Website Survey

Our MVP Testing Went Wrong. Here’s How I’d Do It Differently.

Time to read: 5 minutes

This is inspired by a recent conversation that I had on Quora.

I was suggesting that the OP should do some research before generating “business ideas”. He disagreed:

I think that if it’s easier to launch an idea than to research it, it’s a more useful approach to just test it.

You see, a year ago, I would have said the exact same thing. And I would have been terribly wrong!

Here’s how MVP tests go wrong

Let me share some first-hand experience with you:
A year ago me and the team had an idea about a product that would help marketers. We decided that we would do a super-lean product market test—we would just create a landing page, describe what we’re offering and ask people to subscribe.

We spent quite some time thinking about what our marketer customers would want so that we define the features of our MVP.

Let me repeat this: We were thinking about what our future customers would want. We were effectively imagining what problems they had.

We put a lot of effort into that landing page. We did A/B tests on it. We posted on forums and asked our potential customers what they thought about it.

They gave us vague answers.

We got few subscribers.

Was it because they didn’t want the product?

No. (as we later found out)

It was because we weren’t speaking their language.

Lost in translation

If you’re going to validate the interest towards your MVP with a landing page (which is the cheapest method timewise), you need to make damn sure that any potential customer who clicks on your link is going to relate to the text on your landing page.

You might have the most awesome idea of all times, but if you can’t communicate it to your potential customers, if you can’t show explain to them why this is something that they need, you are most likely not going to get any sign ups.

“We can’t market it, so let’s just build it”

Back to the story. Feeling frustrated with our landing page, we decided to put some last efforts in our idea. And so we followed the traditional “Lean Startup” methodology.

Here’s what the lean process suggests:

Image originally published on
Image originally published on

Good in theory, but in practice it turned out to be:

We built [BUILD] the MVP product itself, because we thought that maybe we sucked at writing marketing texts (which, it turns out, also wasn’t the problem). We sold it to a few customers. They actually wanted it and liked it!

However, talking to customers [MEASURE] we discovered that they were missing some basic (to them) features [LEARN]. And we had included perks that no-one needed and used. And so, we started fixing the product.

We guessed about 20% of our potential customers’ needs.

We actually sucked at guessing. And, probably, so do you.

That was the culprit the whole time.

Perhaps we should have started with the MVP itself, then?

A counter-question:

If you can’t get people to be interested in your idea, if you can’t make an offer now, how the heck are you going to make that offer later?

Think about it. If you can get your potential customers to sign up before you code the MVP, then you’ll be able to get them to purchase it after. That’s why promotion comes before coding.

But, “building a landing page” is also not your first step towards a software product business.

So, where do you start?

Here’s how I’d launch an MVP now (process viewed from 10k feet about the ground):

First things first: Who are your future customers?

Being passionate about building things we forget the most important part—someone must want to purchase them.

In marketing terms: You need to define your potential customers.

But by “define” I don’t just mean: “Yeah, I’ll help marketers”.

Take the time to describe them and research them:

  • Who are they?
  • What distinguishes them from other people?
  • What do they do for a living?
  • What things do they like?
  • Where do they hang out?
  • Why do you like them?
  • Why do you want to work with them/sell them something?
  • etc.

Second: Don’t make hypotheses. Find out what they want

Don’t imagine. Don’t pick your brain for ideas, like we did with our product. Don’t ask them “What do you need?”, because they can’t say. Or if they say this won’t be the right answer.

Instead, research your future customers and find out what problems they have.

Once you have found what problems they have, you can base your MVP hypotheses on these problems. Your hypotheses will still be unproven, but they will be based on data, rather than your imagination.

Third: Learn your audience’s language


Like I told you, we had a tough time explaining to marketers how we could help them. And it was exactly because we didn’t know what problems they had, their language and what they valued. We didn’t know what made them tick. Because we weren’t marketers.

If we could speak their language, we could have found out a lot more about them before building the actual MVP. We would’ve gotten a lot more subscribers on that landing page.

So, spend some time analyzing your target customers. Go to their specialized forums, note what words they use and how they describe their problems.

Fourth: Make a valuable offer

Now that you can speak their language, you can explain to them what you’re offering.

That’s where your “coming soon” page comes into play. Not before that, but just now—when you already know whom are you talking to and what pains you are killing for them.

The landing page will contain your promise—a brief description of what your product does and how it helps your target audience. In their language.

Ask your potential customers to sign up and be notified when you launch the product.

Now you can start coding your MVP.

Bonus tip: Don’t let your email list go cold.

While you’re building the product itself, make sure that you keep talking to your future customers. For example, email them blog posts with solutions to their problems, share your progress with your MVP and ask them what they think.

This way, when you launch your product, you will have a group of people who feel engaged—they have been with you all the way.

In summary

You might have a valuable product idea, but if you can’t communicate it to the people who will be using the product it is worth nothing!

Therefore, you won’t receive a response from them.

Don’t assume what your potential customers want—research them. Use the data from your research to shape your product idea. Then create a landing page and explain your product offer.

About the author Gergana Dimova

I use my non-magical persuasion methods to help small business owners, digital agencies, entrepreneurs and consultants get more leads and sales. You can learn more about working with me here.

Leave a Comment