Table of contents
Ensuring that your customers are happy and engaged is critical to the success of your business.
There are countless statistics that back this claim:
The question is how to help your customers be successful so that they continue to use your product/service and buy more from you.
The answer: Build a customer feedback program—a centralized, systematic way to collect the voices and opinions of customers throughout the customer journey (in our case, using surveys), which, if acted on, delivers clients products and services that truly meet their needs.
Having a team focused on customer feedback may seem like a luxury, but you can build a Voice of the Customer strategy using limited resources. Moreover, practically any company can benefit from a centralized, systematized method of listening to customers. Why?
You know that each of your customers is unique—whether it’s in the types of products/services they use, how they use them, or who within the company uses them. Couple these differences with the fact that your customers are made up of a diverse group of people, and you’ll find that there’s no one-size-fits-all solution that supports each of their preferences and needs.
Think about your own customers. Do you know the following with absolute confidence?
According to McKinsey, measuring customer sentiment throughout the customer lifecycle is 30% more predictive of clients’ overall satisfaction than evaluating how they feel at a single point in time. Surveys are perfect for this because they give customers a quick, convenient way to provide candid and thoughtful feedback, and require little effort to send multiple times to even very large groups of people (more about this in the “Sending” section).
Once you understand the customer experience in more detail, you’ll be able to take steps that deliver clients more value. For example, you’ll be able to train customer-facing employees in the areas they need to improve on most; and you’ll be in a position to adjust your product roadmap to better meet the needs and wants of your customers.
And eventually, you’ll experience big business improvements:
Organizations that measure the entire customer lifecycle see a 20% increase in customer satisfaction and up to a 15% increase in revenue.Source: McKinsey & Company
We’ll walk through each step of building your own customer feedback program (using surveys) with expert guidance from SurveyMonkey’s Voice of the Customer team—a team that runs the customer feedback program for a platform that supports 95% of Fortune 500 companies, in addition to thousands of smaller organizations around the world. That experience has given them an understanding for developing and running customer feedback programs.
To help make our tips more relatable to your company’s situation, we’ll review, throughout the guide, how two types of organizations approach each step of building their own customer feedback program:
Choose the version that makes the most sense for your organization, and we’ll walk you through the process step-by-step.
Let’s begin with the most fundamental step: planning your customer feedback program.
Quick note: We’ll be talking about the Net Promoter Score® (NPS) in the coming section. If you’re unfamiliar with the term, it’s considered the world’s leading metric for measuring customer happiness and loyalty. And since it’s a foundational element of a customer feedback program, you’ll see it mentioned throughout this guide. If you want more context on the NPS, you can find an in-depth breakdown about it in the Design section!
If you’re interested in measuring customer satisfaction using another question, like the customer effort score, you’ll still be able to apply this guide’s framework for building your customer feedback program. Also, this guide serves as a great starting point for building an automated customer feedback program.
Net Promoter, Net Promoter System, Net Promoter Score, NPS and the NPS-related emoticons are registered trademarks of Bain & Company, Inc., Fred Reichheld and Satmetrix Systems, Inc.
Building an effective, centralized, and systematic customer feedback program from scratch isn’t a simple task. In order for your program to be effective, it must hold a central role in your company and wield enough influence to deliver insights and recommendations to teams that can put them to use.
That means you need goals, a strategy, partnership with key teams in your organization, and the blessing of your company’s leadership to build a truly effective customer feedback program.
In this chapter, we’ll walk you through how to plan a customer feedback program that:
If you’re going to justify the need for a customer feedback program, you should come up with a goal for your program that aligns with the goals of your company. It’s best to come up with a clear, attainable goal for your program before you do anything else. That way, it’s easier to zero in on focused questions that deliver actionable insights that will inform your goals.
To land on the right goal, you’ll need to assess the current challenges your organization faces. Then, reverse engineer those challenges into a clear-cut, realistic goal for your program. Here are some example scenarios, along with goals that match them:
Goal: Understand why customers are churning so you can take steps that will reduce the churn rate by 3% per quarter.
Goal: Achieve a relatively high NPS in your industry. This involves either using a competitive NPS survey to know what your rivals’ scores are, or using an industry-wide benchmark (if it’s available) to see where you stand.
