Customers are getting harder to please. According to HubSpot:

  • 55% of consumers in the US and the UK don’t trust the companies they buy from as much as they used to
  • 65% of them don’t have faith in company press releases
  • 81% of them prefer to get advice from friends and family rather than a business

It’s also getting tougher to acquire customers. ProfitWell found that overall customer acquisition cost (CAC) is up almost 50% during the past 5 years, so it’s crucial to keep as many current customers as possible. That’s where customer satisfaction comes into play, because it’s typically cheaper to retain a customer than acquire a new one. High churn rates have toppled many companies.

Luckily, you’ve probably already invested in a communication medium that offers plenty of ways to track customer sentiment: email. You can use email to get a handle on how people feel about your company and its products and services, and then you can tweak your triggered and promotional emails accordingly, along with your products and services and many other things.

When you improve your customers’ satisfaction, you not only grow near-term sales but you create advocates who will recommend your company to their friends, family, and colleagues. That’s a key element of evangelism marketing, in which your customers preach the gospel, so to speak, of your business to others.

Here are 3 steps to using email to help improve customer loyalty and reduce churn.

1. Deliver 1-question surveys via email to calculate NPS, CES, and CSAT

You’d be surprised what kind of information you can glean from 1 question, especially since the low level of effort means customers are much more likely to respond than if you ask them to fill out a multi-question survey. There are a variety of options for 1-question surveys, but 3 of the most popular ones measure different aspects of customer sentiment:

  • Net Promoter Score (NPS)
  • Customer Effort Score (CES)
  • Customer Satisfaction Score (CSAT)

Net Promoter Score (NPS) is a tool that helps you determine the loyalty of your customers. It ranges from -100 to +100, but any score above zero is considered good, while +50 is seen as excellent and +70 means your most of your customers aren’t leaving for the competition any time soon.

Calculating NPS is fairly simple: You ask your customers, “On a scale of 0 to 10, how likely are you to recommend our [company/product/service] to others?” You can hone in on a specific product or service, if you’d rather measure customer loyalty to it, rather than your entire company. All answers between 0 and 6 are detractors, 7-8 are passives, and 9-10 are promoters.

Toss out the passive answers and subtract the percentage of detractors from the percentage of promoters. For example, if you surveyed 1,000 customers and the passive/detractor/promoter percentages were 15%/31%/54%, your NPS would be +23. Track your NPS over time to see how it changes as you tweak your email strategy, roll out product updates, and make other decisions.

Here’s a good example of an NPS survey from Insurify, an auto and home insurance comparison service:

Any 1-question survey you send by email should be simple like this one. Your goal is to get them to click a response and give you a result, so there’s no reason to muddy the waters with cross-sells, upsells, and other promotional messages. Note that Insurify uses a color range to help cue what the numbers mean, and a simple image reminds the recipient what the company does.

You can take a similar approach when sending Customer Effort Score (CES) and Customer Satisfaction Score (CSAT) emails.

Customer Effort Score (CES) measures how customers feel about their experiences with a product or service, including support interactions. They typically choose their responses on a 7-point scale from Very Difficult to Very Easy, although you can also use approaches like “[Product Name] met my expectations,” with a 7-point range from Strongly Agree to Strongly Disagree.

The scale might be something like this:

  • Very Difficult
  • Somewhat Difficult
  • Fairly Difficult
  • Neutral
  • Fairly Easy
  • Somewhat Easy
  • Very Easy

You can also try a scale of 1-5 that drops the “Fairly” responses or a 3-point scale where the choices are Difficult, Neither, and Easy. Those might be better approaches if you don’t get a good response rate from a 7-point range, which could signal that your customers are having trouble parsing the difference between answers like Fairly Easy, Somewhat Easy, and Very Easy. If they struggle with deciding how to respond, they could decide not to bother.

Customer Satisfaction Score (CSAT), as its name implies, measures customers’ satisfaction with a product or service, including support interactions. Like CES, you can use a 5- or 7-point range from Very Unsatisfied to Very Satisfied (or variations on that, such as Awful to Great), as well as a simple 3-point range: Unsatisfied, Neutral, and Satisfied.

There’s no industry standard for measuring CES and CSAT, but the conventional wisdom is that you should average the responses and/or look at the median score, if you’re worried about very low or very high scores skewing the average. Anything above a 5 on a 1-7 scale, or above a 4 on a 1-5 scale, means you’re on the right track.

