If you add nonfunctional experiences to an existing product or website, you also make it harder to use and less rewarding. For customers who are evaluating your product based on reliability and credibility, a broken link or dead end makes them wonder, “What else can’t I trust in this product?”

For example, years ago I was moderating usability test sessions for a financial services application that I had just redesigned. User response was unexpectedly dismal and I had no idea why. One participant immediately told me, “I would never use this website.”

Why not?

He pointed to the bottom of the screen. “There’s no privacy link,” he said.

Was I talking to a privacy nutcase? I wondered to myself, and asked, “How important is it to you to read the privacy policy before using a website?”

He looked baffled, and said, “All I know is, bank websites are supposed to have that little picture of a lock and a privacy link at the bottom, or else you’re not supposed to use them.”

By accidentally omitting one tiny credibility marker, I had inadvertently tainted customer reactions to the product. Before the next participant arrived, I quickly added in a lock image and a fake privacy policy link to the demo. Customer feedback immediately flipped from strongly negative to positive.

Even if you’re working on a standalone demo, you may be dealing with customers who have exacting standards. I’ve seen executives who get riled up over a typo on a PowerPoint slide. In the startup world, the phrase “If you’re not embarrassed by your first version, you waited too long to release” gets thrown around a lot. In the enterprise world, if you are embarrassed by your first version, you may not get the chance to show a second version. Invest a few extra hours in spellchecking, making sure your links aren’t broken, and cleaning up your images.

On the other hand, if you make a demo or prototype look too good, you run the risk of customers thinking it’s already built (or in the process of being built). They may delay purchases or upgrades assuming they can just wait for the new version. Then what happens if you invalidate your hypothesis and decide not to build the product or feature after all? You may wind up with customers who are disappointed and angry.

Use a sketch

Even if you’ve clearly explained that you are doing research to find out what customers want and not showing a product direction, and reiterated that what they will see is subject to change, it’s psychologically hard to see something and then lose it. A useful way around this is to create something that looks good, but is clearly fake. You may want to use a program like Balsamiq, which allows you to quickly create consistent, clear, but cartoon-like sketches. No one seeing a Balsamiq mockup (Figure 8-1) mistakes it for a finished product.

Figure 8-1. Balsamiq wireframes approximate layout and interactions but are clearly not working versions

Companies with existing customers need to define MVP a bit differently than a startup would.

At Aetna, product development uses the term “minimum sellable product.” At Yammer, we emphasize the word viable. It’s OK—even required—to go a little bit above and beyond the minimum to create a cohesive experience.

It’s different for startups

For startups, an MVP is the minimum amount of product that provokes a nonneutral response. That’s because new companies face different risks than established companies. As Steve Blank says, “On Day One, a startup has no customers.... On Day One, a startup is a faith-based initiative built on guesses.”^[61] For a startup, the first risk to overcome is “Will anyone care enough to respond to this idea?” Even a tiny confirmation of interest like clicking a link or scrolling down a page is something a startup can’t count on.

In the TripAdvisor example, that could be a banner ad that people either click or do not click. Not clicking is a neutral response: it shows that people aren’t even curious enough to notice the potential product. On the other hand, clicking doesn’t really tell you all that much. It tells you that people are curious, somewhat interested, or that they noticed the ad. It doesn’t tell you that they are likely to whip out a credit card and book some travel (Figure 8-2).

Figure 8-2. The result when I clicked the “read” button on a Virgin America in-flight entertainment console; unfortunately, there’s no way for Virgin to know whether customers secretly long for books or are just bored and pressing buttons

I’ll explain how important it is to choose the right customers by telling you just what happened to me when I chose the wrong sort of customers.

At Yodlee, I made the mistake of giving demos of new prospective product ideas to a few of our more conservative customers. I’ll let you in on exactly what we were thinking because it’s a really common misconception:

Those demos did not go well: the customers were so unreceptive to new ideas that they could not provide any useful answers. The implications for the relationship were even worse. These customers were afraid that we were going to turn our backs on the product they were using and getting value from. They were anxious that they were going to be stuck with a dead-end product while we directed all our innovation energy elsewhere. We spent a lot of time doing damage control afterward.

What worked better for me back then was to collaborate with our account managers. Which customers are comfortable with change? Which ones engage us frequently (even if they’re often complaining)? Which ones ask questions about our future product direction? I would then reach out to these customers, explaining that we were exploring a new direction and wanted to learn more from them.

Subscription billing provider Recurly taps into its customer support team’s insights in a similar way: “[Support] knows which customers are beta-tolerant and interested in specific features, so we can show them early prototypes and watch how they’re using the product today.”

