Having a solid grasp can make or break your mission, and failing to do so can create problems with innovation. Australian-based shoe e-retailer Shoes of Prey had to learn this lesson the hard way. Once a darling of the investment community, Shoes of Prey collapsed earlier this year when it discovered a large gap in customer's intent and their actual behaviors.

After conducting market research, company leaders concluded that mass-market customers would respond well to shoe customization, an offering that many heralded as innovative. Unfortunately, the mass market didn’t bite. As CEO Michael Fox wrote in a Medium post, “While our mass-market customers told us they wanted to customize … what they were consciously telling us and what they subconsciously wanted … were effectively polar opposites.”

Undoubtedly, thinking like users is one of the best approaches to entrepreneurship. However, it’s often difficult to determine how to begin, particularly when you’re also dealing with innovation problems. Here are three steps to better understand customer needs.

1. Follow up on outlier answers.

Just because you haven’t heard a specific answer doesn’t mean it’s invalid customer input. Expand your understanding of users not just through interviews, but by observing their behaviors, too. In addition, strive to understand the context in which your product is most and least effective.

2. Get “real” with objective tools.

It’s too easy to become anchored in your assumptions when you’re close to the product you’re designing, so it’s critical to create some distance in the validation process. This requires a vigorous understanding of both what you’re trying to understand and how you’ll carry out processes. These elements really matter.

Explicitly document your assumptions early on. This allows you to challenge them as you continue to research and work through innovation issues. That process can also help cultivate the habits of awareness that will allow you to ask questions with some objectivity, as well as talk less and listen more.

3. Experiment with framing and note the results.

Frame research questions in multiple ways to see how it might affect user feedback. In a 2008 post-presidential election poll, Pew Research Center found that questionnaire design impacted answers immensely. Pew asked all respondents: “What one issue mattered most to you in deciding how you voted for president?” However, some received a set of multiple-choice answers to choose from, while others had to type their answers into empty text fields. When the economy was listed as a multiple-choice option, 58 percent of respondents selected it as a key determinant, whereas only 35 percent of respondents who received the open-ended question volunteered the economy as an answer.

Any researcher can achieve similar insight through similar methods of framing. For example, try restating your question in a negative way, changing labels from percentages to actual values (e.g. “25 percent of people” versus “four people”) or reordering questions randomly across a large sample of users to wash out anchoring effects.