If you use a 3-point scale, you might want to take an approach similar to the one NPS uses: toss out the 2s (Neutral) and subtract the percentage of 1s (Difficult) from the percentage of 3s (Easy) to create a range of -100 to +100. As with NPS, a score above zero is good, +50 is excellent, and +70 or more is “Wow, people are thrilled with us!”

2. Use Sentiment Analysis to track customers’ tone in the emails they send to you

While many people like to interact with businesses on social media, they still use email and are likely to communicate with your business that way too. However, it can be hard to decipher how they feel, especially if a message is nuanced and people in your company who read it may disagree whether it’s net positive or net negative. And sarcasm, such as “Wow, great job taking a week to send my order,” can sometimes be hard to figure out.

That’s where you can use Sentiment Analysis, which uses Natural Language Processing (NLP) to process the raw text in the emails sent by your customers. Then it uses Artificial Intelligence (AI) to add a layer of emotion decoding, tone determination, and overall sentiment output. It’s useful for giving you an objective view of how your customers feel, without the bias people naturally bring to any situation.

You should use Sentiment Analysis in conjunction with survey emails to determine how your customers currently feel about your company overall, or a specific product or service, and then track the changes over time. Sentiment Analysis is a good way to add nuances to the data produced by surveys – for example, people may seem happy overall based on their survey responses, but a deeper dive might reveal a specific trouble spot to hone in on, such as lengthy shipping times.

Admittedly, Sentiment Analysis is much trickier to implement than survey emails, so you should do plenty of homework and spend time researching the available tools before creating a proposal for your engineering team.

3. Decide when to send surveys and how to act on the results from them and Sentiment Analysis

Create a strategy for sending your 1-question NPS, CES, and CSAT surveys, which should be triggered by certain actions, such as:

  • Key milestones
    • Time elapsed since sign-up (30 days, 60 days, etc.)
    • Actions taken in an app (the Xth time they’ve done something, for example)
    • User onboarding completion
  • Shortly after an order has been received, to give them time to form an opinion about a product
  • Immediately after a customer support or other 1-on-1 interaction, when their feelings are still fresh in their minds
  • When customers are considered churned, either by canceling their accounts or lapsing their activity for a certain period of time
  • When customers’ status changes in some meaningful way, such as on-and-off customers becoming more regular in their patterns or vice-versa
  • Just before and right after a renewal date, or when you issue a major release or announce important news

If you use multiple survey metrics, each one should have a specific purpose. While you can, for example, use NPS to gauge how a customer felt about a support experience (“How likely are you to recommend our company to others after your customer support call today?”), you should maintain consistency and always use it that way. One way to look at NPS, CES, and CSAT is like this:

  • Send out an NPS survey once or twice a year to get a big-picture view of how customers feel about your company.
  • Send a CES survey when a customer takes an action that can be a pain point for many people, such as calling your customer support phone number.
  • Send a CSAT survey to see how customers feel after receiving a product, or after you release a software update.

As you begin tracking survey scores and potentially adding Sentiment Analysis to the overall picture, decide how you will respond to your findings, such as:

  • Rewarding happy customers, and reactivating churned ones, with incentives or other kinds of offers
  • Addressing a thorny problem area with an announcement: “We’ve heard your dissatisfaction with our shipping times and we’re pledging to reduce them,” for example
  • Handling different customer cohorts in specific ways: For example, you might find that customers who live in certain areas have different needs than customers in other areas
  • Making changes to products and/or services
  • Altering your strategy for triggered and/or promotional emails – your customers may feel spammed, they may want to hear from you more, or perhaps you’re missing an opportunity to email them at a crucial juncture in their customer journey

If you see discrepancies in the data, such as strong NPS and CES numbers but weak CSAT scores, probe deeper to understand why. You might want to send follow-up surveys to customers who answered in seemingly contradictory ways. It’s possible that they are happy with your company except for a specific thing that they’re unsatisfied with. Then figure out a way to fix the problem with that cohort.

Over time, you should see an increase in customer loyalty and a decrease in churn. If not, you may want to try an in-depth survey with an incentive attached (a chance to win one of several gift cards, for example) to get at the root of the problem.

~ Erica