Definition of LTV

[fish_bone]

Should I include non-paying users in the data set when calculating LTV, or should I only include paying customers?

E.g. if I only have four users, who paid the following over a period of time:

User A: $0

User B: $0

User C: $2

User D: $4

If we discount forecasting, does that mean that the LTV is $3 or $1.5?

Depending on the answer, ROI is either equal to LTV / CPA or LTV / CAC.

 

[613jono]

@fish_bone LTV is often seen as super ambiguous. But you should include a random cohort of your users, not just paying ones. You're trying to work out how much users will bring to your company: it's about estimating all your users to work out if you're on a path to monetization.

Good quote from Martin MacMillan, Founder of Pollen VC:

Put plainly, Lifetime Value is all revenue a single user is expected generate from the time they download the app until they stop using it. It’s quantified as a window of time in which an app is seen as turning a profit over operating costs.

That is from a series of posts on using LTV.

• Part 1
• Part 2
• Part 3

You can also use Pollen's LTV calculator here to simplify things.

 

[fish_bone]

@613jono That would then mean that to calculate your ROI, you need to divide your LTV by your CPA (cost of acquiring any user), not your CAC (cost of acquiring a paying customer). Correct?

 

[613jono]

@fish_bone To show the ROI of your UA campaign, you want to highlight the ratio of LTV to CAC. That's pretty easy to do as long as you keep all those acronyms straight.

Working out what that ratio means is a little more complicated. Benchmarking is pretty sketchy because it depends on a bunch of other factors - time, for instance; or relative competition in your category/country.

Some genres will have long slow payoff cycles - e.g. strategy games - where other categories - e.g. arcade games - may have a quickler payoff cycle.

LTV changes over time.

An LTV calculation is just a snapshot of a proposed trajectory of an app. As the metrics change - say a content update improves monetization - this can change your LTV over time. Smart marketing managers calculate their LTV constantly to see how it evolves over time based on the product and the audience acquired.

 

[fish_bone]

@613jono But your LTV considers all users (both non-paying and paying customers), while CAC only considers your paying customers. The cost of acquiring a paying customer is higher than acquiring a regular user, so the LTV vs. CAC ratio will therefore be off. The LTV to CPA ratio should be more interesting since both factor in all types of users and are therefore comparable.

For example, if

LTV > CPA

then

ROI factor = 1

e.g. you're making more money from campaigns than you're spending.

 

[613jono]

I really recommend Part 2 of the series on LTV I linked to above, it shows a bunch of examples of how to calculate these metrics and how various changes might effect them.