Choosing customers based on our existing relationships was a more appropriate approach than choosing them based on account size or brand name. But the quality of an existing relationship isn’t the best metric to use. It’s also not applicable for many products and services—mass-produced physical goods, low-touch services, consumer software, and plenty of other businesses simply don’t have a regular communication channel with customers.

Your best customers are the people who are getting the most value from your product. Sean Ellis, CEO of Qualaroo, developed a simple set of survey questions to identify the most passionate users of his products. You can create your own version of the eight-question survey for free at http://www.survey.io, but the most critical question is this one: How would you feel if you could no longer use our product? (See Figure 8-3.)^[64]

Figure 8-3. The “how disappointed” question from Sean Ellis’ Customer Development Survey at http://survey.io

Find out who would answer “very disappointed.” Those are the customers you should reach out to and learn from.

Those are the people who are the most passionate, who consider you almost more of a partner than a vendor. (That doesn’t just apply to enterprise products; think of any friend who eagerly recommends his neighborhood butcher, his favorite energy drink, that amazing music-finding service.)

At Yammer, we use data to identify our most passionate users. Our analytics team segments the top 1% and top 10% of people who return and actively participate, and our user research team samples those populations to collect net promoter scores^[65] as well as to engage in customer interviews to learn more about how they use Yammer, what their work environments are like, and how we could make their lives better.

What’s the biggest risk to most products? It’s not that your product isn’t good enough: it’s that customers aren’t even interested enough to try it.

Customers can’t get value from your product until they use it. And given that you’re competing with a million other things, it’s incredibly hard to grab their attention. Your product might legitimately be the best, but what does that matter when every other product is making the same claim?

Customer development enables you to understand your target market’s pain points and come up with solutions. But that’s not the same as capturing the way your product makes customers feel.

Kathy Sierra talks about the important distinction between selling a great product and selling a product that makes great users. As people who build products, we’re naturally biased to want to talk about features, technical sophistication, and triumphs of design and engineering. What our customers want is to hear that they’ll feel smarter, get their jobs done faster, look better, feel less stressed, or get to replace an unwanted solution. When you echo those words back to other prospective customers, they get it (Figure 8-4).

The best person to sell to your customer is...your customer.

Figure 8-4. Your focus should be on making your customers feel awesome, not on telling customers that your products are awesome; http://headrush.typepad.com/

Customers are not accustomed to products changing direction. They’re accustomed to hearing vendors sell, not ask. This is doubly true in the enterprise software space. Unless you are extremely clear in your communications, customers will interpret your questions as a commitment to build something. They see demos as a window into what is coming, not an exploration of what is possible.

To understand the impact of the need to overcommunicate, consider what would happen if you were at your favorite pizza place and the manager came over to ask you about burgers. Lacking context, you’d probably give inaccurate answers. Perhaps you’d say no or rush through answers because right now, all you can think about is pepperoni and four-cheese. Or you’d answer based on your weekend situation (buying lunch for your kid’s soccer team full of picky eaters) as opposed to your weekday situation (working lunches with beer and premium burgers). Or you’d come back to the restaurant expecting to see burgers on the menu and be annoyed if they’re not served. Or you’d avoid the restaurant, assuming that the pizza quality will decline once the kitchen is split between pizza and burgers. The restaurant would need to explain why they’re asking, emphasize that they’re learning from customers, and clarify that no changes are being made yet.

Even if your hypothesis is not out-and-out wrong, it’s going to be significantly modified based on the real-world feedback you get during this process. The features you think are most important are probably not the most important to the customer. One customer will rave about a feature that others consider trivial. Your emphasis will inevitably change, and you definitely don’t want to make promises you can’t keep.

You also don’t want customers to feel that they should be polite. In many business and cultural settings, complaining is viewed as rude or unprofessional. As one Yammer user told me, “When I give feedback in surveys, I try to only say as many negative things as I can say a nice thing to balance it out.” (Yikes. Presumably the worse your product is, the less honesty you’ll get from her.)

What does that mean for your conversation with customers? It means you need to clearly communicate three things:

It’s important to state these points clearly and firmly from the beginning, and reiterate them throughout your conversations.

Here’s a template that I’ve used recently for Yammer customers:

I’ve emphasized some of the words here (not in my email) because I want you to see how I’m weaving in the expectation that this is exploratory by using words that imply a distant future. This serves as a second qualification; it weeds out customers who are motivated by asking for specific roadmap enhancements. It’s not useful to talk with customers who have an agenda to push for a specific feature.

