The biggest customer development mistake that most people make is: not doing customer development.
Even when an interview feels awkward or you’re not sure you’ve gotten any useful responses, you’re still learning.
And also: there are some common mistakes that can cause you unnecessary frustration:
- Before – not preparing for “unpopular answers”
- During – multi-part questions, let-me-help troubleshooting, watering down the magic wand question
- After – leaving out the verbs in the summary
Before – Not preparing for “unpopular truths”
What will happen when the customer answers that they don’t have that problem? If the customer doesn’t need the solution you’ve already envisioned? When the customer isn’t getting value from the product you’ve already built?
Here’s what happens in most situations: you learn from customers, you summarize the unpopular truths, and you share them… and your manager/team doesn’t react well:
“You must not have talked to the right people.”
“What did you ask, anyways? You probably didn’t explain it well enough.”
“They just don’t get it.”
Or perhaps they say, “that’s interesting… but it’s not enough evidence for us to change course from our initial direction.”
The point of customer research is to mitigate risk and help us to make smarter bets. It allows us to adjust our prioritization and decisions based on that new information. If research can’t change minds, why do it?
That’s why we talk about invalidating hypotheses. We need to approach customer development looking to be disproved, not ready to claim victory. And we need to bring others along on the journey.
It’s important to explicitly tell stakeholders, “we might hear that customers don’t need this” and to share the questions you ask to reduce false positives.
Provoke them to add or modify your questions: “if I’m asking these questions and the customer is telling me they’re not interested, what would make you confident that wow, the customer genuinely isn’t interested? What level of data would it take to convince us to change course?”
Common stakeholder reactions to customer research
- Willing to pivot away from an early idea based on interviews alone
- Receptive to interview insights if they’ve personally participated in at least some of the interviews
- Will want to combine qualitative insights with quantitative data to strengthen their confidence
- There’s nothing that will change their minds
For that last one, is it worth doing upfront customer research at all? In my experience, no.
If the decision is already baked, it isn’t worth researching prior to product launch. There are only so many things we can focus on, and I’d prefer to ask questions where we’re ready to listen and respond to the answers. Instead, I’d invest that time in evaluating the customer experience post-launch and making recommendations for the next version.
During: Multi-part questions
“Could you tell me about how you manage your email inbox and your other communication tools?” It’s hard for the interviewee to know what part to answer first, and to remember multiple parts of a question. The cognitive energy they use to hold that question in their hand takes away from the effort they put into responding.
When you combine >1 questions, you get <1 good answer.
It’s okay to ask multiple similar questions in a row instead!
Instead of: “Could you tell me about how you manage your email inbox and your other communication tools?”
I’d start with: “What’s your strategy for handling work communications?”
The first tool that the person talks about is the most important, or most frustrating. You’ll get to compare how they talk about email vs Slack, or Zoom vs unscheduled phone calls. If they choose to mostly complain about one tool, then the others aren’t a problem they need to solve.
During: watering down the “magic wand” question
One of my favorite questions to ask in research interviews is “the magic wand question”:
If you could wave a magic wand and change anything about how you _____ – doesn’t have to be possible! – what would you change?
This is completely different from “if you could wave a magic wand, what would you want?”
I’ve tried asking both versions in the same interview with the same person and gotten completely different answers. When we ask some form of “what do you want?”, we are getting away from the ‘why’. That question feels like it’s asking for a solution, not an underlying problem. People are happy to give you solutions, and we’re happy to hear them – and they’re not what we need.
It’s critical to nudge the interviewee to put themselves in the mindset of doing _____.
They need to remember the frustration or limits (or pride, anticipation, satisfaction) they’re feeling. That emotion is what motivates people!
The “it doesn’t have to be possible” is also necessary. It’s inviting them to think beyond incremental solutions or that feature that your competitor offers. This is the space where new product opportunity lives!
Sometimes people don’t want to answer the magic wand question and insist, “what I want most is X!” That’s okay, because you can acknowledge their desire and then ask them to tell you more about it. “I’d love to understand how you’d see yourself using X…”
What if people don’t ‘get’ the magic wand question?
Sometimes you share limited cultural context with your interviewee, and you may wonder if they will ‘get’ the intention behind the magic wand question. In those cases, you can give an example:
If you could wave a magic wand and change anything about X – doesn’t have to be possible! – what would you change?
[pause, if they seem confused, continue with…]
For example, I used to get stuck in traffic on my way to pick up my kids, and I wished I could push a button and have my car go straight up like a helicopter and fly over the traffic jam!
Then they get it!
During: let-me-help troubleshooting
Here’s something that will happen – particularly if you’re working on an existing product: you’re talking to a customer, they mention a problem, and voila!
- You know the solution.
- You know that this feature does exist already!
- You recognize this as a known issue and want to tell them the fix.
- You suspect a bug and want to try to reproduce it so you can file it.
Stop! Don’t interrupt your interview to go into let-me-help, troubleshooting mode.
As soon as you do, you reset the expectations of the interview.
When you say “oh! I can help!” you’re signaling that now, you are the expert. The customer is here to learn from you, instead of you learning from them.
Once you’ve done that, most interviewees will follow your lead and defer to you. They’ll talk less about their experiences and ask you questions instead. That’s a lost learning opportunity.
You can help – later. Take a note of what you want to say, and send it in a followup email. You can thank them, and add a note like:
“By the way, you mentioned wanting X feature. I wanted to let you know that it exists – though we clearly haven’t done a good job enough of exposing it to customers! You can find X under the settings menu (see screenshot). Let me know how it works for you!”
After – leaving out the verbs in the summary
You’ve done 10 interviews, you have a pile of notes and you synthesize them down to the most important bullet points. Just make sure you aren’t leaving out the verbs!
Let me give some examples:
With verbs: Interviewees wanted to pay using their local currency
Verbless: More payment options
With verbs: Admins need to be able to remove offensive posts and prevent CEO from getting flooded with ‘like’ notifications
Verbless: Need better moderation tools
Verbs are actions. They’re behaviors. You can track them and watch how they change.
When we leave them out, it’s too easy to summarize rich meaning into these ‘jargon phrases’ where everyone can apply their own personal definition. And then we lose sight of the actual problem. We inadvertently replace the actual problem with everyone’s individual impression of what “more payment options” means.
“More payment options” – if you weren’t there for the original conversation, you might reasonably think that meant:
- supporting more payment providers like PayPal, AliPay,
- supporting payment plan options like Klarna
- allowing people to pay in their preferred currency
- still accepting only one currency but showing the conversion to customer’s preferred currency
- offering monthly or yearly or invoice-based billing
- and probably many more
Verbs aren’t the magic ingredient – but they’re a useful signal. When you keep the verbs in your summary, you’re more likely to preserve the root problems. That allows you to build solutions that actually fulfill the customers’ jobs-to-be-done.
In summary, the 5 customer development problems to avoid:
- Make sure you’re asking the right questions and setting the expectation for ‘unpopular truths’
- Don’t ask multi-part questions. If you accidentally do, stop and reset: “I’ll ask 2 separate questions. First…”
- Keep the magic wand wording: “If you could change one thing about ____ – doesn’t have to be possible – what would you change?”
- Don’t troubleshoot during the interview – follow up later with advice, solutions, bug fixes
- Keep the verbs – don’t over-summarize your insights