a big launch and hope for the best." And if it doesn't immediately work then throw cash at advertising.

This is a dangerous, flashy way to launch a business. And often leads to a glorious wreckage.

One huge failure. Game over.

It's also how most, if not all, businesses launch.

The better way

The right way is to avoid one big failure by having

as many small failures as possible. As quickly as possible. Fail fast, fail often.

Sounds weird but hear me out.

We only learn through failures and mistakes. We cannot, by definition, learn from a complete success. There's no learning moment there.

We

need to fail in order to learn.

But: we need each of those failures to be survivable.

The regular (bad!) model allows all our failure to coalesce in one big launch moment. It's all of nothing.

The better model is to make lots of plays, make a lot of mistakes and small failures, learn from each failure and improve.

'Learn' startup

This model is called "lean startup", a term popularised by Eric Ries in his book of the same name. I'm calling it "leaRn startup" to hammer home the point.

The basic idea is:

• start with a hypothesis
• come up with a test
• validate by launching In the realm of business for example this means constantly launching new products, services and content to see what customers

actually resonate with.

We don't use market research. We don't make "educated" guesses. We don't just hope for the best.

We build and launch. And see if the customer responds.

And the only response we care about is them opening their wallet and paying.

Nothing else matters for our tests.

Lean Startup Machine

I actually won a lean startup competition when I lived in New York. This is where I learned the power of this methodology.

It was a 2 day hackathon where we came up with a product and got the basics built. We then were judged by a group of successful entrepreneurs and a winner announced.

There were some seriously impressive startups being created that weekend. Really cool ideas. Multi-million sized projects. Amazing teams.

But we won.

Not because our idea was better. Not because the projected market was larger. Not because out team was more impressive.

We won because we got members of the public to give us cash.

We got total strangers to pay us for this product that didn't exist yet, with the promise that we'd get it to them if/when it's ready.

Anyone can say "that's a great sounding product." People say nice things (white lies!) all the time. Especially if you are talking to people who like you.

But what people

say doesn't matter. It's what they pay for that is the only real proof.

Minimum viability product

In practical terms what does this mean?

It means you need to launch your product/service immediately.

Not in a few months. Now.

And you use the success of this launch to gauge the next steps.

If people bite your hand off for it you know you are onto something.

If there's no interest then the idea needs tweaking. In lean startup it's called a "pivot".

We can do fast launches by building something called a "minimum viability product" or MVP.

Basically this is something that captures the

value of your business in its most essential form.

AirBnb is a famous example here. It began literally as one air mattress on the floor of an apartment in SF. People rented it instead of a hotel during conference season.

Groupon's MVP was hitting up an email list of 500 people working out of an office building selling them pizza. Then calling up the local pizza joint and bulk buying 2-for-1 pizzas and pocketing the difference.

The purpose of the MVP is purely to prove that there is a market demand for what your business intends to sell.

You do not wait until you have an office, team, fancy logo and letterhead to sell. You sell immediately and see if anyone gives a damn. If no then this is a tiny low cost failure. Instead of wasting months and thousands of dollars you have your answer immediately.

So let's build an MVP for your business:

Act as a lean startup consultant
My business idea is [detail your business]
My ideal customer is [customer details]
The primary problem I solve for my customers is [problem]
Suggest 5 potential MVPs I could design and deploy
The MVP is to confirm market demand. 
Each MVP idea should be fast (`<1` day) to deploy, cost less than \$100 and have a success/failure condition.

Fill in as many details about your business as possible. As always, more specificity is best.

ChatGPT will generate some MVP examples. Here's one of five:

Importantly there are Success/Failure conditions. Use these as jumping off points to set your own. Remember, the purpose here to validate where there is a market or not.

The MVP should be super simple to deploy. Not something that'll take a month to prepare.

And remember - you can always kill it before having to fulfil!

In the example above it's a workshop that we get sign ups for. You'd set a date, create a zoom invite link and publish the event to your channels. If you don't have channels in place you'd run an advert.

If you get the 10 sign ups (success condition) great - run the workshop.

If you get less than 5 signups (failure condition) - kill the workshop.

The goal here is not running a workshop.

The goal here is to gauge customer demand!

Premium Prompt - Sequential testing

One MVP does not a business plan make.

Instead you need a set of experiments, all giving you more and more information about customer demand.

Here's a prompt to set this up for you.

Act as a lean startup consultant
My business idea is [detail your business]
My ideal customer is [customer details]
The primary problem I solve for my customers is [problem]
Suggest 5 potential MVPs I could design and deploy
The MVP is to confirm market demand. 
Each MVP idea should be fast (`<1` day) to deploy, cost less than \$100 and have a success/failure condition.
Each MVP should be sequential, following on from the last, from smallest to largest commitment to create and deploy. Each should only be deployed if the previous success condition is met. 
Provide the 5 MVPs and a basic flowchart decision tree.

This will

sequence the MVPs like so:

These are the first 2 (of 5).

As well as a decision tree:

Deploy each test one by one, each time growing in confidence that the business idea is strong.

You'll get

massive learnings about your customers by following this process of incremental building.

Pulling it together

We want to fail.

That's how we learn.

We just failure fast and fail often. Rather than saving it all up for one cataclysm.

This takes a lot of mental rewiring. We'll cover that in the next Part.

A reminder of what we'll be covering :

• Part 1: The upside of downfalls
• Part 2: Fail fast, fail often
• Part 3: Mental tweaks
• Part 4: Antifragility
• Part 5: It's dangerous to go alone