Once a customer agrees to have an exploratory conversation, I start by mentioning that we extended this invitation only to a limited number of our most visionary customers. This sense of scarcity seems to encourage more specific, deep feedback. It also motivates people to refrain from sharing information with other customers or on social media. (Your lawyers will probably want you to get an NDA signed before you show anything to customers regardless.)

During your conversations, you’ll probably hear customers ask, “So when is this going to be available?” or “How likely are you to build this?” Don’t give a politely noncommittal response with room for misinterpretation. It feels a bit awkward at first, but you must clearly repeat that this is exploratory.

Henry Wei, Senior Medical Director of Clinical Innovation at Aetna, agrees. “We have to be very explicit and say ‘We’re trying to figure out what you really need. We’re not trying to rush this into existence if it’s not what our customers will get value from.’”

It may take some practice, but ultimately you’ll find that many of your customers will appreciate this style of discourse. Customers are often promised features and fixes. While they may hear promises from their vendors, those promises tend to fade after the sale closes. It’s rare that customers hear from a vendor who is actively trying to identify and address their problems.

Being invited to talk about their specific experiences drives a level of customer involvement and engagement that nothing else can touch.

Such listening also yields another gift you can give to your customers: the ability to know just a little bit about what their competitors are thinking. We commonly anonymize what we’ve learned and then share a summary with customers we talked to. (For example, “We talked to 22 customers, including 15 Fortune 500 companies, and the major concerns we heard were X and Y.”)

When it’s impossible for your company to have an untainted customer development conversation with your target customers, something needs to change. Either your company or your customer needs to be someone else.

Microsoft faced the problem of unavoidable bias when working on a total overhaul of Hotmail.com (now branded as Outlook.com). Once the leader in free email addresses, Hotmail had long been perceived as outdated, poorly designed, and even embarrassing.^[70] Microsoft as a company suffered from a different problem: so many people use the Outlook email client in their workplace that they were likely to resist any changes to that familiar look and feel. Approaching potential customers as either Hotmail or Microsoft was unlikely to get objective feedback. Instead, Microsoft user researchers removed all traces of Microsoft from their conversations and prototypes. They used market research firms to recruit participants who met the target customer criteria, from cities across the United States. The researchers simply described themselves as being from a company working on ways to improve email and asked questions about the ways that customers were currently reading, composing, triaging, and sending email.

Conducting customer development without revealing your company affiliation doesn’t need to be complicated. All you really need is a different domain name! If you contact people from an @someplausibledomain.com email address, it won’t even occur to them that you may be from another company.

Another way to do incognito customer development is to speak to people other than your current customers. While the best customer development interviews happen with your actual current or future customers, you can often learn a lot from the right proxies.

This may be a different stakeholder than your target customer, someone who lacks decision-making abilities but has a deep understanding of how things work in her particular environment—think administrative assistants, early-career product managers, or project managers. These people are easier to access, incredibly knowledgeable, and often underappreciated. You might even choose people from your own organization (if you haven’t mined your customer support staff for insights, put this book down now and do it).

Depending on your business model, you may also have target customers who are users but not buyers. For example, think about an ad-supported service. While working on an online finance application, I didn’t have access to the financial executives who were our target purchasers. But there was nothing stopping me from talking to their customers—the people who used their online banking and bill pay services.

Using Craigslist,^[71] I was able to recruit a dozen online banking customers for interviews and a couple hundred for survey research.

Through those interviews I learned how their banking customers were behaving and which online banking features they used. When we finally got the chance to pitch our target purchasers, we were able to say, “We’ve talked to your customers and we know what they need.”

I learned that 88% of the customers I surveyed who had supposedly switched banks were still paying bills out of their old bank accounts because the process of setting up bill-pay information online was so tedious.

From the consumer perspective, this was merely a nuisance: they still had to carry two ATM cards and remember two PINs. For banks, it was a huge financial loss. Banks try to woo you to switch away from their competitor, but they don’t make money just because you open a bank account: they make money when you keep most of your money with them. Banks want you to directly deposit your paycheck and then pay all of your bills from their account.

That opening guaranteed us a second meeting that we may not have otherwise gotten, and the insights from that research became a whole new product.

When multiple consumers told me the same problem, “I haven’t switched my bill-pay account because it took me so long to set it up the first time,” it was obvious that consumers and banks would benefit from a tool that automated the bill-pay account switchover. This insight became Yodlee BillPay Account Accelerator (http://corporate.yodlee.com/solutions-payments-suite-yodlee-billpay.html). While I am one of the joint patent holders for that product, I take no credit (or blame) for its horribly long name.