Following on from my last post exploring how virtual worlds make money, how much they can expect to make, and the market size, this is the first in a series of articles that will dig into the numbers that drive both the costs and revenues for causal virtual worlds.

In this post I'll explain how to calculate the lifetime value of a player (LTV). In the following two posts we'll look at the user acquisition model, and then dig into the cost of hosting, moderation, and support.

The LTV is a key metric for profitability: the LTV of all paying customers must be greater than the cost to acquire those customers, plus the cost of servicing all free and paying customers.

The lifetime value is driven by three metrics:

• Conversion of free players to paying players
• Spend per paying player
• Length of active membership

Its important to monitor and optimize these three metrics, as we'll see later, small changes can make big differences.

Average Revenue Per User
We start by taking the conversion from free to paying and the average spend to calculate the Average Revenue Per User (ARPU).

ARPU = % Converted x Ave Spend

Lets plug some numbers into our hypothetical world. We'll assume that 10% of our users will signup for a £4 monthly subscriptions. The ARPU is:

£4 × 10% = £0.40 ARPU

If the world also includes micropayments, perhaps through mobile or credit cards, we can extend the model:

ARPU = (% Subscribing x Subscription) + (% Micropaying x Ave Spend)

If we assume that 15% buy through micropayments, and spend an average of £2 per month, then the ARPU is:

(£4 × 10%) + (£2 × 15%) = £0.70 ARPU

Note. A well performing world can expect an ARPU between $1-$2. Our £0.70 falls within that well performing range. However, it would be unusual to launch and hit that target within a few weeks. Instead, developers should take the time to optimize their world and improve the conversion of free into paying players.

Churn
Churn is a measure for the number of players that leave your world every month and is typically calculated as a percentage of the total active user base. For example, a 33% churn means that on average 1 in 3 players leave the world every month.

We can calculate the lifetime of our average player with the formula:

LT = 1 / % Churn

So, for our 33% churn our player lifetime is:

1 / 33% = 3 Months

Lifetime Value
The ARPU calculation gave us the revenue per user per month and the churn rate allowed us to calculate the average user lifetime. We combine these two numbers to get our lifetime value:

LTV = ARPU x LT

So, in our hypothetical world we have a LTV value:

£0.70 × 3 Months = £2.10

Lets revisit churn, remember that the player lifetime is calculated as 1 / % Churn. This means that if we can half our churn we double our revenues. This becomes an increasingly big lever as we start to approach low churn rates.

For example, with a 20% churn:

LTV = £0.70 x (1 / 20%) = £3.50

If we cut the churn down to 10%:

LTV = £0.70 x (1 / 10%) = £7.00

User Acquisition
At this point it would seem reasonable to assume that to acquire a new customer we can afford to spend up to the LTV less the cost of servicing that player. In other words:

Max Acquisition Cost = Lifetime Value - Servicing Cost

However, this does not take into account any viral effects.

While some of your world traffic will come through your own marketing activity, such as ad networks, and search advertising, another portion will come through your players telling their friends. Its this second portion that gives our viral growth.

When we talk about viral growth we often think of exponential viral growth, the kind of growth that Facebook has seen. To be exponential growth you need each of your players to invite and convert more than 1 new player. That is, for every player who joins, they bring another 2 new players with them.

Because the cost to acquire these viral players is effectively zero it can be tempting to optimize for achieving exponential viral growth. However, achieving long lasting exponential viral growth is difficult, and may not be the best strategy.

With a solid conversion to subscription and micropayments you can afford to spend money on advertising or other marketing activity. In contrast, social networks have to focus on low cost viral growth because their ARPUs are significantly lower

Continuing our hypothetical virtual world, if we have a viral growth where on average 20% of our new players invite 4 people then we have a viral factor = 20% x 4 = 0.8.

This means that on average each player invites 0.8 other players, who invite another 0.8 players, who invite another 0.8 players, and so on. This geometric series can be simplified to:

1 / (1-0.8) = 5 Players

So, for every player that we acquire through our paid marketing activity we get 4 for 'free' through viral effects.

Getting to the bottom line, we can calculate the total value of acquiring one customer as:

Total Value = LTV x (1 / Viral Factor)

For our hypothetical model we see:

£2.10 x 1 / (1-0.8) = £10.50

In other words, we could afford to spend up to £10.50, less servicing costs, to acquire one new player. Although, its unlikely you would need to spend over £10 acquiring one new player!

By paying attention to all the metrics discussed in this post its possible to optimize the funnels so for profitable growth through a function of marketing - be that search, affiliates, ad networks, and so on.

In the next post I'll explore two of the costs for servicing your players, hosting and moderation.

At the end of the series I'll share the excel model that combines all of the topics discussed into one template virtual world business model. Of course, any feedback received from these posts will be included.

If you'd like to review the model and input sooner email me matthew.warneford@dubitlimited.com or send me a message on twitter here.

Lastly, learn more about our virtual world platform that gives creatives the tools for designing a world without needing to program.

Matt

Update on Sunday, August 2, 2009 at 9:17AM by Matthew Warneford

Read the second article in the series: Calculate how the addressable market size affects your virtual world business model