Goal: To consistently improve your customers’ experience, you aim to increase the NPS of customers who’ve been with you for over 6 months by 3 points per quarter.
Keep in mind that if your goal is too generic, it’s likely to be less attainable and not provide enough direction. And if it’s too specific? It may not make enough of an impact to justify a centralized customer feedback program.
Here are some goals that don’t work—with the first goal being too general and the second being too specific:
Maybe running a customer feedback program is your full-time job; for most, including SurveyMonkey’s own Voice of the Customer team, it’s only one of several things you do. Either way, you’ve got to be realistic about the time and resources you have available and what effect you can really have. The impact you can make depends entirely on the size of your company, your customer base, and the resources you can put towards the program.
If you’re at a scrappy startup, there’s a good chance you’ve got several responsibilities in addition to your customer feedback program. And even if you don’t, there’s a good chance you don’t have a lot of time, which means you need to be super strategic with how you use it. You’re going to need to identify a goal and develop a strategy that’s lightweight enough for you to handle on your own. A scrappy startup should prioritize a single, attainable goal that’s absolutely vital to its success.
If you’re at a more established company, you’ve likely accumulated a larger customer base, and built a more diverse and specialized set of teams. All this is to say, things aren’t necessarily going to be easier. You’re going to want a more holistic customer feedback strategy that focuses on more than one area or challenge of your business.
The best thing about customer feedback programs is that the information you get from them can benefit practically every team in your organization. Here are just a few examples of the ways a well-run customer feedback program can help lift your entire organization up.
If your executives don’t consider your program to be mission-critical, you won’t get the resourcing you need to ensure its success.
To help leadership understand why your program matters, you’ll need to speak their language. You should already have an attainable goal that’s tied to your organization’s priorities, and a plan for how your program can benefit teams across your company.
Here are some customer feedback proof points that align with common program goals. Share these as part of your pitch to leadership, but if possible, complement them with research that’s specific to your industry. The more relevant your data is, the better it will resonate with colleagues.
When our Voice of the Customer team first began creating a customer feedback program, the team knew that doing so would allow us to better understand our customers and take steps to improve the customer experience. To get buy-in, they presented leadership with an NPS program that used persuasive research by Bain, including the fact that an NPS industry leader outgrows competitors by more than 2x. Though leadership was already keen on running a more robust NPS program, the research created urgency in getting the program up and running faster.
Over time, our Voice of the Customer team has been able to use our own survey data to show happier customers do, in fact, churn at a lower rate than unhappy customers. This has given the program even more credibility with executives and has allowed it to become an increasingly visible and important part of employees’ day-to-day work.
Now that you have your team together and the executives are bought in, you can choose the type of NPS survey to use.
But first, some background on the Net Promoter Score: The score is derived from the Net Promoter Score question, which asks, “How likely would you be to recommend [Company name] to a friend or colleague?” The respondent is then asked to give a rating on a scale between 0 (least likely to recommend) and 10 (most likely to recommend).
Your customers are grouped into cohorts based on their scores:
To calculate your NPS, take your percentage of Promoters and subtract it by your percentage of Detractors. To put it in its equation form:
NPS surveys can be divided into three groups: relational, transactional, and competitive. They each have distinct purposes, designs, and sending styles. Here’s a look at each of them:
Transactional NPS surveys are sent to customers following a defined action, and ask them to evaluate a specific experience. You can use this survey for anything, from following up on a phone call with a support representative, to getting feedback from a webinar your team hosted.
A transactional NPS survey aims to understand a specific part of the customer experience. And pinpoint how, exactly, that experience is performing over time.
SurveyMonkey uses transactional surveys at different stages of the customer engagement. For example, when the customer is using the product, they can click a button to provide feedback via a survey. The survey they receive depends on whether they’re in the process of creating a survey, setting up their collector (how they’ll gather their responses), or analyzing their feedback.
Try a transactional NPS survey if you…
But don’t use it when you…
Relational NPS surveys consistently collect feedback from a sample of customers (who can be chosen at random but shouldn’t include any one customer more than once per quarter) to get a holistic understanding of the experience. They can be sent to a fraction of customers on a weekly, monthly, or less frequent cadence. While relational surveys themselves may be low effort, keep in mind that certain parts of your program, like follow-ups with individual customers and the quality of your reporting, can demand the most time and resources from your team.
