Three Traps with Customer Surveys & How to Avoid Them

Surveys are a powerful tool to gauge customer sentiment, but most organizations fall short by failing to create mutual value when asking for feedback. Let's unpack why that is and how to avoid it with an example.

There are three common traps that organizations fall into:

1. Not providing an incentive in return for asking for a lot of the respondent's time

2. Asking questions that are adjacent to what they're really interested in

3. Making it cognitively hard to answer questions based on scales and selectable responses

Let's use an example with United Airlines to walkthrough each of these traps and how slight adjustments can add value to both the organization and the respondent.

• After each flight, United Airlines sends a survey via email to some of the passengers with the subject "Tell us how we did"

• The first question is presented in the email: "How likely are you to recommend United Airlines to a friend, relative or colleague?"

• Beneath this is a scale from 10 to 0 with qualifiers at either end of "extremely likely" and "not at all likely", and each number is clickable to start the survey with that answer selected

• After starting, there are 5-10 minutes worth of questions about the entire experience

Not providing an incentive
United asks for 5-10 minutes of a person's time. While this might seem trivial, it's enough time for a respondent to question whether it's worth continuing the survey.

This effect is compounded when the respondent receives a survey for a subsequent flight and might skip over the email entirely.

Here are a few ways you can avoid this trap:

1. Offer a universal benefit upfront in exchange for a complete response, which for United could be miles (their form of loyalty points) or entering into a raffle for a quirky prize (like their upcycled keyring made from parts of expired life vests)

2. Send a meaningful follow up to express gratitude for the respondent spending their time to answer questions, which for United could be delivered via email and app notification a day or two later (thus creating another positive impression of the brand)

Asking questions that are adjacent
United starts with a question asking about the likelihood of recommending them to a friend, relative, or colleague. Their survey continues to ask about practically everything in the customer experience starting with the booking experience all the way to baggage claim.

However, is this what they really care about?

Probably not. Instead, they're likely concerned with measuring loyalty and likelihood to purchase from them again, plus how their initiatives are being received by passengers.

Here are a few ways you can avoid this trap:

1. Ask what you want the customer to do next that aligns with how you deepen the relationship, which for United is likely something to the effect of "will your next flight be on United Airlines?"

2. Ask only specific questions that are sufficient to measure whatever you're attempting to assess about that experience, which for United could be aligned to a specific innovation or improvement initiative (e.g. the "gate agent to seat" experience) instead of everything they're doing

Making it cognitively hard to answer
United asks a question to assess likelihood of recommending them with a 10 to 0 scale. There are additional questions that follow this pattern deeper in the survey.

These are cognitively hard for respondents to answer unless they're choosing an extreme value (10 or 0), especially as anything in the middle (8 to 2 or 7 to 3) is entirely subjective.

Is one person's 8 the same as another person's 6?

Here are a few ways you can avoid this trap:

1. Provide answers as unambiguous phrases that can align to different numerical ratings or outcomes, which for United could be "I already recommended United" or "I will recommend United" for the positive outcomes to their first question

2. Use a smaller Likert scale or three options to capture positive, neutral, and negative responses; which for United are three options of "likely", "maybe", and "unlikely" with optional qualifiers of "very" or "somewhat" to offer a total of five options

TLDR: Offer an incentive upfront. Ask what you want the customer to do next and specifically about initiatives you're measuring. Provide relatable answers.