Once you understand the value of a customer in each stage of the customer journey, you can start making good decisions about not only how to spend your marketing resources but also how to invest your engineering time. For example, you can use Google Analytics to measure your cost of traffic or transactions as they relate back to your Adwords campaigns, but you need to know what those actions are worth to you to make smart decisions.

Let’s start by thinking about the stages in the customer journey. For a transaction driven website there is often a natural funnel, and this might look something like:

• First time visitor
• Repeat visitor
• Registered visitor
• First time buyer
• Repeat buyer

The value of traffic

Let’s say you have 10,000 unique website/app visitors in a month and 500 transactions. That’s a 5% conversion rate. Let’s further assume that your average transaction is $50 with a 50% margin factoring all variable costs, so you are clearing $25 per transaction. But what’s the value of a visitor? That’s important to know, as it should serve as the upper limit for the marketing team when calculating the average cost to drive traffic to the site.   The simplest analysis suggests it is total margin divided by number of visitors or (500 x $25)/10,000 = $1.25. That is straightforward enough; but all traffic is not created equal. Visitors do different things when they come to your site. They might make a purchase, or register to be alerted about sales or to get a whitepaper, or just leave and never come back. Google Analytics refers to the positive actions as micro conversions (e.g. registration) and macro conversions (e.g. a sale).

The value of repeat visitors

You can use an analytics tool like Google Analytics to break the visitor behavior down further. How many of the visitors were repeat visitors and what percent of purchases were made by those repeat visitors versus first time visitors. Let’s say the data shows a standard 80/20 breakout where 80% of the visitors are first time, but they only account for 20% of the purchases. That suggests that 8,000 first time visitors generated 100 transactions worth a total of $2,500. So the value of a first time visitor on a first order analysis is $2500/8000 or only $.31! The value of a repeat visitor is $2.00, derived from $10,000 in margin divided by 2,000 repeat visitors. That suggests that marketing efforts designed to drive awareness and first time traffic to the site should not exceed $.31 per visitor. Programs focused on getting visitors to come back would make sense as long as they cost less than $2.00 – $.31 or $1.69, assuming that you had already paid $.31 for their first visit. That might include retargeting ads for example.

The first order analysis is very simple, but still valuable. Second order analysis would take into consideration the percentage of first time visitors that return on their own. Let’s say that is 20%. If returning visitors are worth $2.00 and 20% of first time visitors will come back on their own, then a first time visitor’s value is increased by the chance that they will become a returning visitor, numerically by $.40. So getting first time visitors to the website is worth $.71 because some visitors will transact on their first visit, and some will naturally come back and are much more likely to transact on that second visit. This refined analysis suggests that you should be willing to spend $1.29 ($2.00 – $.71) to get first time visitors to come back.

The value of registered users

Getting visitors to return is an important aspect of any online business, whether it makes money on transactions or advertising or subscriptions. For simplicity, we’ll stay with the transaction example above. Retargeting, specifically showing users ads for your products on other sites that they visit after leaving your site, is an effective tool to generate repeat visits and purchases. The other effective tool to generate repeat visits is email. But to email users, you need their email addresses! That is why so many sites try to get you to register.   Some sites ask, and some sites may require you to register. In the early days of Zulily, they forced visitors to register if they wanted to even see the deals on the site. Other sites may require registration to get beyond basic information. Forcing users to register to get access to a site is called a registration wall. We’ll get back to that.

To know what we gain when someone registers, we have to know the likelihood that we can get them to come back using email re-marketing. Remarketing can be very effective if it is personalized based on your behavior the last time you were on the site.   For example, if you were browsing shoes on a website and you later get an email saying that the shoe sale has just started, then you are highly liked to return.   We need to know what percentage of people will respond to our email campaigns (open rate x click through rate x number of emails sent). Let’s be generous and assume that 50% of registered users who get the email campaign over some relevant period of time return to the site. From our work above, we valued repeat visitors at $2.00. But only 50% of registered users are going to respond and visit again, so a registered user is worth $1.00. I am necessarily simplifying the analysis for the sake of clarity and brevity. We could account for the value of maintaining brand awareness from our email campaigns along with many other factors.

In all likelihood visitors who took the time to register are going to be better customers on average. The analysis above can help you decide whether to invest in a registration system. You’ll have to make some assumptions about how effective your email re-engagement marketing campaigns will be.   Once you have registration in place, make sure you segment your sales by those who registered before they purchased for the first time to understand how much you should invest in getting people to register. Follow the steps above, dividing sales from that segment by the number of users in that segment. Remember that any visitor who purchases registers in the process of purchasing, so track registration prior to purchase.

When you have determined the incremental customer value derived from getting someone to register you can then determine what level of incentives you should use to motivate that behavior or how much technical development resource to invest to optimize that process.

