Subscriptions are an increasingly common way to buy products online, whether consumables like coffee and energy supplements, or personalized lifestyle boxes and fashion.
Research by McKinsey shows that 15% of online shoppers have subscribed to an ecommerce service:
In general, the “subscription lifestyle” is good news for ecommerce stores. As shoppers opt for the ease of online subscriptions, merchants can focus their efforts on retargeting, loyalty and discount programs, attribution, customer retention, reducing churn, and more.
But with every yin comes a yang. In this case, accurately measuring the holy grail of ecommerce metrics—customer lifetime value (LTV)—isn’t easy.
Important things, however, rarely come easily. And if, like so many others, you’re realizing that your marketing spend needs to get more efficient fast, an accurate LTV number is an essential first step.
LTV really matters for subscription ecommerce
LTV matters in ecommerce, but it especially matters for subscription stores. Why? Because both turnover (churn) and profit depend on your customers’ desire to continue engaging with your brand—and making recurring payments to your store.
Because of this, LTV is the metric that should receive most of your attention. It’s your best indicator of churn (the live-and-die metric for subscription stores), your best projector of profit, and perhaps your best aide in decision making when it comes to marketing and sales campaigns.
You may already know that. You may also know what makes tracking it so difficult in subscription ecommerce.
Tracking ecommerce subscriptions is uniquely challenging
Here’s the typical customer flow for an ecommerce store selling subscriptions:
- Customer visits online store;
- Customer chooses subscription;
- Customer goes through subscription checkout flow (via ReCharge, Bold, etc.);
- Customer completes first purchase.
But, unfortunately, that basic setup won’t allow you to differentiate among:
- One-time orders;
- First-time subscription orders;
- Recurring orders.
Why you must differentiate among order types
As you know, it’s a good thing to segment your data and determine if there’s a difference between customers who order once versus those who commit to a subscription. And, of course, different traffic sources convert differently for different order types.
In our experience, first-time buyers who aren’t sure about which product or product type to buy (e.g., which flavor of soda or “subscription product variant”) usually have a higher average order value (AOV) because they test multiple products before commiting to a subscription.
That behavior differs from those who commit to a subscription right away. It also differs from repeat buying behavior, such as coming back to make a similar one-time purchase after 6–9 months, which might be a good indicator that they will eventually become a subscriber.
But it’s no guarantee, and a basic tracking setup won’t give you visibility into these differences.
For recurring orders, subscribers won’t need to touch the website. Their orders will be created automatically by the billing tool (e.g., ReCharge) when the renewal date comes. The customer will interact with the website from this point on only if they wish to:
- Skip an order;
- Modify subscription details;
- Cancel the subscription.
This creates major attribution issues. Google Analytics will know nothing about the source of these recurring orders, so it will tag them as Direct traffic, and, therefore, won’t link them to the original user data:
Customers who choose a subscription in Shopify, for example, may not be linked to the user who commits to a payment in ReCharge. As a result, transactions are often improperly tracked and attributed; in many cases, a refund or cancellation isn’t tracked at all.
In one example, although 19% of the traffic comes from paid search, none of the ecommerce transactions are attributed to it. Instead, almost every transaction is Direct.
What merchants need to do is to link steps in the customer lifecycle to bring together a unified view of the customer in Google Analytics:
Merchants need to track each subsequent recurring payment to the same pre-checkout user journey (including the marketing campaign from which they came), along with other custom dimensions in Google Analytics to determine lifetime value.
There are manual and automated ways to do this:
- Manual. Add the platform (e.g., Shopify) Customer ID as the User ID when the original order is sent. Then, on a weekly or monthly basis, download the customer list from Shopify as a CSV and upload it as a user import via Google Analytics data import, with LTV as a custom dimension.
- Automated. Use an automated tracking solution, like Littledata (our product), Elevar, or Glew, to attribute traffic. Ideally, the solution will integrate with your billing tool (e.g., ReCharge, Bold), so that the work is done for you.
Obviously, we’re biased toward the latter solution. But regardless of whether you choose a manual or automated option, a few issues at the core of subscription analytics linger:
1. Aggregates lie. At this point, all order data is still aggregated. You can’t segment and drill down to truly understand customer behavior over time. This makes it harder to get accurate data insights, and even if you do get accurate insights, they’re less actionable.
2. For most users, standard subscription reports are difficult to interpret. You might think that running a subscription business will make you an expert in cohort analysis.
But the truth is that many ecommerce managers find it tough to interpret standard reports, let alone set up product or marketing action plans based on data insights.
Standard reports in Google Analytics hide the most valuable insights because they report only on a session-level scope. The insights on which traffic sources drive the highest LTV require a different approach—exporting data to Google Sheets.
