Learning What Customers Want

As a startup founder I need to know if we are building the right product or focusing on the right feature. Our resources are limited and we might not get a second chance. Focusing on the wrong thing can be damaging and sometimes catastrophic. We can sit around the conference table and write lots of ideas on the whiteboard. Then we can psych ourselves into thinking they are billion dollar ideas or they are doomed to failure.

Or we could just talk to customers and learn what they want.

That’s interesting … tell me more

Why is it so important to talk to customers?

1. It’s easy
2. We can uncover customer needs and build a product that focuses on solving their most important problems.
3. It can potentially save us from going on a wild goose chase and burning through all of our money.

Still not convinced? Read the Four Steps to the Epiphany by Steve Blank. He invented the lean startup methodology and formalized the customer development process. Steve recounts many teams stacked with talent that executed well and spent millions to bring products to market that nobody cared about. And he talks about how talking to customers is the best way to get your company to product-market fit.

How to get feedback from customers

There are many ways to get feedback from customers, but the three best ways I’ve found to uncover customer needs are interviews, user studies, and surveys. When starting a product, do open-ended customer interviews with the goal of understanding what life is like for your customers. Do this and save some time to test your high level ideas.

Then you’re going to want to do customer surveys. Surveys allow you to easily see if your learnings from user interviews can scale. 10 interviews may not be representative of a large customer base, but 100+ surveys is going to make you a lot more confident. Plus, you now have a large pool of potential customers, many of which signaled an interest in buying your product.

Once you’ve settled on a product, then you can move on to the design process and validate your ideas. I suggest using lo-fidelity wireframes and getting feedback with user studies. You’ll be taking notes on their first impressions and trying to figure out what works and what doesn’t with the user experience. At that point, you can build the product and start developing a detailed sales and marketing plan.

User Interviews

You’ve got an awesome idea and I’m sure you’re antsy to build it ASAP. But before you jump into product development, do your homework and understand your target customers. You’ll want to talk to a minimum of 5 customers and should target talking to between 10 to 20. Every person you talk to is going to bring a new perspective and everyone is going to strengthen your understanding of the market.

For that reason, user interviews should be required for starting a company. Otherwise, you could spend a year building a product and launch to the sound of crickets. They’re extremely useful for new products, but probably overkill for new features. You’re better off putting together a survey for new features. They’re also helpful for a product reset, since they give you qualitative data on why you aren’t getting to product-market fit.

When you decide to do interviews, there a few things you need to consider. Who is the ideal person to talk to? How I can I find them? What’s the format of the interview? What are my goals? What questions should I ask? What should I do with the results?

Target Customer

Come up with a hypothesis of who your target customer is. You want to understand a niche and identify common problems for similar customers. I’d group customers like this:

• B2B: Industry, Job Function (E.g., Sales), and Company Size
• B2C: Demographics, Psychographics

Finding People to Interview

Start by interviewing all of the relevant people that you know. They’ll give you great feedback and are happy to talk at length. Unfortunately, they can try to please you with their answers. They want you to succeed and don’t want to lower your confidence. The closer you are to someone, the more you have to set expectations with them at the start of the interview. Let them know “I really need to know if my baby is ugly. I don’t need a cheerleader; I need honest feedback. I’m excited about this project, but I need to know the flaws before I spend months or years of my life on it.” And once you finish up the interview, be sure to ask for recommendations on who else you should interview.

Once you talk to people you know, here are some more strategies to find people:

1. Search for local businesses or people in your target market. Send them cold emails and see if they’d like to talk about an EXCITING NEW PRODUCT that could help their business.
2. Search online for influencers in your target market. Ask them in a short email if they can give you some feedback on your idea. The worst thing that could happen is that they ignore your email. If you do get a response, please keep the interview short and be respectful of their time.
3. Go to a coffee shop and ask if someone has 15 or 30 minutes to talk. You may need to give people a reward (E.g., $5 gift card) for their participation.
4. You can also run an adwords campaign targeting people and feed it back to your landing page. I suggest using a simple launchrock page. Then you can ask for email or phone to keep them updated on the products launch, then follow up a day or two later inviting them to a research interview. Or you can directly ask on the landing page if they would like to participate in a research interview for a new product.

Here’s another reason to do the work now and get ten or more interviews: You are doing lead generation and qualifying prospects before you build your product! If you did your interview correctly, you will have found their pain points, talked about how you address them, and gotten a dollar value for how much solving the problem is worth. That way when you launch, you’ll have a group of 10+ customers who are primed and ready to buy. These people are telling you exactly what to build to solve their pain points. All you have to do is build it.