Registration wall

Getting back to the registration wall, why not just make all visitors register? How would you decide if this was a good idea for your business? Well, two things are going to happen when you put up a registration wall. Some people will register and some people will just leave and likely never return (abandonment). You can guess at what percent will do each or you can put up a registration wall for a brief time or for a fraction of traffic as a test and get your registration and abandonment data. If setting up a test requires an engineering investment, you can calculate the breakeven assumption and gut check that. We’ll come back to that.

We’ll use our example above to value a first time visitor at $.71. So for every new visitor who abandons we lose $.71, because we have driven the likelihood that they will come back or purchase on their first visit to zero. Registrations are worth $1.00 according to our example above. To calculate the incremental customer value from the registration wall you can use the following formula where Y is the percent of visitors who register: (Y x $1.00) –  ((1 – Y) x $.71). If 50% of users register the incremental value is $.50 – $.36 = $.14. If only 25% of visitors register, then the incremental value is $.25 – $.53 = -$.28, and you would be losing value. You can calculate the breakeven registration rate, which in this case is about 42%. If less than 42% of users register, you will be losing customer value.

The value of a first time customer and repeat customer

Let’s keep going along the journey. Some of your customers will transact with you only once, and some will become regular customers. One-time customers perhaps had a one time need, or found you through a search result to a specific product or via an ad or a sale or friend referral. They likely don’t have an ongoing need for your product/service or already have other sources that work well enough for them.  In some cases you can motivate a first time buyer to become a regular buyer by incenting them to come back to your site to explore your offerings. In other cases, first time buyers will love the value proposition you offer and become regular customers organically.

A repeat buyer is your holy grail. Repeat customers have the highest lifetime value (LTV). In a prior post, I provided an illustration of calculating an LTV which by definition is the value of the repeat purchaser. For purposes of our calculation below, lets assume that the LTV is $100.

At the top of this article, we were clearing $25 in margin on a single purchase. So again on first order analysis the value of the first time purchaser is $25. But if 40% of those first time purchasers organically or as a result of your email remarketing campaigns go on to become regular customers, then their value increases by the increase in value between first time and repeat buyer ($100 – $25 = $75) times the percentage of those that make the transition, in this case 40% x $75 = $30. So we should be willing to spend something less than $25 +$30 = $55 to get a first time buyer. We should also be willing to incent our first time buyers to become regular buyers. That amount of that incentive should be less than $100 – $55 = $45.

Caveat

There is almost always more than one way to calculate the value of your customers through their journey. The more data you have and the more granular your segmentation the more effective your analysis can be. I have simplified the analysis and the scenarios above for clarity and brevity. There are always more nuanced ways to think through these calculations and this is meant to be a starting point for those without a lot of experience in this realm.

Next Steps

The more granularly you follow the steps in the customer journey and also segment your customers by how you acquired them and their purchase behaviors the more often you will be able to make smart decisions about how to increase value to the enterprise. If you assign those interim values in Google Analytics for example you can quickly see whether an A/B test is working for you. You can also make informed decisions around investments in marketing dollars used to acquire new and repeat traffic, motivate customer registration, and incent first time and repeat purchases.

I hope this has been helpful. Please reach out with any questions.

Cheers,

Pete

Peter.Baltaxe@gmail.com

While working on some course material for entrepreneurs, I realized that it would be useful to build a foundational set of posts on calculating and using customer lifetime value.  In a follow-on post I’ll discuss how to think about customer value through out the steps of the customer journey and use that to make technology and marketing investment decisions.

Whether you are thinking through a new business idea, evaluating marketing channels, or trying to raise money from investors, you need to understand customer lifetime value (LTV). You will use LTV to convince yourself and others that you can make money on a per customer basis. Your LTV should exceed your cost of customer acquisition so that when you have enough customers to cover your overhead you will be profitable as a business.

This basic concept applies to any type of business, whether you make money on consumer transactions, SaaS subscriptions to businesses, or advertising. I will run through a transaction business below (think e-commerce for physical or digital goods), and then show how the approach is modified for other business types.

There are four important factors in understanding customer lifetime value: transaction margin, purchase frequency, customer lifetime, and a discount rate.

Transaction margin

I have seen examples of calculating LTV using gross revenue, but this is disastrous when using the LTV to determine what you can afford to spend to acquire a customer. I strongly recommend using transaction margin rather than gross revenue. To calculate the variable margin from a transaction take the revenue of the transaction minus the variable costs. Be honest with yourself about the costs in addition to the direct product costs (cost of goods) and shipping and handling costs (for physical businesses), including: a returns percentage, a percentage of transactions that require customer support, credit card fees, marketing discounts and coupons, etc. For example, if 5% of transactions require support, then add (.05 times the average support cost) to your total variable cost per transaction. Any cost that grows directly with the number of transactions should be subtracted from the gross revenue to calculate the transaction margin. If you have variable sales or marketing costs like a one-time commission to an inside sales person or channel partner, those are customer acquisition costs and are not part of your margin calculation.