A relational NPS survey aims to understand the overall customer experience. So it tends to focus on broader topics, like the customer’s level of satisfaction and the general quality of your product/service.
Use a relational NPS survey when you want to…
And avoid it when you’re looking to…
Pro tip: Don’t try to launch transactional and relational NPS survey projects at the same time. Instead, choose one to get started with. Once the first survey is running smoothly, you can carry your learnings over to the next NPS survey.
Competitive NPS surveys involve getting feedback, anonymously, on you and your competitors from people in your target market (individuals who match the criteria of your ideal customers). Collecting the feedback anonymously limits bias in your responses—allowing you to accurately evaluate how you stack up against the competition. Similarly, surveying the wider market, and not just your own customers, helps you get a fair assessment on where you stand.
Note: This survey template only allows the respondent to evaluate a single competitor. If you’d like the respondent to potentially evaluate an additional competitor, copy page 2 of the survey and edit that page’s copy with the brand’s name (you can repeat this process for however many brands you want to add). Finally, to ensure that respondents only evaluate the competitors they use, incorporate skip logic on question 3.
Use this type of NPS survey to…
But keep in mind that…
With these 3 types of NPS surveys in mind, which would our scrappy startup and established business choose? Let’s find out!
The scrappy startup is focused on improving customer retention. Since the factors that influence retention vary from client to client, they use a relational NPS survey to see which clients (especially their Fortune 500 accounts) are at risk of churn.
The more established business knows that once clients reach a certain milestone in their level of engagement, they’re more likely to buy more from the business. As a result, the business uses a transactional NPS survey to clients who reach that milestone. The survey helps them see which clients are, in fact, ready to spend more and which aren’t.
The process of creating your survey isn’t always straightforward, and it should be a team effort. In this section, we’ll break down how you can work on the survey with colleagues, the types of questions you can ask, and how to ask your questions so that you get honest and thoughtful responses.
Imagine that you’ve finished your survey and feel prepared to send it—only to have colleagues request last-minute edits! This can really throw off your timeline. Before we even get into how to create your survey, it bears mentioning that it’s a good idea to involve your colleagues in its design.
It not only stops them from giving feedback at the final hour, but can also help you improve the quality of your survey and ensure you’re asking all the critical questions.
As a side benefit, your colleagues can spot and fix the small things. Whether it’s related to grammar, the wording of your questions, or missing answer choices, a pair of fresh eyes can be all it takes to fine tune your questionnaire.
When considering who to partner with, keep the following in mind for each team:
Once you’ve decided who to collaborate with, here are some different options for sharing your survey:
Once you know who you’ll work on the survey with and how to collaborate on it, you can begin to brainstorm its format.
So how long do you want your customer feedback survey to be?
In general, your NPS survey should be relatively short (think 5 questions or shorter). You’ll only want to ask what you really need to know and are prepared to take action on.
Anything that’s non-essential isn’t worth taking up your customers time—or your own when you review the responses.
Furthermore, because your survey is likely to go out multiple times a year (more to come on how to send your survey later in this guide), making it any longer than it needs to be increases the likelihood that customers burn out from taking them in the future.
Here are the different lengths that your relational and transactional NPS surveys can have:
The 2-question (transactional or relational) NPS survey: This survey keeps you ultra focused on the NPS, and it can be used as a relational or as a transactional touchpoint. It just includes your NPS question and a follow-up open-ended question that asks why they chose that rating. Here’s an example of how it can look:
Transactional NPS survey: This type of NPS survey builds on the 2-question version by asking the respondent a few more questions (up to 5 questions in total). These additional questions should be straightforward to answer and should center on the event that triggered the survey to be sent.
Pro tip: Keep your survey focused on your primary goal. If you need additional data, you can get it by sending an additional survey. Alternatively, you can try collecting the data through more personal means, whether it’s interviewing customers directly or receiving feedback from your customer-facing colleagues.