Interview Format

Interviews don’t have to be formal and certainly don’t have to be paid. You want people to be comfortable and primed to give you honest feedback that you can learn from.

I generally reserve an hour for interviews, but am happy to stop at 30 minutes unless we’ve got momentum. Print out a copy of the questions before the interview. Make sure that you leave 2+ blank lines between each question. Don’t type on your computer. It’s loud and can distract your subject.

Have two people. One will give the interview and the other will take notes. If you’re in a bind, you have to ask questions and take notes. Good luck and I hope you can write fast.

So you grew up in New Orleans?

Interview Questions

It is all about them. It’s not about you making a sales pitch or looking good and gaining confidence. The purpose of this call is for you to learn and develop an understanding of your customer. You should start with vague questions at first and gradually move to specifics. I’m deliberately vague at the start, since I want the time to go on a tangent and learn about unexpected things that are important to them.

Background / Informative Chit Chat

• Have you worked here long?
• How did you get into sales?

Then you want to get into the meat of the research. Plan out your follow ups in depth and have a choose your own adventure set up for the interview. Get to know not just the what, but how and why. People love telling you about a great experience and a really bad experience. You’ve got to quickly get through areas they don’t care about and find the outliers where an impact and a boatload of money is made.

Their Specific Workflows

• How do you do find prospects?
• Do you use a CRM?
• What do you like about it?
• What do you dislike about it?
• How important is it to you?

I’d cover a few specific workflows that pertain to your product or feature. Then you can move on to get feedback about your solution. Keep it at a high level so they can understand what you are offering. Then gradually give them more information so you can learn what to focus on.

Feedback on Product

• Would you be interested in a tool that helps you find leads?
• The tool would help you do X. Here’s how it would work …
• Is this a good fit for you?
• Why or why not?
• How much do you think a tool that does this would cost?

Lightening Round

I sometimes like to throw in a rapid fire section at the end where I ask if they think features are important or not important. I tell them not to explain why. I just want to get rapid fire, survey-style feedback at this point. Here are some examples from an interview I did on a sales prospecting tool:

• Import from another tool
• Filter the search results
• Cross reference where someone works and find potential coworkers

Want more? Here are some real Interview Questions from when we considered building a sales prospecting tool.

How to improve your interview based on feedback

In your first few interviews, you’ll hear things that differ from your expectations. Take those differences and update your questions before you talk to your next interviewee. You‘ve got to find out if the differences are outliers or part of a trend.

If you’ve done a few interviews and didn’t get much value from some of the questions, drop them. This will give your interviewee time to speculate and go down a rabbit hole. That’s where the real learning happens.

What do do with the results?

Each successive interview should bring you closer to identifying a customer need that you can solve. If you have done ten interviews and are left scratching your head, the market is probably not there. If the customers are excited to talk to you and want your help with the same pain points, you’re on to something. Grab a hold of this and don’t let go. Focus your future interviews on this problem and try to solve it like a Rubik’s cube. You’re not going to get the answer immediately, but you’re going to get closer and when you actually do, then you can print money.

If you are getting stuck and can’t identify a huge pain point, try going up one level of scope and looking at a larger set of problems. First, ask a less narrow set of questions and go for generic “what are the bottlenecks that frequently hold you up”, “if you had a genie and three wishes to help you in your job, what would you ask for”, “what are the most frustrating things about your job” type questions. If that still doesn’t hit on anything, then you can expand who you are talking to. Talk to people with different job titles and different responsibilities. Think back to your previous jobs. Eventually you are going to find some massive inefficiencies that your business could fix.

Customer Surveys

Surveys are great for planning your product roadmap. You can see what people like about your current product and what needs improvement. Then you can describe potential new features and see what they are most interested in using. Since you can get false positives in a small sample size, surveys work best if you have a large group of users.

Your goal here is feedback at scale. Think of it like a conversion funnel. We need awareness at the top of the funnel and we need them to finish the form. I’ve had great luck using Intercom to target our users with an in-app message or a short email with a link. Once you get them to the survey, the best way to do this is to write a survey that is quick and easy to fill out.

Survey Says … !

Survey Questions

The feedback isn’t personalized, so you need to be much more focused than in an interview. Furthermore, you want to be clear and concise in your questions, so you don’t confuse or bias your users. I like to say statements and have respondents use the Likert Scale to show how much they agree or disagree. I also use checkboxes to get a feel for preferences (which features/tools have you used? which of these are you interested in?) but try to avoid short answer at all costs.