How ecommerce subscription businesses can calculate LTV accurately
LTV is really split into two categories:
- Historic LTV. The sum of gross profit from all historical purchases for a single customer.
- Predictive LTV. A predictive analysis of transaction history (and shopping behavioral indicators) that predicts the LTV of a single customer. As predictive analytics work, this number should (theoretically) grow more accurate with each purchase or customer action.
Most run-of-the-mill LTV calculations for historic LTV look like this:
In this case:
- T = Average monthly transactions;
- AOV = Average order value;
- ALT = Average customer lifespan (in months);
- AGM = Average gross margin.
We suggest a different method, which ensures the most accurate LTV calculations to prevent churn.
There are two ways to do it. In both scenarios, you need a User ID to tie all orders to one customer. We recommend using a Shopify Customer ID and pushing it in a custom dimension in Google Analytics.
From there, you can view LTV in two ways. These two methods differ and cannot be combined due to scope:
- Use Method #1 to see LTV/Customer/Source. This is best if you want to drill down to see which sources of traffic drove higher lifetime value for a subscriber.
- Use Method #2 if you want only LTV without source data. This method is quicker and easier but requires an automated tool.
Method #1: Use session-level scope to understand LTV by traffic source
This report cannot be made in the Google Analytics interface and requires some data manipulation.
Step 1: Query the data from Google Analytics using Google Sheets.
The result of the query should look something like this:
Step 2: Use pivot tables to prepare the data for Data Studio.
Select all the data and select Pivot Table from the Data menu:
In the Pivot Table setting in the Row section, add the custom dimension with the User ID and “ga:sourcemedium”:
At this point, you’ve summed all the transaction revenue/source/customer. Now you can import it into Data Studio.
Step 3: Visualizing the data in Data Studio.
After creating a new data source and importing the Google Sheet with the pivot table, modify the aggregation type for transaction revenue. This is done to view the median LTV per channel. (We use median to limit the impact of outliers.)
Add “Source Medium” as a dimension and Transaction Revenue to the Metric section. The results should look like this:
If you wish to view the LTV per channel, repeat the process but instead of “Source Medium,” use the “Default Channel Grouping” dimension from Google Analytics.
Method #2: Push Customer ID and LTV to custom dimensions with an automated tool.
Method #2 is quicker and easier. In this scenario, we already have the Customer ID and LTV in custom dimensions via an automated tool.
Similar to the previous step, this data still needs to be exported. With the data exported, we can calculate the median LTV based on all customers (or any subset of them).
Customer acquisition cost (CAC)—the other half of the equation
Your LTV/CAC ratio is essential, but too many subscription brands are shooting in the dark. The prerequisite data for accurate calculations includes:
- A Customer ID linking Shopify and Google Analytics;
- A Client ID to track a customer source to recurring orders.
These two elements are essential to calculate LTV and CAC for subscribers, especially months after subscribing, when churn becomes a looming threat.
With these data points, you can determine how successful your acquisition campaigns are, say, three months from the acquisition date (to see whether you’re acquiring one-time buyers or loyal subscribers).
Because your cost of acquisition should ideally be one-third of your lifetime value, you want to aim for a 3:1 LTV/CAC ratio. A higher CAC (e.g., 3:2) often isn’t profitable in the long term; a ratio with far greater LTV (e.g., 5:1) usually means that you’re limiting growth.
For LTV/CAC, use a single source of truth
We recommend Google Analytics for your one source of truth. This is because it has all the data from each customer touchpoint (assuming you’ve tagged each of your campaigns with UTM parameters).
Reporting CAC only on Facebook data, for example, is an incomplete approach—you know only a part of what works (or doesn’t work) within your marketing strategy.
Customers have multiple touchpoints, from multiple platforms and multiple devices. In that modern, multi-touch, multi-device customer journey, a single source of truth offers a better big-picture understanding of how sources of traffic influence your conversions.
Having a single source of truth means consistent campaign tagging for traffic sources outside of Google Ads, and using the Cost Data Upload feature to upload the costs from other Ad Networks (e.g., Facebook, Bing, AdRoll, etc.)
Measuring LTV/CAC by channel and customer
A note of caution: No LTV guarantees that a channel is profitable or helpful on your path to scale.
In terms of LTV by customer, it’s more of the same. When accounting for CAC plus product and shipping costs, even a customer with twice the LTV may not be profitable:
Accurate LTV calculations matter because churn and profit rely on long-term engagements—and recurring payments.
Tracking subscriptions, not just one-time purchases, is a beast of its own when it comes to attribution. Subscription tracking requires a unique approach to differentiate among order types in Google Analytics.
With the right data, you’ll be able to see which channels drive profitable growth, and which aren’t worth the investment.