Can CLV calculation ever be accurate?

Nozzle’s approach to Amazon customer lifetime value

CLV is a tricky metric, but a powerful one if you have a clear and accurate understanding. To a large degree, we built Nozzle to solve CLV calculation problems for our customers, and help formulate CLV-driven strategies and tactics for customer acquisition — e.g. PPC. Our ability to help you understand CLV and customer buying trajectories is central to the value we provide. Here, we are going to explain how we do it and what understanding CLV lets you do.

Additional resources: Check out the two webinars we’ve conducted on this topic if you want even more details

• How Important is Customer Lifetime Value on Amazon?

• Customer LIfetime Value on Amazon 

Why CLV is so important

Customer lifetime value (CLV) — sometimes denoted as LTV — gives you:

• A better understanding of what your products are really worth
• A new approach to break-even ACoS and PPC spend
• Insight into which products are most valuable to long-term profits
• A longer-term outlook

How Nozzle calculates CLV

There are two calculation methods for CLV:

1. Simple CLV: useful for a rough calculation to provide general figures. The most straightforward customer lifetime value formula is a historic model. Here CLV is equal to the total value of each transaction for a given customer, multiplied by your average gross margin.

CLV is such an essential metric that even downloading files from FBA and using pivot tables is a worthwhile endeavor for the value it will deliver.

However, there is a better way. 

1. Complete CLV: calculated on a year-by-year or month-by-month basis. You can factor in models of changing revenues and costs, and compare projected ROI of different investment possibilities to calculate present value. This is the way we approach it.

• Step 1: Calculate your revenue on a per-customer basis. You can do this on any timeframe — for example, month-by-month or quarter-by-quarter. 

• Step 2: Subtract sales tax, Amazon fees, and cost of goods sold, along with advertising spend. 

• Step 3: Analyse that data using a cohort analysis based on when customers first purchased.

• Note: Cohort analysis involves breaking apart data sets into related groups for analysis — this should align with the timeframe you chose in step 1. For example, comparing people who first converted in Q1 of 2019 with people who first converted in Q1 of 2020, and how they compare in their respective first and second quarters, etc.   

Something to note here is that, as the graphic above shows, the main element of the “cost” equation that you control is your advertising costs. Unlike costs baked into each sale (which scale linearly with each sale) advertising costs are often an upfront cost to acquire a customer. Your ability to spread out this cost across the complete gross revenue a customer will generate lets you be far more strategic with this critical point of control. 

Finding and accessing this data in Amazon is the main challenge. Robust CLV calculations depend on:

• Knowing where to capture customer and product data
• Being able to attribute profit to a customer over their entire lifetime
• Having sufficient data to forecast lifetime value based on past behavior to date
• Generating models of behavior that match the data well 

And it all needs to be carried out at speed for you to benefit. We solve these problems by tapping into a number of Amazon APIs and pulling data from Amazon MWS to crunch the numbers using proprietary AI algorithms, and then presenting the figures back to you in customizable dashboards. 

However, to illustrate why it’s worth the challenge 

What you can do with Nozzle CLV analysis

Customer lifetime value matters. Fundamentally, it allows you to understand the true value of your customers. Knowing this lets you accurately determine how much you can afford to spend in order to acquire a new customer — benchmarking ACoS targets with a longer-term vision of profitability. In total, CLV enables you to:

• Create a competitive advantage in customer acquisition strategies
• Understand and improve customer retention rates
• Reveal the level of brand loyalty you command 
• Quantify price elasticity within your market — how sensitive your customers are to price.  
• Target decision-making to deliver long-term profitability

What outcomes you derive from CLV depends a great deal on your business and your business goals. Here are some outcome strategies that CLV can help develop or validate.

1. Improve bidding strategy by changing your break-even ACoS/RoAS

Amazon calculates your ACoS (or RoAS — given the recent shift in ad metrics on Amazon) based on the assumption that you will only ever make one sale of that product. A standard view of break-even ACoS looks to make sure that your ad spend does not exceed your pre-advertising profit margin — ensuring single-sale profitability. If you are unfamiliar with this process, check out our article — How to Calculate Break-Even ACoS.

But CLV gives you more insight into your true costs and returns. For example, if your Average Order per customer is 1.5 for a specific product (as in the example below), you can bid much higher than your competitors, win more business and sell more — all while still being assured that you will turn a profit long-term.

In this case, your break-even ACoS has increased by 50%, just by knowing your Average Order per customer value for that ASIN.

