Case Study: How 3 startups built their customer discovery practice

Posted on September 11, 2023.

When serial entrepreneur Steven Cohn set out to launch his fourth startup, Winware, he spent the bulk of his time where few founders do at the outset of their new ventures: talking to customers. Instead of hacking together his first prototype or hiring his first engineer to start writing code, Cohn dove deep into his target audience. Winware’s goal is to use a machine learning algorithm to optimize SaaS products to drive better product adoption, increased trial conversion and improved revenue retention. For Cohn, now a veteran of three exits, there were still many unanswered questions. For example, he still didn’t have a full understanding of the challenges his target audience faced using data, analytics and marketing automation tools to drive product-led growth (PLG) conversions and customer retention. To get those answers he started speaking to folks he believed were his target audience. 

Before building any product Cohn, on his own, spoke to more than 300 different companies. He met with a variety of roles within each of these companies. Each of these conversations helped him narrow down his target audience so he could start focusing on his ideal customer profile (ICP). His daily routine consisted of anywhere between 2 and 4 interviews that were recorded and later shared with the team he brought on. The rest of his time was spent finding folks to speak with. He tried multiple tactics to recruit interview participants. Initially Cohn did cold outreach on LinkedIn. With a relatively low success rate, he then began reaching out to his personal network as well as his investor’s networks. Finally, he began experimenting with content marketing to drive inbound leads – not for sales (at least not yet) – for learning. Writing compelling content on LinkedIn began to attract the audience he was looking for. As they engaged with his work, he reached out to this warm pool of leads for further insight into their thoughts. Building a startup in a crowded market required Cohn to build a deeper understanding of the problem Winware is solving and why and how existing tools do not solve these problems.

You will be wrong

No matter how confident you are in the product you’re building or the problem you’re solving, at some point you are going to be wrong. You’ll choose the wrong target audience. You’ll prioritize the wrong feature. You’ll pivot into a space that doesn’t have the growth potential you anticipated. Perhaps you’re addressing a problem with decreasing urgency or maybe your value proposition is just too complicated for your audience to understand. The sooner you can learn where you’re right and where you’re not, the sooner you can redirect efforts towards a more productive direction. 

Just because a sector is crowded doesn’t mean there’s no room for new tools and services to exist. Without understanding where the gaps are though, it becomes increasingly difficult to determine how to position your offering. Launching a product will tell you, quickly, if you’ve found product market fit. It will also bankrupt your startup equally as fast. Building a practice and cadence of regular customer conversations is a much cheaper way to learn the same thing with a tiny fraction of the investment. But it will require patience. 

Understanding where you fit in a crowded market

First-time founder Matt O’Connell saw an opportunity in the customer insights and productivity sector. He identified two sides of a typical company’s workflow. First, he identified tools that help teams execute more effectively – Asana, Jira, ADO, Trello, Linear, etc. These tools live close to delivery teams and support the sequencing and tracking of work. Then there are more strategic tools for improving alignment and creating focus. These tools typically take on things like OKRs, performance management software, and analytics. But what are teams doing to connect those two worlds? And how do we ensure the focus on the customer doesn’t get lost in the shuffle? This is where O’Connell positioned Vistaly. His goal is to insert Vistaly into existing product development tool stacks to maintain the customer focus in everything the teams do. 

In an average week O’Connell has 15-20 customer conversations. Most of these are 1:1 or 2:1 calls with customers. They are often interviews and are designed to be informal to keep the interviewee comfortable and sharing freely. Before the conversation, O’Connell and the Vistaly team identify what they’d like to learn. This allows them to keep the conversations focused and no longer than they have to be. This is particularly helpful if a conversation starts to go off-track or stops yielding useful insights.Time is precious – cutting interviews short when they stop providing value is in everyone’s best interest. Having a clear learning goal defined upfront helps make that decision easier. 

Getting creative in finding people to speak with

Since most of Vistaly’s prospective customers are already using Slack, they’ve begun reaching out to customers to asynchronous conversations in Slack. This simple technique breaks the ice with the target user and makes it easier to get the conversation started on their terms. Once the team builds a rapport with the customer they can ask to schedule a live call with them. This dramatically increases the conversion rates for getting interviews scheduled. Several times, a simple Slack outreach about a specific learning goal has resulted in an immediate response from the customer, “Do you have 10 minutes to chat about this now? I just experienced that with my team.”

Another way O’Connell and the team learn from their customers is by observing where they are experiencing friction with the product. They often will reach out for help to resolve an issue. This is where customer support typically steps in. The Vistaly team likes to capitalize on these exchanges to learn more about the root cause of these issues. They don’t always have success turning reactionary customer success calls into discovery calls (some people want a problem resolved and then move on), but when they are successful the learning payoff is substantial. 