Track your Surveys

Always start by collecting contact information. This lets you connect the user back to your analytics, you ensure that you’re not double counting the same person, and it lets you connect with a valuable customer later on. You can also personalize future correspondence to this person based on their responses.

Building a relationship with your customers is powerful. It’s a great experience for someone to make a mild complaint on a survey and have the company immediately solve their problem. You can also call them back to have a longer qualitative interview and build a tailor made solution. Or later on, you can get feedback on your design.

I begin the survey by getting feedback on the current product, then I check to see if they are using any substitutes to the new features I’m proposing, then I get feedback on what they like about the new features. Lastly, I ask for how much they would pay for the new feature. After all, if it provides no monetary value, then it ain’t worth building.

Feedback on the Current Product

• What kind of articles would you like us to include? (checkboxes for 5+ categories)
• I like the articles provided in the product. (likert scale)
• I prefer X to Y (likert scale).
• I would like the ability to do X. (likert scale)

Potential for the new idea

• How do you track customers? (5+ checkboxes, Other)
• I regularly send email to existing customers (likert)
• I‘d like a new feature that makes it easier to send email updates to customers (likert)
• How much would you pay for a new email feature (radio buttons: $5, $10, $20+, nothing)

Here’s an example survey I used last year. Note: I removed Email entry.

User Studies

You’ve gotten feedback from your interview. Now you need to get specific feedback on your high level design and user experience. With user studies, you take anywhere from 15 minutes to an hour and watch how users interact with your new design. Your goal is to find usability issues and confusion in the UI. The users will identify problems and you’ll have a chance to redesign and fix them before you launch.

Presenting your design

There are many different ways to show your design to users. You can show them a finished product (and you should if this is a big launch), but you will want to get feedback before you develop the software. The cost of making a design change is much, much greater after you have written an app than when it is just a few sketches on paper. If you still want the feel of a finished product, you can use tools like inVision to create a clickable prototype without writing code. You can also print out a high fidelity mockup of your screens and talk to them about it. Or you can put together wireframes and give the user a bare bones set of mockups for your product.

I prefer wireframes. If you do a high fidelity mockup or a clickable prototype, people can focus on the details (color, typography, etc), instead of the user experience. Plus, wireframes are easier to put together. I draw mine in pencil, but they’re easy to build in something like Sketch. Tons of people use Balsamiq, an easy to use wire-framing tool, and I have had good experiences with Pencil, a free OSS tool.

Users are going to go bonkers over that rotating banner, amirite!!?

Conducting the Study

Your goal is to not bias the user. It is natural to help your users, but you have to resist. If you tell someone how to use your product and give them training, then you don’t get to see what was confusing. Get the user to think out loud and be extremely vague. As in the court of law, do not lead the witness. Only after the user is done with a section should you tell them “what they were supposed to do with the app”.

I typically start a user study with the agenda and jump straight into the mockups without an explanation. Explaining what it does and how it works will bias your users. You don’t need to provide any more background information than the topic and can start with a statement like this: “We are going to talk about a new product that helps salespeople find prospects on social media. We haven’t gotten feedback from users yet and want to make the product as usable as possible before we build it. Don’t be afraid to call my baby ugly and let me know of any things that we should fix before we launch”.

Listening Time

I start the user study by showing them the first wireframe and saying something like “what do you think this is” or “what are some things you could do on this screen”. That gets them talking. They will point out the confusing parts and say things like “I guess this is a dashboard”.

Once they are done talking, then you can start asking them how they would perform basic tasks like “How would you find a user?” Then you shut up and take lots of notes while they are talking. Resist the temptation to help them if they get lost. This is when you’ll get your most valuable feedback. Ask them clarifying questions, “what do you think you should do?”, “why did you click that button?”, “what do you think this link will do”, “what will happen if you click the cancel button?”

With wireframes, you are looking for clarity of user experience. You’ll have to keep your questions and tasks short, since they aren’t going to be able to imagine how the design would look when state is modified. However, when you move on to testing a prototype, you should ask users to perform complex tasks and see how your whole design works together.

Once you’re through with question time, be sure to ask if they have any more feedback. I like to ask “Is there anything that felt out of place?”, “is there anything that was missing”, and “what would you change?”

Here’s an example set of tasks for a user study we did before a major product update last year.

Good luck!