Before we go into the details of techniques for finding customers, let’s address the skepticism you’re probably feeling. You may think, “So you’re telling me that people are willing to spend time talking to someone they don’t know who has no product to show them?”

I’ve found that no one believes this at face value. People are busy. They hate telemarketers, advertising, and spam. How is this different?

To start believing that people—useful people—will want to talk to you, you need to understand the people you’re looking for.

How do these human tendencies help when you want talk with prospective customers about your idea?

You send out an email from your personal email address that addresses the recipient personally. The recipient sees it and says to himself: someone is asking for my help.

It’s one of the few globally universal psychological constructs—regardless of culture or income, we derive happiness from investing our resources to help others.^[15] Because you’ve specified why you want to talk with this specific person, the recipient has a feeling of ownership. He can’t ignore it, assuming that someone else will deal with it.^[16]

It seems that we particularly like giving our time to causes that align with our own self-identity.^[17] So when you send an email asking someone to help you out by talking about something he’s already interested in, that’s a win-win.

Your recipient responds and commits to a conversation with you. As you begin the conversation, you make it clear that she’s the expert. You want to learn from her experiences.

Wanting to sound smart isn’t as egotistical as it sounds; we are rewarded when we gain the respect of others. Psychologist Abraham Maslow defined this as esteem (achievement, respect of others, respect by others) and included it in his hierarchy of needs.^[18] When I thank someone at the end of a customer development interview, it’s common for him to reply with something like, “No, thank you! I’m so glad that I had something useful to share!”

My theory is that most of us are unrecognized experts in the things we do every day—whether it’s keeping the family fed or debugging code or coordinating large meetings. It’s a pleasant change to hear from someone who doesn’t take that for granted.

As your interviewee recounts his frustrations, he gets to vent and experience catharsis. But more importantly, as you ask follow-up questions, you give him the feeling that things can get better. That’s not what we’re used to:

Unless we personally experience the problems that we are trying to solve, it may still be difficult to imagine that customers will be eager to talk to us.

Your first stop should be your immediate circle of friends and coworkers. (Presumably you’re trying to enter a market that you and your connections know something about already; if your personal network is completely irrelevant, that’s probably a sign that you’re trying to enter a market you know nothing about, which is a terrible idea.)

How should you start? Just because you have 500 people in your contact list doesn’t mean you should blast out an email with 500 recipients.

You don’t have the time to interview a bunch of people who are obviously not relevant to the problem you’re solving, and you don’t have endless social capital to burn through.

It’s not that you’re going to ask your old boss, your child’s soccer coach, and your Aunt Mabel for an interview (unless, of course, they have the problem that you are trying to solve). While you may already know some prospective customers, it’s more likely that you’ll find people within the larger universe of second-degree connections. You’ll need to ask your connections to introduce you to their friends, coworkers, and family members who have the problem that you’re trying to solve.

For many of you, personal connections and direct introductions are not going to lead to enough relevant people. Most likely, you’ll need to try a variety of approaches to figure out what works.

Finding people on LinkedIn

LinkedIn is typically the easiest way to find people who work in a specific industry or hold a certain job title. LinkedIn also allows you to narrow your search based on skills and expertise.^[21] (Of course, job title and experience are usually more relevant for enterprise products than for consumer products.)

You can probably find at least a couple of people who fit your criteria and are first- or second-degree connections. That may be all you need to seed your initial search: one or two receptive people can help you identify more people you should be talking to.

You can send messages directly to your first-degree connections, either through LinkedIn’s own messaging system or by finding the person’s contact email address on their profile. (You’ll need to be logged in to LinkedIn to see her full profile.) If someone is a second-degree connection, LinkedIn allows you to ask the common contact to forward a message to him (Figure 3-2).

Figure 3-2. When sending messages to a second-degree connection, include a ready-to-forward snippet to make the introduction easier for your contact

I don’t recommend forwarding messages to a third-degree LinkedIn connection. With two degrees of separation, the sense of being vouched for fades a lot. If you want to reach out to third-degree connections and maximize your odds of getting a response, you’re better off upgrading to a premium subscription and using LinkedIn inMail to directly contact people.

LinkedIn includes inMail credits with premium subscriptions. Because it costs money to send these messages, the recipients are somewhat less likely to perceive them as spam. Your inMails only count against your total if the recipient accepts your message—so you can do some tweaking of your message cheaply. A premium subscription gives you access to additional filters, which can also help you find people.