The Vistaly team is small. Making the time for customer interviews isn’t easy and can often lead to other, suboptimal compromises. Their target audience are often busy professionals with jammed calendars. To find time with them requires flexibility and accommodation. As a result, O’Connell will often have interviews scattered throughout the day. This can be painful for someone with an engineering background. As O’Connell said, “I’ve learned the value of “flow state” – being completely absorbed and focused on something, often to the point of losing track of time. Some of my best work happens in this state, but that’s hard to get into when conversations break up your day, and you have to constantly be aware of the time.” One thing he’s found helpful is reserving a few large chunks dedicated to customer discovery work throughout the week. While this solves a problem for him, it doesn’t always align with his customers’ availability – which always comes first. 

Implementing a practice of continuous conversation

Putting together a practice of regular customer conversation is crucial to the success of your startup. As the founders of Winware and Vistaly shared, the depth of insight they’ve collected over the hundreds of interviews they’ve done has allowed them to make better decisions, reduce burn rate on initiatives that won’t yield results and focus their teams on the most probable path for success. However, the practice takes time and a toll on the founders’ calendar. How can you put together a sustainable program within your company to develop this muscle and build in that continuous customer conversation? Let’s take a look. 

In terms of investment, the main currency here is time. The cost of doing customer interviews is rarely financial. Instead, it requires a deliberate decision, by the founder, to allocate some percentage of time to learning. This has an explicit tradeoff – time spent learning is time not spent delivering. The insight gained from the interviewing process will provide better direction on what to build and how to implement and design it. In an early-stage startup, everyone in the company should make time in their schedule to speak with customers. This starts with the founder(s) and applies to early engineering, product and design hires. Once the team starts to grow, not everyone will speak to customers directly but everyone should have access to the insight and the time to consume it. Cohn, the founder of Winware, encourages his engineers to put in a JIRA story called “watch customer interview” so that they can build time into their sprints to view recorded customer interviews from earlier in the week. Depending on your sector you may need to provide some compensation to your interviewees. This often comes down to a gift card to Amazon or Starbucks, usually in the $50-150 range depending on the job title of the person you’re interviewing. Aside from running the actual interviews, you’ll need to account for the time spent recruiting and scheduling them. This may take up to 4 hours per week of your time. If that starts to become untenable consider outsourcing this to an external recruiter. It will cost more but will provide ideal candidates for you to speak with. 

What does a customer interviewing process look like?

At the bear minimum, your interviewing process should result in 3-5 customer conversations on a weekly basis. Here’s what that looks like on a daily basis:

  • Monday: decide what the most important thing you need to learn this week is and identify who you’d like to speak with
  • Tuesday: begin recruiting efforts to find at least 3 customers to speak with and reach out with an interview request along with any incentives you are offering for participation
  • Wednesday: Follow up and confirm scheduling for the interviews. Start putting together the list of questions you’d like to ask during the sessions.
  • Thursday: Conduct interviews. Ideally, you’ll be able to schedule them all in one day. There will undoubtedly be exceptions to this. 
  • Friday: Meet with the team to review the findings from your interviews and discuss how it will impact existing plans. 

Rinse and repeat this process on a weekly basis. Make sure you record every interview and have it automatically transcribed. You can then use AI tools to look for patterns in the transcriptions that you may have missed over dozens of conversations. 

Making the conversations valuable

Interviewing customers isn’t hard but it does require practice. Think of it as a muscle. It already knows what to do but the more you go through the right motions, the stronger the muscle gets. Interviewing is no different. There are endless resources on how to run a great customer interview. The most important thing to get out of your customer interviews are stories. You want your participants to tell you how they go about the tasks that you’re trying to support and improve. You want to understand where their work is easy and where it’s hard. You want to get a better sense of what causes them pain in doing a good job and what they’re doing today to alleviate that pain. How well is the competition meeting their objectives? Where are they failing to do so?

To support this type of insight there are (at least) two powerful questions you can ask in every customer interviews:

  1. Tell me about the last time you [did the thing we are trying to help you with]. This prompt asks your participant to describe to you how they went about shopping for shoes online, inputting some data, running a query or applying for a mortgage. Whatever problem it is you’re solving for, ask them to tell you about the last time they went through the same process. Listen for them to tell you that one part of it was hard or easy and dig in to find out “what made that easy for you?” or “why was that hard to accomplish?”
  2. If you had a magic wand and could make this experience perfect, what would that look like? Your goal with this prompt is to get a sense of what they imagine your service should provide. This is inevitably where they’ll start to list off a wishlist of features. This is a good time for the second half of this prompt. Every time you hear them say something like, “It would be so much better if the product allowed me to do [something]” you respond with, “If you had that feature, what would it let you do?” What you’re trying to learn here is not what features they’d like to have but what goal they’re trying to achieve. Ultimately, you and the team will decide how to help them achieve that goal but understanding what the goal is for them is the key. 

