How to convert your first-time buyers into loyal customers with data
I don’t think I need to tell you how important high customer lifetime value (CLV) is for you when building a profitable business - but few companies are using a systematic, data-driven approach to improve their CLV.
Last month, we started an article series to help you establish a data-driven framework in improving CLV, to help you use data you are currently collecting within your organization to systematically increase and expand your CLV.
Our framework divides the generation of customer lifetime into two stages - the user’s acquisition phase, and the user’s loyalty phase.
In the previous chapter, we discussed the first phase of the customer lifetime value, which is the user acquisition phase.
In this phase, our discussion focused on optimizing our ads and sending them to the right people to maximize the acquisition of new users into online stores and generate purchases, providing value for your business.
Next enter another important phase of customer value, the loyalty phase, which is drastically different from acquiring customers because your interaction with your customers is much longer, and not as linear as what you would get from simply acquiring them.
To make sense of this messier phase, let’s begin by exploring several crucial areas of customer retention, along with key metrics that you should establish in each of those stages.
Before proceeding, I highly recommend you reviewing our cohort analysis article, as data analysis for post-purchase behavior is almost all cohort-based, and having a basic understanding of the inner working or cohort analysis is crucial for the full understanding of the content to come.
Overview - The retention stage
The retention stage is arguably an even more important staging than the customer acquisition stage described in the previous article, even though most companies are spending considerably less time on it.
The reason we are making this argument is because scaling up your retention impacts your bottom line much more significantly than scaling up your new customer acquisition efforts.
This efficiency stems from the fact that most of the instruments you use in this stage, such as email marketing, referral platforms, retargeting ads, are significantly more cost efficient than sending massive prospecting ads to get your initial customers through the door.
Furthermore, there is a hidden “customer referral” factor that we also need to consider in this stage, which drives more new, highly qualified customers through the door at almost no additional cost.
In our framework, we have identified two primary sources of value from your customers those produced by your customers through:
Making more purchases
Referring other customers through the door.
Let’s begin by discussing the first source of value - those produced by customers through more purchases.
The core metric we use to represent this source of value is the customer lifetime value or CLV.
Even though technically, we are primarily discussing the part of your customer CLV that is beyond their initial purchase.
The measure itself is nevertheless one of the best and simplest measurements for the value your customer have contributed to your company over an elongated time period (In fact, most of the metrics discussed in this article are going to include that first purchase component, as it is much easier to compute).
To deepen our analysis of CLV, it can be further divided into two components:
Purchase frequency
Average order value
As computed with the simple formula CLV = purchase frequency X average order value, both are incredibly important when it comes to creating an exponential increase to your CLV.
To compute purchase frequency, customer retention rate and average orders per customer are two of the most frequently used metrics.
Customer retention rate helps you understand whether your customers are coming back in the first place, and average orders per customer help you understand how many times on average your customers are coming back.
As for average order value, common metrics include average order value (with particular focus on how it changes within a customer’s lifecycle), and the amount of revenue generated by your customer beyond their first order.
Computing average order value for all of your customers helps you understand if your customers are buying more or less per order after your initial purchase.
And the amount of revenue generated by your customers beyond their first order gives you a proportional idea of how much of your growth is purported by your loyal, repeat customers.
Going onto the second source of value, which is customer referrals.
Everyone in the eCommerce industry understands that word of mouth promotion and user-generated referrals are probably the most effective way to generate additional revenue and loyal customers.
But few are measuring it in a systematic way.
The metric that represents your customer referral effectiveness is often convoluted to compute and elusive due to its complexity of computation.
Ideally, we would want to compute the CLV of customers referred by all of existing customers and compute for metrics like “average referral value”.
However, metrics like that require not only a decent amount of data science skills to compute.
It also requires you to establish a referral system in which you can identify customers referred by another - both are rare in eCommerce companies that are under $100M annual revenue.
With that said, it is not hopeless for us to measure referral efficiency. There are simple secondary metrics that are much easier to compute, the most popular being the “net promoter score”.
Net promoter score is the average of answers to the following question in your customer survey: on a scale of 1-10, how likely are you to recommend this product to someone else?
While not directly measuring the referral “action”, it is a very good proxy of the “tendency” your customers are going to refer your product to someone else.
To summarize this section, in the customer retention stage, there are two primary sources of value: those from the customers themselves, and those from their referrals.
To measure value created by customers themselves, we use CLV as the primary metric and average order value & order frequency as secondary metrics.
To measure value created by customer referrals, we can either use primary measures such as “average referral value” if your organization is sophisticated enough to compute it, or secondary measures such as “net promoter score” for a much simpler measurement process.
With the framework established, we are going to move onto talking about actions that you can take to improve your retention effectiveness in each part of the framework.
But before we move further into this discussion, let’s sidetrack a little bit and discuss the effect of time on all metrics that we have presented to you above.
Time’s effect on your cohort metrics
Most of the metrics we presented in our frame in the previous sections, such as CLV, average order value, are a type of metric that we call “cohort metrics”.
“Cohort metrics” are intrinsically different metrics than metrics such as “revenue” or “orders” because they are forever changing, or “never golden”.
For example, let’s say that we are measuring the CLV of your customers that have made a purchase in Jan of 2020.
If we were to look at their CLV in March of 2020, it will be a number that is much lower than when we look at their CLV in June of 2020.
This is because the additional 3 months give more customers opportunities to come back and make purchases, and some customers may even, in the 6-month period, come back multiple times and make multiple purchases, further improving their CLV.
While it is expected for a customer’s CLV to stabilize over a long period of time (such as 1 year or 2), the period it requires for it to do so is generally unacceptably long for the analysis to be still time-relevant for us to make decisions.
This means a few things.
First of all, our CLV is never going to be one number, it may change and vary drastically depending on how it is computed, and when we are looking at the data.
Secondly, building upon the first point, each organization needs to create a standard for CLV computation internally, so cohort metrics such as CLV can be compared and contrasted systematically.
Generally, I recommend creating a snapshot of your customers’ cohort metrics 3 months after their initial purchase and use that snapshot as a reference to compare and contrast your customer value acquired from different channels and time periods.
What this means, however, is that your cohort metrics are useless for your customers acquired within the most recent 3 months - their cohort metrics are simply incomparable as earlier customers due to this time issue.
Of course, the number of months/weeks you set a snapshot of your cohort depends on your purchase cycle, decision making speed, and other factors, but the snapshot method remains one of the most effective ways to compare cohorts effectively.
Conclusion
With the entire “loyalty stage” journey described and complexities explained, I believe that you have made an essential breakthrough in the perception of user value.
Next part, we are going to cover how to make your customers come back and buy more, along with tips on conducting those analyses effectively at your organization.
As usual, comment below if you have any questions or concerns about this article, love to answer questions or just talk!