In some cases, searching LinkedIn serves as a lightweight customer development tool. You may find that your advanced search for a specific job title and industry combination doesn’t return as many results as you expected—in other words, your target market may be too small.

One of the challenges with LinkedIn is that you are competing with recruiters, some of whom take a spray-and-pray approach to any person who meets some vague keyword searches. How do you differentiate yourself? Be specific, be personalized, and be brief.

Introducing Yourself via LinkedIn

LinkedIn messages have a 1,000-character limit, so you’ll need to be brief in describing your purpose and method. Here’s the information to include:

• Acknowledge why this person specifically can help you

• Briefly (5–10 words) state the type of problem you’re working on

• Clearly state a minimal request for time (e.g., a three-question survey or a five-minute call)

Your message may look something like this:

Hi, my name is ________, and I’m trying to learn more about how small companies choose SaaS products. Can you take two minutes to answer this three-question survey? [URL]

I appreciate your time, and I’m happy to return the favor however I can!

Thanks,

__________

Surveys are not intended to replace customer interviews. In this scenario, the survey serves two purposes. It allows you to ask a couple of screening questions to ensure that this is a relevant target customer, and more importantly, it serves as a foot in the door to make initial contact. Because contacting people via LinkedIn is somewhat impersonal, you’ll need to make your initial request as lightweight as possible in order to get responses. Once you’ve started a conversation with someone, she’s far more likely to agree to follow-up questions or a longer interview.

Finding people on Quora

Quora works well as a tool for finding people for both consumer and enterprise products, though the site remains somewhat skewed toward technology topics and the Internet-savvy. (Quora initially launched by invitation only; the early community was heavily concentrated among San Francisco and New York technical circles.)

Quora has a much smaller user base than LinkedIn or Twitter, but tends to attract more passionate and vocal people. I’ve typically had better luck connecting with people via Quora, and I suspect this is why. (I’ve also noticed that I’m personally more likely to respond to messages sent to my Quora inbox than to my LinkedIn inbox.)

If you use Quora, you’ll want to do some homework first. Quora users are fairly protective of their community. Be a contributing member, not just a taker. Cover the basics: upload a photo, complete your profile information, and follow relevant topics. Spend some time reading questions and answers in the topics relevant to your product area. Try to write some answers, or at least comment on some answers, in a way that adds value to the community without self-promotion.

As a bonus, going through that exercise will probably give you ideas for additional questions to ask. People who have written questions or comments on the topics you’re following are a good place to start. They’ve self-identified as experts who are probably willing to continue talking on the topic. But even the people who have followed the question have demonstrated some intent and may be receptive to your messages (Figure 3-3).

Please do not solicit customer development participants in a Quora thread. Your question will be down-voted swiftly and annoy the very people that you’re hoping to learn from.

Instead, you’ll want to craft individual private messages for each person. When you find an interesting person, click on her name to get to her profile page. From a user’s profile page, there is a link to send her a direct message in the sidebar. (It’s somewhat hidden; see Figure 3-4.)

Figure 3-3. People following the Quora question “Customer Development: How did you test and run a hypothesis?”

Figure 3-4. In the lefthand sidebar of a user’s profile page, you can use the Message link to send him a private message

Start by sending only a couple of outreach messages to Quora users and then wait to make sure that your message was effective. (You don’t want to burn through all of your prospects if your message needs tweaking.) You may need to wait a couple of days to receive a response if people have turned off the notification that emails them when they have a new private message.

Using blog posts

There may not be very many people who care enough to author an entire blog on the area you’re exploring, but the ones that do exist are definitely written by people you need to connect with! A few of the more comprehensive blog aggregators are:

• Alltop.com

• Blogarama.com

• Blogs.com

Of course, there are far fewer blogs entirely devoted to a particular topic than there are individual blog posts on a topic. For finding standalone blog posts on a specific topic, you don’t need a blog aggregator; you need a search engine.

Trial-and-error searches, patience, and a lot of open browser tab windows will eventually yield results. One tip: if you’re using Google Search, click on the Search Tools button and change the “Any time” option to “Past year” (Figure 3-5). You need to find people who experienced your problem recently, not back in 2005.

Figure 3-5. Use Google’s “Past year” filter to omit outdated search results when hunting for blog posts

Using a landing page

Building a landing page and using Google AdWords to direct traffic to it was the original customer development tactic. Eric Ries wrote about it and many companies (including KISSmetrics) used that method. But as more businesses have begun using AdWords, the price per click has gotten much more expensive. The cost of acquiring visitors via AdWords may be prohibitively high unless you are in a narrow niche.