Relational NPS survey: This type of survey also includes a few follow-up questions. However, unlike the transactional NPS survey, these questions include a brand attributes question and a “key drivers question”—a matrix question that’s a bit more complex for respondents but effective in identifying the factors that drive your NPS either up or down. Because of the value key drivers bring to your analysis, the relational NPS survey is the gold standard option for tracking changes in your NPS.
Need help deciding which NPS survey length to use? Let’s find out what our startup and established business would do.
The scrappy startup is keen to understand why customers are dissatisfied (and eventually churning)—so they’re set on using a relational NPS survey to find out. They’re particularly interested in asking a key drivers question and a brand attributes question to understand the root causes of their NPS, but given their limited resources in managing the responses they collect, they refrain from asking additional follow-up questions.
The more established business wants to decide which customers are targets for sales upsells and discover any that should be spending more on a given product/service. Since they have a team that can manage responses to several questions throughout the customer journey, they decide to run a series of transactional NPS surveys at different customer touchpoints to find out. But they’re also interested in tracking customer sentiment about their company and comparing it to their competitors. Since they’ve got more resources available, they can run all three NPS surveys.
Regardless of your NPS survey’s length, always ask the NPS question at the very beginning. Otherwise, your NPS data can be influenced by the format of the survey.
Why? Because your customers might get grumpy when they realize how long your survey is taking them, or an earlier question may jog their memory of a specific experience with your team. In either situation (and in whatever other ones might arise), the respondent might be influenced to answer your NPS question in a certain way.
In case you need more convincing, including it at the beginning will lead to more completed NPS responses than asking it later on, as some respondents may drop out of the survey early.
On one of our long-standing transactional NPS surveys, our Voice of the Customer team realized the NPS question was in the middle. After moving it to the very beginning to be consistent with other surveys, our average score on this survey increased by 10 points!
It’s worth noting that since a slight difference can make one survey incomparable to another, we needed to restart our trend analysis once we reordered the NPS question.
As you write your questions, you’ll want to keep them straightforward and unbiased. Otherwise, the integrity and value of your customer feedback gets compromised. Here are 5 tips and tricks to ensure that you can collect the responses you need:
The NPS question is crucial to evaluating customer satisfaction and engagement, but without direct context that explains the rating, it can easily be misunderstood. To avoid any confusion, ask a neutral, follow-up question that looks something like the following:
Once the responses come back, you’ll be able to understand how your Promoters, Passives, and Detractors really feel (more to come on this soon).
Even better, ask 5 questions. An NPS survey should be less time consuming than other customer surveys that are sent one time and focus on areas like customer research.
Keeping your NPS survey short and to the point forces your team to prioritize the questions you want to ask most—based on the response data you hope to track and act on over time. And for customers, it allows them to put more thought into answering the questions you include.
Also, consider that customers don’t have to respond. Any feedback they offer is a gift to your team. A short, simple survey honors that gift and allows the lines of communication to remain open in the future.
That way, you can use skip logic (a feature that determines the question or page a respondent sees next based on how they answer the current question) to personalize your open-ended follow-up question based on their score.
For example, if the respondent is a Promoter, your skip logic can take them to an open-ended question that asks, “Please help us understand why you chose the score above” And if the respondent is a Detractor, your skip logic can take them to the question, “How can we improve the rating you gave us?”
Also, once a respondent goes on to the next page of your survey, their responses from the previous page get saved automatically. This means that if a respondent doesn’t feel like explaining the reasons behind their rating and leaves your survey, you’ll still know their NPS.
Note: Since this tip involves asking different questions to respondents (based on their NPS rating), you’ll need to separate your analysis of the open-ended feedback by Promoters, Passives, and Detractors as opposed to looking at all of the responses together.
To clarify, the only one we recommend you include is the one that follows your NPS question. Here are some reasons why:
Typing in answers to free-response questions on mobile is extra time consuming, so it can easily lead to more people leaving your survey or providing shorter responses.
If each survey is time consuming to respond to, your customers can eventually burn out from taking them all together.
Pro tip: Make your free-response question (and all of the follow ups to your NPS) optional. This helps you respect your customers’ time, and it also makes them less likely to leave your survey.
