Oomiji’s Guide to Survey Design

Even if you’re not using Oomiji, you may still be conducting research among your customers. Customer insights are a real advantage for any company in a competitive market. Many companies ignore customer research and make assumptions about their customers based on anecdotes and internal bias. But anecdotes are not data and should not be used as such.

The advantage of Oomiji is that it combines survey research and NPS scoring with segmentation by emotions, motivations and perceptions, a much stronger tool to predict customer response than using closed-ended questions to predict future customer behavior.

Oomiji’s AI analysis, segmentation, and engagement tools are only as powerful as the questions behind them, and this, of course, is true for any research platform. Our guide gives you working examples organized by goal, along with notes on what makes each question effective and what to avoid. Admittedly, it is slanted toward Oomiji, as the platform can segment down to the individual using any survey question, including respondent language. However, if you’re not a professional researcher, we think you’ll get value out of this guide.

A note on question types: open-ended questions produce the richest language for AI analysis and segmentation. Scaled and closed questions are useful for benchmarking and sorting, but they are not substitutes for open-ended responses. Where both appear in a section, lead with the open-ended questions.

How many questions to ask?

Companies often have a long list of questions they want to ask their customers. On Oomiji, you can do that but don’t ask them all at once. People get survey fatigue at 8 to 12 questions. Asking three or four questions will get you a much higher response. An Oomiji advantage is that you can ask another three or four in the next month or quarter and continue to learn more about your customers. When composing questions, the question to ask yourself is: What information is most important to marketing and selling my product? If the information is of mild interest but not key to your sales, save it for another time.

Net Promoter Score Questions

See: Understanding Satisfaction and NPS in Oomiji

NPS questions are a great way to start and get you a surprising amount of information by using both Oomiji’s segmentation capabilities and one-click aggregate analysis. As NPS consists of only two questions, one closed-ended and one open-ended, you’re likely to get a much higher response, and you’ll learn perceptions of your product or service directly from your customers. It’s also a great way to get clues to attract more high lifetime value customers.

One-Click Aggregate Analysis

Both the Insights and Satisfaction sections of the platform offer One-Click Aggregate Analysis that will quickly analyze your survey results and provide descriptions of Key Consumer Segments, Trends and Patterns, Verbatim Analysis, and Actionable Recommendations. This feature will save you time and money in analyzing and interpreting customer insights for better planning. It will also provide you with the language your best customers use so you can use it to connect with potential high CLV customers on social media and websites.

Sample Questions

See: Choosing and Configuring Question Types

Our guide to different types of questions and their format is below. While you can change the language, try to stay within the guidelines to ensure that you get high-quality responses.

Question type key:

  • [open-ended]  Produces narrative language used for AI analysis and segmentation
  • [scaled / closed]  Useful for ranking, scoring, or grouping by preference
  • [NPS follow-up]  Paired with the NPS score question to surface the language behind the number

What to Avoid

The most common ways clients degrade their own data before analysis even begins.

Common mistakeWhy it matters
Leading questions“Don’t you think our service is excellent?” primes the respondent toward agreement. You get confirmation, not honest signal. Oomiji’s AI will surface the language you fed them, not what they actually think.
Double-barreled questions“Was our product high quality and easy to use?” asks two things at once. If the answer is yes to one and no to the other, the response is useless. Each question should have exactly one variable.
Too many questionsSurveys longer than 5 to 7 questions see sharp drops in completion rate. Oomiji tracks per-question completion, so you will see exactly where respondents drop off. Shorter surveys with better questions outperform long ones with mediocre ones.
Only asking closed questionsScaled and multiple-choice questions can tell you where sentiment lands, but they cannot tell you why. Open-ended responses are where the language for segmentation, messaging, and AI analysis comes from. A survey with only closed questions produces sorted data, not intelligence.
Vague or generic phrasing“How was your experience?” is so broad it produces low-effort, meaningless answers. The more specific the question, the more specific and usable the response. Replace category words (“experience,” “service,” “product”) with concrete references to what actually happened.
Asking what you already knowIf you already have demographic or purchase data on a customer, don’t use survey space asking for it again. Use surveys for what you cannot get any other way: motivations, perceptions, emotions, and intent.

A useful test before sending any survey: read each question aloud and ask whether a thoughtful person could give you a genuinely surprising answer. If the range of possible answers is obvious or narrow, the question needs work.

Whether you’re an Oomiji client or not, these questions will help you learn more about your clients. Look for a platform that appends every response you get to that respondents data file as you can then segment by virtually any factor that you’ve learned. It makes for a very powerful customer insights and engagement tool.

Let’s set up a demo to show you how the platform works and makes customer surveys, segmentation and engagement easier. Just contact us at info@oomiji.com or set up your meeting here.

Leave a Reply

Your email address will not be published. Required fields are marked *