Purchase frequency

If you are selling coffee like Starbucks is, your purchase frequency might be five times a week. If you are selling cars, it is more like once every three to five years. Purchase frequency is an incredibly powerful lever, and the role of retention marketing is to get your customers to come back as often as possible.   Subscription businesses are popular because you are automatically increasing the purchase frequency. If you don’t have a lot of history, you’ll have to make an educated guess as to what the typical purchase frequency might be.

Even before you have real repeat customer activity to analyze there are a couple of ways to triangulate to an educated guess. One approach is to benchmark your offering against other businesses and try to get some understanding of their purchase frequency by talking to them or their customers. Alternatively or complementarily, you can gather insight into purchase frequency from customer interviews, and then haircut the stated frequency as actual purchase behavior is often less than stated intent.

Lifetime

This is the most subjective measure for early stage companies that don’t have a lot of customer history. I saw an example of a hypothetical LTV calculation for Starbucks that estimated the lifetime of the customer at 20 years.   That’s pretty optimistic, but by now they may have the history to justify that.   It is tempting to think that your customers will be with you forever, but you will lose customers for many reasons.   Looking at the music business at Amazon for example, which I was directly involved in, we were acquiring incremental CD customers and we had ten years of history of Amazon customers to analyze. But some percentage of the CD customers moved to digital downloads as the CD format lost favor. Digital downloads then decreased across the industry in favor of streaming music services. As another example, Quicken consumer software had a long history of loyal customers but those customers are now moving to the free Mint SaaS product. No company or product is immune to losing its customers. If you are selling baby supplies, your customer will outgrow the need in a year or two. The general lesson is that your customers’ needs and wants evolve and their options for solving their problems and achieving their goals are constantly evolving.

For an early stage company with limited cash reserves, I would advise being very conservative on lifetime for another reason. If it takes a long time to recoup your customer acquisition cost, then you may not have the cash left to see the day that your margin from that customer recoups the cost of acquisition. At a minimum it will constrain your ability to grow if it takes a long time to recoup your customer acquisition cost. Ideally you can recoup your acquisition costs within a year, and I would think hard if your business model requires more than two years to recoup the cost.

Discount rate

This is the percent by which you should discount the future cash flows when performing an LTV or ROI calculation. Established companies typically use a number such as their weighted average cost of capital (WACC) to discount future cash flows for investment analysis. This might be 5% to 15%. For early stage companies I would suggest discounting the future cash flows by a much higher percentage for two reasons: 1) to account for the high degree of risk of those future cash flows, and 2) unless you have hoards of cash, you will need to acquire incremental cash through the sale of equity. In the latter case, you are giving up some percentage of all future cash flows of the company, and that hopefully is worth a lot.

Putting it together

Lifetime Value (LTV) is calculated by summing the margin on the transactions over the lifetime of the customer, while using the discount rate to reduce any margin that after year one. So if an average customer buys $50 of product with a margin of 50%, their average transaction value is $25. If they purchase four times a year, then their annual value is $100. If you estimate a lifetime of two years, and discount the second year’s cash flow by 25%, then their LTV is calculated: $100 + ($100 x .75) = $175. The present value formula for the year 2 margin is Margin x (1 – discount rate); for year 3 margin, it is Margin x (1 – discount rate) x (1 – discount rate), and so on.

Using LTV

If you cover your customer acquisition cost on the margin from your first transaction, then you are in great shape. If you don’t (and many companies don’t), then you need to calculate the LTV. It is critical to understand that your business will work on a per customer level. The margin they generate for the business over time, i.e. the customer lifetime value, needs to exceed the cost to acquire them. Then once you understand your marginal profit per customer, you can calculate how many customers you will need to cover your overhead at various stages of growth.

Calculating customer acquisition costs is obviously important to this process. How you calculate your acquisition costs is a function of the type of marketing channels you use and is beyond the scope of this post.

Revenue vs. margin

As I mentioned above, I have seen examples of using revenue rather than margin to calculate a customer lifetime value. This makes sense if you have no per customer or per transaction costs. For example, in some SaaS businesses all the cost is in the overhead of developing the software and in the hosting infrastructure. There might be a minimal additional server cost per customer or a small amount of support, but that is likely not worth factoring into your calculation. You can use revenue rather than margin, since they are essentially the same, and compare that to your customer acquisition cost. Businesses based on an advertising revenue model fall into this category as well. You pay to get your eyeballs (your customer acquisition costs) but there is not an incremental cost to deliver value to an incremental site visitor.

Next steps

Often a company uses multiple channels to acquire customers. To the extent possible you should track the source of those customers against their behavior over time so that you establish an LTV for the customers from each channel and compare it to the customer acquisition cost for the channel. You may find for example that customers you acquire through an aggressive discount program don’t become loyal customers (they are bargain hunters) and so their value is negative when you compare their LTV to the cost to acquire them.

Once you know your LTV for your repeat customers you can start to make informed decisions about many other aspects of your user experience and how you allocate financial and technical resources. That is the topic of another blog post.

I hope this has been helpful. Please reach out with any questions.

Cheers,

Pete

Peter.Baltaxe@gmail.com