What if I told you…when you buy something online for the first time, a company knows a) how much you’re going to spend over time and b) how long you’ll continue to spend with them?
Well, they don’t KNOW, but they’ve got a pretty good idea on average. Lifetime Value (abbreviated as LTV) is an average that gives product managers and marketers an idea of how much you would be “worth” as a customer. Based on the activity of their current customers.
Get enough new customers for the right price, and over the long term your business is…
Companies are willing to give you a free trial, or free add-on, or signup discount, because they know you’re worth more than that as a new customer. This idea is not a new concept for most people, so please pardon this basic overview, #PeteArmy.
So it’s a calculated figure that you use as an assumption to drive business decisions.
Now here’s how you calculate it.
You can use revenue, profit, or contribution as the $$, any denomination of dollars depending on reporting needs. That can change the name of the metric slightly, so for simplicity let’s use revenue.
Bare bones LTV: The simplest version takes the transactions of all customers, and divides by the number of customers.
LTV in set “lifetime”: If you want to box your LTV in over a set period of time (“How much do we make over a three year “lifetime” with this customer”) you just need to put date limits in your data and only pull the “current” customers and every transaction they ever made.
LTV with net present value: These first two examples take current and future $$ and treat them the same. That’s not sophisticated enough for some models, and “discounting” the future $$ compared to ones today can be a way to compare decisions where you may be trading current for future value. You can discount these future transactions using an NPV calculation, but you’re getting into some fairly robust calculations.
You’ll need to assume:
- Current trends for # of transactions, size of transaction will hold
(Will future customers exhibit similar purchase patterns)
- A close enough/relevant discount factor for your net present value calculation.
(Does this represent the tradeoff involved in forgoing current $$ for future $?)
It’s often helpful to calculate out “estimated value per period” to get there.
Again, you can use that $ value for different metrics than revenue. If you have thin margins, or you work in services with an overhead rate, you can take out the cost of selling/fulfilling, or some other figure important in your internals.
Where I first learned most of this stuff I was doing analyses for nonprofits, to learn how much we could expect in donation $$ from a donor. We’d subtract the costs of acquisition and all the marketing spend during a lifetime, and we would have “Net Long Term Donor Value.”
My past few gigs have been in ecommerce, and COGS enters into the equation very differently, when you’re selling products/services. The concept is a tool that different businesses and products will use differently.
Here’s how you can use it.
User Acquisition/performance marketing campaigns:
Most “acquisition” campaigns operate at a loss. Meaning you might spend $10K on Google Ads, and be very happy with making $6K back as long as they bring in enough new customers. A first purchase doesn’t pay for the whole campaign on a net basis, and it’s still a success. If you’re an enterprise business account, you represent a major win as ONE customer. How do companies know this? They estimate your LTV.
And even as one consumer, you’re worth more than that signup discount, guaranteed. When performance marketers are setting their budgets, they use a metric called Cost Per Acquisiton or sometimes Cost per Action (generally abbreviated as CPA. Over the lifetime, you pay this # back hopefully.
Sometimes this CPA is a hard target — you might set your Google Ads campaign to stop for certain adgroups when “Pogo” starts to cost more than $50 per tracked conversion. Other adgroups like “Pogo Stick” and “Moon Shoes” might still cost you $45 of ad budget per successful conversion, and be within your campaign guardrails for efficiency.
What should those CPA guardrails be? It’s really whatever you like. You might be obsessed with growth, have a bunch of VC money and be willing to tolerate a high cost for a campaign. Or you may want to use CPA as a strict limit to scale efficiently with your ad $, even if that means slower growth.
You could even throw it over an ROI-based target for the lifetime value. “I only want to pay 20% of the lifetime value (or less) to acquire a new customer.” That means, if your LTV is $300, you’ll take campaign results that bring in new customers at a CPA of $60 or lower. That way you’re likely to hit a nice, 4X+ ROI if the rest of your math is good.
Measuring impact & weighing decisions:
When faced with the choice of retaining existing customers OR trying to find new ones, it’s almost always smarter to retain and build trust with the people who know and love you.
If you are faced with a decision involving retention, keeping customers happy, or some other internal enhancement — here’s how you use LTV to inform that decision.
Let’s say you need to decide to A) Pay for new fancy accounting software that will let your accountants pay invoices more quickly, or B) invest in new ecommerce forms that increase transaction size once they’ve launched. Your engineers can only implement one and it’s up to you to make the call.
Using the math from earlier — calculate the change in the green figure with those new forms, and then figure out its impact over the customer’s lifetime.
So taking the period revenue ($10 more annually, on average, then NPV discounted for every period) might mean that over a customers lifetime, you’re going to see $20 more in LTV.
Then you multiply that new change by your customer count and you have a rough proxy of the $$ upside in changing your forms. Even for existing customers, LTV can help.
With the LTV increase, you might have enough to hire an additional accountant and keep bringing in more customers. (Be nice to the accountants and upgrade their software eventually).
I hope you’re still awake after all this math. How have you used LTV in your organization? Would a better explanation of CPA or metrics be helpful? Putting this out there in the world so it can help someone, so let me know!