However, if you have a large Twitter following, a popular blog, or a base of existing customers, you can get a reasonable amount of site traffic. You can build it yourself or use a hosted service like LaunchRock (http://www.launchrock.com), which also helps you promote the site via Twitter. The advantage of using a landing page is that it’s easy to combine a short survey with the ability to follow up and contact people for a longer interview (Figure 3-6).

Figure 3-6. A short survey helps qualify potential customers without hurting your response rate; be sure to collect email addresses so you can follow up^[24]

Intuit calls visiting the customer’s home or office “Follow Me Home”; user researchers may refer to this as an ethnographic interview or a field visit. This method predates the lean startup movement by decades.^[25]

Observing the customer in his natural environment is the highest-fidelity method of customer development. You’ll learn about factors like noise level, neatness of the environment, whether your customer has privacy or is constantly interrupted, whether he has outdated or upgraded technology, and who comes to talk to him while you’re there. If you ask a question about how the customer does something, he can show you, not just describe it.

Face-to-face conversations are more personal; he can see your face and body language, which helps in building a relationship with this potential earlyvangelist. You may also meet other people in the household or workplace for later follow-up questions.

On-site interviews are by far the hardest to coordinate. Privacy concerns may prevent employees from bringing an outsider into their office, or at least require getting an NDA signed in advance. For consumers, a messy home or distractions from family members may discourage the customer from agreeing to talk. The need to obtain permission or approval in advance may also lead to delays or last-minute cancellations.

On-site interviews work well for:

Another high-fidelity method for customer development interviews is a face-to-face conversation in a neutral public location. Even if you don’t know what the prospective customer’s noisy cubicle or messy kitchen looks like, you can see his body language and watch his facial expressions as he describes his problem situations. Because he can see you as well, this method helps build rapport with the customer.

The cons are similar to those with the site visit. You’ll need to identify appropriate venues. Once you’ve found them, your customer has to figure out how to get there, park her car, and (assuming you’ve never met before) figure out how to identify you. I find this method to be the most challenging for note-taking (there is often background noise, and sometimes a lack of reliable surfaces to write on).

In-person interviews at a neutral location work well for:

This is the method that I most commonly use for customer development interviews.

I don’t get any visual information, but I can learn a lot from verbal intonations and pauses. (I also suspect that people may be more honest about things that embarrass or frustrate them when they don’t have to make eye contact.)

More importantly, I can conduct interviews faster so I can learn faster. When I ask for phone conversations, I get higher response rates. Scheduling is easier because no one has to account for travel time or other people’s permission. I can take more comprehensive notes because I can type on my laptop without appearing rude.

Phone interviews work well for:

Video chats share most of the same advantages and disadvantages as phone conversations. You also gain the benefit of seeing the customer’s facial expressions and body language and potentially even what’s on her computer screen. If your target customer is tech-savvy (or younger than 20), this is a great option. If not, approach with caution. After spending 5 minutes trying to figure out if video is working, your flustered customer is not going to be able to give you the best possible information in the remaining 15 minutes.

Video chats work well for:

I’ll admit it: instant messaging (IM) is not my favorite approach for conducting interviews. Text-only communication provides the least amount of information with the highest risk of misinterpretation. Most of us have difficulty interpreting the intent or emotional severity of written messages.^[26] I’ve also noticed through various user research projects over the years that people are more self-conscious and more likely to self-censor what they write than what they say.

Still, sometimes text-based communication is the best available option. Live chat apps give you the opportunity to have a conversation with someone while he’s on your site. If you don’t talk with that prospective customer in real time, you’ll have no way to contact him in the future. You also may want to schedule conversations over IM if you’re concerned about audio quality—talking with interviewees in other countries when voice-over-IP is unreliable or talking with interviewees whose accents make it difficult to understand what they’re saying.

IM works well for:

Once you’ve gotten an affirmative response from a person, then you can go into more tactical detail. Your job: to make it incredibly easy for her to commit to a time to talk.

For a phone call, I recommend offering three or four options, not all on the same day of the week or the same hour of the day. Make sure you are explicit about what time zone you are proposing.

Better yet, figure out in advance if your interviewee lives in a different time zone and suggest appropriate times. A 9 a.m. call isn’t a considerate suggestion if you are in New York and your interviewee is in Los Angeles.