These include details about the customer’s personal background, their company, and your engagement with them. However, only ask a demographic question if you can’t easily retrieve the data ahead of time and only ask questions that are critical for your organization. If you already have access to the data, you should include it as custom data associated with your contacts in SurveyMonkey. Either of these options let you include demographic information in your survey results and allow you to filter by them to see how specific groups of respondents answered your other questions.
When it comes to best practices in writing surveys, our survey experts have a lot more helpful advice. If you’re still curious, check out these 5 features that can help make your next survey a success.
The system you set up for sending your surveys can be one of the trickiest aspects of running a customer feedback program.
You need enough responses to be sure you’re getting a clear signal and that the patterns you’re seeing in your results are actually real. On the other hand, the worst thing you can do is over-survey your customer base. Sending too many surveys can easily lower your response rate—impacting the validity of your results.
In this section, we’ll cover how to walk this fine line by strategically planning:
Deciding the precise number of customers to survey and the exact number of times to survey them depends, in part, on the type of NPS survey you’re planning.
Transactional surveys: Since these surveys are triggered by customer actions, you have less control over how many each customer receives.
That means you have to be thoughtful about which customer touchpoints trigger a transactional survey, so you can space them out throughout a customer’s journey. If you notice that your transactional touchpoints lean heavily toward a certain stage in the customer journey, try to replace a few of them with high-value touchpoints at other stages.
Relational surveys: Unlike transactional surveys, the number of customers who receive relational surveys is completely within your control and requires some reflection.
Start by looking at the size of your customer base, since relational surveys repeatedly sample portions of that customer base over time.
You want to make sure you get enough responses that you can trust your results, while at the same time ensuring you don’t send a customer your relational survey more than once per quarter. That means you have to calculate the number of responses it would take for your results to be statistically significant (or at least meaningful enough to draw conclusions from) with the confidence level and margin of error you want. In other words, you need to find your ideal sample size.
Note: Margin of error represents the range that your population’s responses may diverge from your sample’s, while confidence level is the likelihood that your sample accurately reflects the opinions of your population.
Once you’ve landed on a sample size that matches your priorities, you can decide how often you’ll send your surveys. We recommend sending your relational survey no more than once a quarter, per customer, to prevent overwhelming them.
Let’s imagine two different scenarios for your organization and see how you’d approach surveying customers in each:
Using the numbers above, your minimum sample size is 379 customers (we used the sample size calculator to arrive at this number).
Let’s say your response rates from previous surveys are around 20%. This mean you’ll need to email at least 1,895 customers (379 x 5=1,895) to get the sample size you need.
Since you’ve got a relatively large customer base, you’ll be able to run your relational NPS survey many times throughout the year without surveying any one customer more than once per quarter.
Using the numbers above, your minimum sample size is 278 customers.
Like before, let’s say that your response rates from previous surveys are roughly 20%. So you’ll need to email at least 1,390 customers (278 x 5 = 1,390) to get the sample size you need.
Since 1,390 is greater than the size of your customer base, you’ll need to raise your margin of error to land on a more realistic sample size. It also means that you’ll have to send your relational NPS survey fewer times over the course of the year, perhaps only 2 or 3 times per year.
The takeaway from these scenarios is that, all else equal:
The more customers you have to survey, the more times you can send your relational NPS survey without reaching each customer more than once per quarter.
Here’s how SurveyMonkey’s Voice of the Customer team runs their relational NPS survey:
Since SurveyMonkey works with hundreds of thousands of customers, our relational NPS survey goes out monthly to a small fraction of those who have been active on SurveyMonkey over the past 365 days. This lets us evenly space out when each customer receives the survey (no more than once per quarter), and gives our team a steady and manageable level of feedback to address.
Competitive NPS: For competitive NPS, you’ll use a survey panel (a collection of pre-screened individuals who are willing to provide feedback through a survey) to anonymously reach individuals who fit your target audience criteria. Surveying your target market as yourself can bias your responses, as would surveying your personal contacts or customers who are likely to be more familiar with you than with your competitors.
It is ideal to collect responses from 1,000+ individuals. That way, you can segment your respondents based on demographics and still have a statistically significant sample size. And as respondents likely won’t be familiar with each brand you include, surveying more individuals lets you reach a statistically significant sample size for each brand.