Cohn, founder of Winware, adds that he likes to make an explicit request to sign the customer up as a beta user at the end of each interview. This and requesting an introduction to others who may be interested in the product are like a “real word NPS survey,” Cohn added. 

Make learning the path of least resistance

As a founder you play a pivotal role in creating the culture for your company. If having a deep understanding of the customer is part of the culture you want to create you have to make customer conversations, and learning, the path of least resistance. Build in incentives for your team to conduct customer interviews. Enable them to take time away from their other tasks to do so, as Cohn did at Winware. When new ideas come up for consideration and discussion, ask what customer problem these ideas solve. Mandate that new ideas need validation prior to implementation.

Most importantly, provide your team with unlimited budget and support for talking to customers. Whatever they need, the answer should always be yes. This is, after all, a very low-cost activity. The last thing that should stop you from getting to know your customers is budget. In our second book, Sense & Respond, there is a case study about a German streaming media company called Maxdome. CEO Marvin Lange knew he was up against several 800-lb gorillas in the streaming media space as well as a cultural sensitivity to subscription services. To figure out how to overcome these enormous challenges he provided his teams with unlimited research budgets. Anytime they wanted to meet with customers the answer was yes. In addition, Lange mandated that his executive team would work the local christmas markets in Germany, trying to sell the service face to face to people on the street. Each of these interactions is an opportunity to listen and learn why people do and do not buy. Make this type of curiosity a part of your culture from the beginning. In addition, when your team sees you and the executive team doing it as well, they’ll understand how important it is. 

Celebrate course correction

All of this new insight will conflict with the plans you already have in place. As mentioned above, you will inevitably be wrong about the decisions you’ve already made. This is cause for celebration rather than an error to be avoided. Every time you and the team learn something new from talking to your customers, share it broadly with the rest of the company. Be clear about the original plan, what you learned along the way and how you will be changing course. Publicly celebrate de-prioritization of ideas that don’t meet customer needs or conflict with the insight gained through customer conversation. When your teams see that it’s ok to change course based on evidence, they’ll be more likely to do it. This will save significant resources chasing ideas that aren’t going to work just because someone decided to do it without any customer validation. Continuous improvement requires continuous learning. The agility of your organization depends on your teams feeling comfortable implementing the insights they’re learning from talking to their users. 

Democratize the learning process

The final piece of this puzzle is to ensure everyone on your team, regardless of role or title, understands that talking to customers is part of their job. It is their responsibility to understand what customers want and what’s getting in their way. The examples used in this article are all about founders doing the legwork. That’s a great start because it shows how important it is to the core team. As the team grows everyone else must build this into their weekly practice. This isn’t work for just researchers or designers or product managers and it certainly should not be outsourced to an external agency. This is key, foundational insight into whether or not you’re solving a real problem for real customers in a meaningful way. Everyone must care about it and participate in the validation process. 

Not just the right problem, the right users

With Adobe’s acquisition of Figma, Penpot, the first Open Source design and prototyping platform for product teams, saw a spike in sign-ups. Co-founder, Pablo Ruiz-Múzquíz was cautiously optimistic. While acquisition numbers were going up, it was crucial to understand whether these were the users Penpot was actually seeking for its product. Like the other leaders in this article, Muzquiz began speaking to these new folks. Initially, that conversation was filtered through a series of short surveys. As new customers were validated they were put into a beta program which provided access to the tool in exchange for feedback. Often, that feedback was delivered in interview form. As the Penpot team ramped up its product development efforts to meet this new demand they maintained a regular cadence of 30 customer interviews a week. They had to ensure that they were meeting the needs of these new users, focusing directly on their core target audience while not straying away from their core vision and mission. 

Every new feature Penpot puts in their roadmap comes with validated customer insight to back it up. As these new ideas come to life they’re put back in front of those same customers to see if they’ve met the need. As Ruiz-Múzquíz puts it, “An interview gives you the chance to pitch your vision more informally and get an unfiltered reaction to it. Sometimes that reaction is quite bland [or]… you get a sequence of “wow!”’s and “aha!”’s”. To be clear, the team at Penpot isn’t asking customers “what they want” from the product. Instead, they are listening for unmet needs and then pitching solutions their customers may have never imagined. It’s a mix of vision and customer-driven insight. 

There is no substitute for customer conversations

The closest thing we have to a magic wand to drive the success of our startups is talking to customers. It’s cheap, fast and effective. It doesn’t require extensive training and is an activity anyone in the company can and should do. Immersing the organization in developing customer empathy ensures the work they do always benefits the customer. It plants the seeds of customer-centricity early in the company’s culture and ensures that, as you grow, this will always be a part of how you evolve your business. As with anything in a startup, if it matters to the founder, it will matter to everyone else. Leading this effort by modeling the behavior ensures you’ll see it in your teams for years to come.

Leave a Reply

Your email address will not be published. Required fields are marked *