Lean Startup Adventure, day 5: Designing a Problem Interview

Yann Gensollen
Yann GensollenJanuary 27, 2016

In the previous post in this series, François analyzed our Business Model to look specifically for risks. The riskiest part of the canvas, for us, is the customer segment and acquisition.

Do the customers we identified actually meet the Problem we defined?

The Problem Interview

In order to find it out, the cheapest and fastest way to test our assumptions is to meet potential customers, and see how they react to the Problem.

This Problem interview is described in Ash Maurya's Running Lean book. It is the very first step in validating important risks in the big amount of hypotheses we make while we fill the Lean Canvas. This exercise aims to challenge the couple Problem-Customer segment.

It's a learning tool: based on what future customers say, we will be able to refine our Business Model. So it's important to clear our minds of the Solution at that point, and to only focus on what we are solving, who is the competition, and who has the pain.

Changed priorities ahead

What We Want To Validate

A Problem Interview is supposed to verify falsifiable hypotheses. Based on our Risk analysis, we formulate the following hypotheses:

  • The “Data management” task is identified as part of someone’s job description in the Company
  • General purpose and specialized data administration tools have many shortcomings (low usability, too many tools, not connected)
  • Using these tools leads to suboptimal data management
  • The loss / delay caused by bad admin backend can be estimated

Problem Interview Script

Based on Ash Maurya's Problem Interview structure, we write a script that we'll use during our interviews. Here is our first draft:

1. Welcome - 2 minutes

Thanks for receiving us. We’re here to get your feedback about a web product we’re building.

At marmelab, just like you probably, we handle a lot of data for our activities. We’ve developed several backend administration tools to create, edit, and retrieve this data. Today, we are thinking about the opportunity to transform these tools into a web product, open to new customers. We’d like to share the idea with you, and grab your feedback. We’re at very early stage ; we need to challenge our assumptions with your expertise. Your remarks and critics on our vision will be very valuable for us. They will help us to validate that we’re on the right path, and to align the future product to actual customer requirements.

Welcome to Las Vegas Sign

2. Collect Demographics (Test Customer Segment) - 2 minutes

This is where we try to understand who this customer is - is he/she in the segment we target? an early adopter?

  • Your company tagline says "XXX". Can you tell us more about your main activity?
  • We've done some research on the size of your company. Can you confirm our figures about your revenue, and number of employees?
  • Who are your customers?
  • Do you use any of the following SaaS service?
    • Trello,
    • KissMetrics
    • Basecamp
    • WordPress Hosted
    • Other SaaS service
  • Do you use any of the following software?
    • Drupal
    • Magento
    • Excel
    • SharePoint
    • Other domain-specific software
  • Which data do you collect, store, and edit in your activity ? For instance, an e-commerce company stores customers, products, and commands; a media stores articles, images, advertisements

3. Tell a Story - 2 minutes

This is where we show our focus on customers, open the door for sharing their concerns, and show our understanding of the problem. We choose to tell two different stories, according to the segment the customer we're interviewing belongs to.

Startup Story

We’ve worked with several startups who focus on gathering customer feedback early, in order to validate their initial assumptions. For instance, one startup we’ve worked for, initially launched with a simple landing page, a registration workflow backed by static pages, and analytics everywhere.

Data administration features were deferred to a later stage. In the meantime, the startup team managed their data by hand, using a concierge approach to better understand the concerns of early adopters. The concierge handled subscriptions, customer requests, and cloud pricing data (their core business) using primitive tools like a database browser (phpMyAdmin). In fact, most of the real data administration used to take place offline, using Excel, and only copied to phpMyAdmin once the data is ready.

Very soon, they were struggling with the data import/export between user-friendly but offline tools (Excel) and hard-to-use but online tools (like phpMyAdmin).

SME Story

Many of our customers develop not just one, but several applications. For instance, the system used by one of our customers, an online service for booking restaurants, is divided by domain: one application for user management, another for restaurant configuration, yet another one for reservations, etc. Each application comes with its own backend administration tool.

The problem is that many day-to-day business tasks handled by the people at that company involve more than one tool. For instance, adding a new restaurant requires to add a new user first. Sometimes, they need to re-enter the same piece of data twice in two different tools. Sometimes, they need to switch between two interfaces to check data consistency.

The more a company develops, the more applications they use, the more this problem becomes a source of time loss (to duplicate data, to check for errors and fix them), and a source of expense (to train employees on several tools, to develop new backend tools).


4. Problem Ranking - 4 minutes

This is where we see if they really struggle with this problem.

How many backend administration tools do you use? For which kind of task?

How would you qualify these tools? Are you satisfied with their usability?

Could you rank the 3 following problems by order of importance?

  • You struggle with several administration tools for your everyday job
  • Your administration tools have a poor usability and slow you down
  • You spend a lot of money to develop new administration tools

5. Explore Customer's Worldview - 15 minutes

This is where we understand how they solve this problem today. The script is very loose, we just need a few sentences to relaunch de discussion if we ever reach a dead end.

Who is responsible for solving these problems today in your company ? How do you solve them?

Do you know of any alternative solution/service that would fit your need?

6. Wrapping Up - 2 minutes


We believe that startups need a backend administration tool that would be as powerful and user-friendly as Excel, but connected to the live data, and using a web interface. A backend administration targeting business owners, not developers. A backend administration tool that can be configured without any IT intervention.


We believe that SMEs need a unique backend administration tool for any given backend user. Collaborators should not have to learn many tools to do their job. Instead, they should have all the necessary data to do their job, available under a single interface tailored for efficacy. This tool should be able to connect to several third-party data providers, and hide the technical complexity. It should be simple, user-friendly, handle authentication and authorization seamlessly, and allow not only to read data, but to edit it.

For both

We have developed such a tool for our internal use. We often refer to it as our “Connected Excel”, or “Multi-Backend Administration”. Some of our customers already use it as part of their everyday job, and love it.

Would you be interested about updates on our project, or even a demo of the tool once we have something to show?

As a side note, we would like to interview more people like you, who could be faced with the very same problem. Could you refer us to some of your contacts for a quick interview?

Also, we intend to document our product exploration on our blog. Would you agree to see your responses published as part of a blog post? How about your identity?

Welcome to Las Vegas Sign

Note: In retrospect, this first draft shows how we are influenced by our vision of the solution. Most notably, the wrapping up sounds like a pitch, which is not fair since we announce that we have nothing to sell at that point. After a few interviews, we will follow this script more loosely, and focus more on the actual customer pain points.

Next Steps

With the Problem Interview Script ready, we can now go and meet potential customers, and start learning. We're using our network to find people in the right segment. The next posts will summarize a few of the problem interviews we gave, and the insights we got from them.


Thumbnail picture: Spiral stairs in Saint Istvan Basilika in Budapest, by Jan Fidler

Did you like this article? Share it!