2. Bundling

For some Amazon Sellers, the most effective way to differentiate from the rest is to create your own niche. By bundling products, you make something unique to you, helping you win the buy-box. It also helps you increase CLV by creating higher average order values. However, bundling is only helpful if you bundle products that your customers want to buy. 

Complex CLV analysis goes beyond single ASIN projections to include “frequently bought together” items. You can repurpose this information to inform an effective Amazon bundling strategy. Nozzle makes this easy by creating heat maps that allow you to highlight which products are prime candidates for bundling, and which are most valued together by target customers.

Bundling may become even more useful when you consider the Virtual Product Bundles tool for FBA that is currently in Beta in the US. Brand owners will be able to set-up ‘virtual’ bundles, made up of two-to-five complementary ASINs, which can then be purchased together from a single details page. Amazon does the packaging. That means you can now offer bundles without changing your FBA inbound inventory.

3. Determining long-term customer trends

The benefit of CLV is that it’s time-based. Here, we have an analysis of the CLV trends over several years. In Q2 2018, in the example below, people spent $18.50 on products, and there was a general  upward trend in spending each quarter after. This is great context to go investigate why. For example, perhaps prices increased, or product mix changed.  

However, where this data really starts to get interesting is looking at trends over time. Here, you can see that customers who made their first purchase in Q2 of 2019 only spent $16.50. So, although this business is gaining customers, they need to be wary of the fact that new customers are spending less than earlier ones in their first quarter. If you are spending on ads to attract lower spending customers, it may be time to pause and reflect. 

Note: In this specific instance, this problem seems to rectify itself in Q3. Both cohorts end up spending $21.50 in their respective Q3s. If anything, the long-term CLV trend for this brand is positive, with the exception of a bad start to Q2 of 2020.    

These same CLV tools can also breakdown how segments of your customer base contribute to your total revenue. For example, in the graph below, it shows that the brands top 20% of customers represent 47% of sales. This allows you to understand how dependent you are on existing customers, and tailor strategies accordingly.

In this situation, it might be valuable to spend more to attract new customers, even with a slightly lower value, just to diversify your customer base. This is valuable context for understanding cohort analysis.    

3. Choosing your remarketing timing

By analyzing your buyer behavior, you begin to see how long an interval there is between repeat purchases. This helps you time remarketing campaigns. In the example below, you can see an example of an ever-increasing time frame between replete purchases. Knowing this would allow the brand to align their remarketing with this trend. 

Obviously, the relevance of this will depend on the categories involved — grocery, beauty and supplements being better candidates for this strategy. Big-ticket items where you’d have to wait years for a repeat purchase — maybe not so much. However, by having this data readily accessible, you can experiment with segments of your product portfolio and see how changes to strategy impact sales over time. 

This same dashboard provides additional context about buying trajectories (the horizontal bar chart), which can help with remarketing targeting decisions, and potentially with previously mentioned bundling strategies. 

Lastly, the vertical bar chart here provides context about your overall customer retention strategy. You get a breakdown of orders in any given month by what year those customers first purchased a product. 

In the chart above, the green lines represent people who first converted in 2018. This, obviously, comprises 100% of 2018’s orders. However, when you get to 2019, those who first purchased in 2018 because a smaller part of the total — which is why you see the green bars shrink as a percentage of the total. Looking at this over time enables you to understand how new and old customers stack up to comprise your total customer base, and identify trends within that process.  

If you couple this data with the cohort analysis described in point three, you get a clear picture of how well you retain customers and whether or not you are replacing old customers with new customers that spend less or more on your products — critical to your long-term health. 

Getting more with data

The more accurately CLV can be calculated, the better it can be presented. As we have shown with a few examples, this CLV-enriched perspective gives you an edge. That is because: 

• Selling to existing customers is easier and more profitable than finding new ones. 
• Retaining them is more impactful on your bottom line than new acquisition.

However, the way in which this data is presented is just as valuable as the information itself. By presenting this analysis in automated dashboards, the struggle of combing through spreadsheets and creating pivot tables is removed. You get straight to the insights that are needed to take action.  

Of course, CLV is not suited to every product. But over time, analysis of CLV on Amazon is moving away from just a product-focus to be more customer, persona, and demographics based. Such changes will make CLV analysis even more central to your retail, portfolio, and advertising strategies.

To execute at scale, you will need to deploy the right skills, software, and working practices to become really customer-focused. Get in touch if you want help fine-tuning your approach to customer lifetime value and Amazon analytics in general.