If you use Gmail, I recommend downloading a browser plug-in called Rapportive (http://www.rapportive.com). When you type in an email address, Rapportive pulls associated publicly available data from LinkedIn, Twitter, and Google and displays it in a sidebar next to your message compose window. It may include a photo, a location, a job title, and recent social media posts (Figure 3-8).

Figure 3-8. Rapportive shows you a sidebar with information the person has shared on social media: avatar, location, job title

If you use Outlook on Windows, you can use the Outlook Social Connector (http://www.linkedin.com/static?key=microsoft_outlook) to view relevant LinkedIn information as you’re composing a message. Not all consumers have LinkedIn profiles, but most businesspeople or information workers have a profile that contains at least location and job title.

The best locations for face-to-face interviews are convenient for both cars and public transit, have plenty of space, and don’t suffer from so much popularity that you won’t be able to find a table or quiet area. I usually suggest meeting at Starbucks; they’re accustomed to people having work meetings and their WiFi is reliable. Hotels with large, quiet lobbies are also a good option.

Here’s an example I used at KISSmetrics:

This gives the recipient all of the necessary information she needs to make a decision: expected time outlay, times she can check against her calendar, information about the venue,^[28] and locations (so she doesn’t schedule back-to-back meetings at opposite ends of town).

Odds are that sending a message with several options will eliminate the need for playing email tag to schedule a meeting. But even if the interviewee can’t accept one of your suggestions, you’ve set some parameters so that she can easily propose an alternative that is likely to work.

Once the interviewee has chosen a time, I recommend sending a calendar invitation as well as your cell phone number. It’s also useful to send a reminder email the day before, or in the morning before an afternoon call.

Most of you won’t find yourself trying to cram in interviews from the dozens of prospective customers who immediately agree to talk to you. (If you do, congratulations! That’s a good problem to have.) But you may think that you’ll set aside a half-day and try to cram in as many interviews as possible in that time period. In the beginning, that’s not setting yourself up for success.

You need time to prepare, time to do the interview, and time to go over your notes right after the interview and let the most important things surface. Then you need a few minutes to prepare for the next interview. (The amount of time you’ll need between interviews falls dramatically after you’ve gained some comfort in conducting them, but in the beginning you’ll need the breaks.)

For this reason, I suggest that you start by scheduling no more than one 20–30 minute interview per hour (Figure 3-9).

Figure 3-9. Your interview schedule may look like this; you should block off time before an appointment for prep and some time afterward for summarizing notes, which I’ll talk about more in Chapter 6

Why would you do that?

Over time you’ll be able to schedule interviews closer together. I’m usually comfortable scheduling two 20-minute interviews in the same hour. Now that I have more expertise interviewing, I can typically maximize the information I get from the interviewee in the first 20 minutes. It’s also gotten easier for me to context-switch from one call to the next. It will get easier for you the more interviews you do!

After you send an initial request to talk, I recommend waiting at least a few days to respond. For business customers, I typically wait a week. At that point, I’ll resend the request with an additional message like “I’d still love to talk with you. If you’re available this week, let me know and I can coordinate a time.” About 20% of the time, the interviewee responds to this follow-up email and we successfully schedule and complete an interview. If you don’t hear back, I don’t recommend pursuing the person further. Don’t be annoying.

If you’ve sent out a number of requests and no one has responded, conscript a friend or colleague to read your message and give you feedback. Making your message shorter or clearer, or changing the tone may be all you need.

What if you’ve gotten feedback on your message, improved it, and then sent 10 or more requests that still haven’t netted any responses? Well, consider that your first hypothesis invalidation. Either you’re reaching out to the wrong people or the topic you’ve mentioned isn’t interesting. If this happens, you may want to return to Chapter 2 and revise your hypothesis.

In my experience, somewhere between 5% to 10% of the interviews you schedule will end up in cancelations or no-shows.

When I did usability testing and scheduled sessions all day long, I actually planned my day around this. My rule of thumb was that 1 out of 10 participants would be a no-show. Because of this, I’d typically schedule sessions without a break. (This meant that when I did get a full turnout I had to skip lunch.)

In my experience, customer development tends to have a lower rate of no-shows than usability testing. I think this is because customer development takes a personal approach to recruiting participants; it’s harder to disappoint a known individual than a faceless user research recruiter.

If an interviewee cancels or fails to show up, wait a day or two and then send a message offering to reschedule. Typically the interviewee responds and you’ll be able to reschedule successfully, but if you don’t get a response, let it go.