If responses from 1,000+ individuals ends up being too expensive for you, start with at least 400 respondents.
Note: If you or a competitor are new to the industry, you’ll need to take the results with a grain of salt, as your respondents may have nothing to base their opinion on. Consider running a brand awareness survey on yourself or on competitors using SurveyMonkey Audience instead. It can give you a more accurate sense of how popular you or your rivals are, and with our customizable template, you can create your survey quickly.
As we mentioned before, it’s a tricky balance between surveying too often and too infrequently, but getting the balance right makes a world of difference.
To help you land on the right number of times to send your survey, see the guidelines in the chart below.
|NPS survey type||Survey frequency|
|Relational NPS||Ideal to survey from a group of customers weekly or monthly, but every 3-6 months can be sufficient.|
|Transactional NPS||Send your survey either right after the triggered activity or following a predetermined delay in time.|
|Competitive NPS||Best to survey every 3-6 months to account for seasonality and changes in the industry, but 1-2 times per year may be enough.|
Deciding how you send your survey depends on the type of NPS survey you’re running:
Sending your survey on another channel (e.g. embedding a survey on your site) can bias your responses based on the customer’s experience on that channel.
If you use our email collector, keep the following tips in mind:
Transactional NPS survey: When possible, send it on the channel the customer interacts with you on. Otherwise, send it via email. Here are some common triggers:
Take your customer feedback program to the next level by integrating SurveyMonkey with your customer relationship management (CRM) tool. The integration lets you automatically: send your transactional or relational NPS survey, put the responses back into your CRM, and notify the appropriate colleagues of the responses immediately.
Competitive NPS survey: Survey your target market using a survey panel.
Here’s where SurveyMonkey Audience can help. It allows you to specify your audience (e.g. job title, location, age, etc.) and define the number of people you want to reach out to. It then collects survey responses for you in just a matter of days.
With all of the great customer feedback you’ve collected, it’s time to make sense of it.
No matter which type of customer feedback survey you run, the first thing you’re going to want to do is check your NPS.
Your NPS is your highest-level indicator of whether things are going well, or poorly. It’s also the starting point for what you do with your results next.
See something alarming that requires immediate action? You can address it right away. Want to start digging deeper into your results? Your NPS will provide the first clue on where to start.
Once the responses start coming in, your NPS should be the first thing you see when you click the Analyze Results tab of your survey.
After viewing your overall NPS, check out the number of Promoters, Passives, and Detractors you have. This breakdown drives your score and helps you understand the distribution of customers who are loyal, content (but unenthusiastic), and at risk of churn.
Once you’ve seen your NPS and the breakdown of Passives, Neutrals, and Detractors, you should have a pretty good idea of whether anything is amiss in your results. Not every minor change in your NPS warrants immediate action, but here are a few things that do:
Alternatively, you can take a more granular, systematic approach by basing your red-flag strategy on the individual responses. This means defining the range of scores that require an immediate response, and who should address them.
For example, one global demand generation company escalates every score that’s below a 5 to their VP of client success, while a 5 or 6 rating is immediately shared with the client success manager.
Whichever emergency plan you decide on, skip immediately to chapter 6 of this guide when you’re ready to take action.
You can think of the open-ended responses you put after your NPS question as an opportunity to spot what’s driving your NPS up or down. One convenient way to look at these responses is to use a word cloud in the Text Analysis feature. This feature picks up on the key words or phrases that appear most often in the responses, and shows how often they each appear by their relative size.
Word clouds are mainly useful for presenting your results, but they can also give you ideas for where to focus your analysis.
The words and phrases from your word cloud can carry different meaning, depending on the type of NPS survey you ran:
Assuming you didn’t use the 2-question NPS survey, you’ll want to see how customers answered your additional closed-ended questions (questions that include answer choices, like multiple choice questions). The answers to these questions can help explain why customers answered the NPS question the way they did, so it’s critical to match them up with your Net Promoter Scores to see if they’re related.
Charts can help you visualize the results. They summarize the data in an easily digestible way and help you pick up on key patterns more quickly.
Say you use a multi-select question to ask respondents what they like about using your product. A bar chart, like the following, could help you make subtle comparisons across your answer choices: