A successful subscription-based eCommerce company with several new customers every month may now find itself in a tricky situation - their growth has suddenly slowed down leading to a dwindling top line.
Naturally, the eCommerce company is now looking at ways other than new users to add to their revenue numbers.
One of the key ways of doing so is improving customer retention.
Customer churn analysis and trying to improve customer retention is not a new concept to most businesses. But when sales are dwindling, this often comes to the forefront.
The new Customer Insights Manager at the eCommerce company has been tasked with improving customer retention. Let’s call her Emily.
Emily found herself looking at a whopping 29% month-on-month customer churn rate.
The good part? She knew with such a high rate, there was bound to be room for fast improvement and a chance at real impact.
The bad part? She had no clue what those improvement areas were.
If you were Emily, what would you do?
In this article, we’ll look at how a Customer Insights Manager at a subscription-based eCommerce company went about reducing customer churn rate and improving customer retention numbers using customer sentiment analysis.
Improving customer retention is no small task. You need to continuously monitor what customers are saying and how they are feeling then act quickly on the pain points.
Surely the eCommerce company had some system in place to do this. Emily wanted to understand the current process.
This eCommerce company, like most typical ones, had the following process:
• Customers cancel out for some reason
• If done via customer service, the agents doe their best but end up tagging it broadly such as ‘delivery_issue’
• 2 weeks after cancellation, the churned customers get a 3 questions cancellation survey
• After the data was collected, it was analysed for reasons for churn
Emily analysed the current process and realised there were several shortcomings.
• The completion rate was a dismal 7% therefore the majority of the cancelled customers were not represented.
• The short survey with just 3 questions didn’t yield rich data with details and because the questions were predetermined, the churned customers weren’t free to share everything
The above 2 reasons led to very generic customer churn insights that couldn’t be used to truly gauge customer sentiments.
Now that Emily knew what the problems were she used this information to rectify this approach to customer churn analysis. Let’s look at that in the next section.
Emily realised to improve customer retention, she needed to look at 3 critical areas:
1. Customer data collection
2. Customer data analysis
3. Putting the insights to work
The bottleneck was the first step. If the eCommerce brand could gather data without having to rely on seeking feedback from churned customers, they could then come up with accurate, granular, and actionable insights.
Then they could address the pain points efficiently.
This is where customer sentiment analysis of customer service conversations comes into the picture.
Leveraging customer service data where your customers are sharing their unfiltered, detailed, and unsolicited feedback about what’s troubling them is the best way to understand customer sentiments.
With these in mind, Emily then followed a 6 steps approach to understanding and leveraging customer sentiments to improve customer retention.
Step 1: Adopt a holistic approach to customer data collection
One of the key steps to improving customer retention is to understand why customers are cancelling in the first place.
But the insights that typical post-churn analysis provides are incomplete and often do not provide the true picture.
Emily decided to include customer service data in churn analysis to uncover the true reasons behind the cancellation.
There are several reasons why customer service data can complement cancellation surveys to give you a holistic view and we’ve highlighted them in our previous article on customer retention here.
Emily's approach: The most common cancellation reason encountered by the eCommerce company was a discount error. By analysing customer service data, Emily found out that people mentioning discount errors also mentioned app crashing.
This means that it’s not the discount code that was at fault, it was a product bug that was causing the app to crash whenever someone tried to enter their discount code.
Step 2: Conduct a smarter customer churn analysis
Once you have all the data, it’s important to look at it through an objective lens. There could be several ways to dissect the data.
For Emily, maintaining and improving revenue numbers was the most crucial and therefore what made the most sense for her was to segregate the data into high, medium, and low LTV customers.
She then looked at what kind of problems each of these groups was facing separately.
Emily’s approach: By looking at the high LTV group, Emily realised that some prevalent problems for the other groups such as the product being expensive were not an issue for this group.
The major issues for them were 'allergic reactions to ingredients' and 'side effects'. So to avoid these customers from churning, Emily passed on this information to the marketing team who would then put out clear messages and guides on allergens and related side effects.
A point to note here is that your customer service data also needs to be tagged and categorised efficiently for it to provide solid insights.
Most companies use a manual system to tag tickets which may not often give the best insights.
While manual tagging can solve some issues, it might be best to look at automated systems to tag and categorise all your service tickets. This will ensure you can go as granular as you want with the data.
Step 3: Dig deeper into customer sentiments
A first-level customer churn analysis can give you some insights into customer sentiments, but often lacks the granularity that is needed to make business changes.
By digging deeper into those problem areas you can figure out the root causes and nip them in the bud.
Emily understood this is where marrying cancellation surveys with customer service data is crucial. By analysing post-churn data, you can uncover the overarching reasons for cancellation.
Then digging deeper into those reasons through customer sentiment analysis of customer service data can give you the true insights that will move the needle.
Emily’s approach: Emily found out that bad app experience was a key reason for churn from all the cancellation surveys.
But by digging deeper, analysing customer sentiment of related service data, she uncovered that what was leading to the bad app experience was a lot more granular - the app crashing when someone tried to make a payment over 100 pounds through a visa card.
Step 4: Find non-apparent patterns and trends
This step is all about taking your customer churn analysis to the level. This is where you need to be smart about how to look at the data. Each company can have a different perspective on this.
But coming up with possible hypotheses manually could prove to be difficult. One of the best ways of doing this is using an AI-based reporting system that automatically comes up with correlations.
One example of this could be correlating earlier customer service contacts with cancellations for every customer.
With a starting point at hand, you can then easily work with the data and slice & dice it to come up with new patterns.
Emily’s approach: Accurate, granular analysis provided Emily with insights like customers who complained about product quality in the first month and about delivery issues in the second month is 90% likely to churn by the third month.
And you can go as granular as you want.
Step 5: Prioritise and fix urgent issues and low-hanging fruits
Based on all the analysis above, the problem areas and pain points should be treated like a project.
With each having its score based on its prevalence, urgency, impact, and effort.
Emily’s approach: Take the same example from step 2. Fixing the messages that go on the website and educating the customers about potential allergens and side effects can help boost customer retention.
Step 6: Proactively prevent customer churn
The ultimate way to improve customer retention is to tackle customer pain points before they become reasons for churn.
With a strategic system of churn analysis in place, it is easier to find those areas and combat them before they become bigger issues.
Emily’s approach: One way she did this was by creating customised in-app messaging that goes out to each group of churn-prone users.
Another way was sending a special discount coupon, a surprise gift to users in their third month who were churn-prone.
If you’re in the same boat as Emily or are in a similar situation to the eCommerce company we spoke about, it’s time you focus on improving customer retention.
One of the major ways of doing it (that most probably none of your competitors is doing) is complimenting your post-churn analysis with pre-churn prevention via customer sentiment analysis of customer service data.
We’ve talked about the 6 ways in which you can go about decreasing the customer churn rate of your eCommerce business.
By changing the way the eCommerce company did churn analysis, Emily was able to find the real reasons behind the high churn rate, fix them quickly, improve the overall customer experience, and ultimately improve customer retention.