Don't miss
  • 2,232
  • 6,844
  • 6097
  • 134

ARM yourself: how to reduce CPA and make more money

By on November 16, 2010

This is the second post in a series I’m calling ARM yourself.

Acquisition used to seem so easy. You simply added a button called “Invite your friends” on your website. They all clicked on it, you sold your business for squillions and bought a Maserati.

Not so fast. Not only was it never so easy, but those days are definitely gone. Between Facebook’s closure of the viral channels (mainly because users complained it was spam, but also to force its partners to pay for Facebook ads) and increasing consumer scepticism, acquisition is getting way harder.

Not only that, but CPA is going up as more people compete for customers than ever before:

  • Zynga and other highly profitable Facebook games companies want to keep their market position and are spending on customer acquisition to maintain it
  • Free to play companies like Bigpoint, Gameforge and Innogames are venture-backed and keen to show continued growth
  • Asian companies like Shanda and TenCent want to make further gains in the Western market
  • Traditional MMO players like Turbine are taking games like Lord of the Rings Online and Dungeons & Dragon Online free to play, and will be spending their marketing dollars on lead generation.

If CPA is rising, and virality is dying, what to do?

The first thing to do is to make sure that you have the basics covered:

  • Harness the social graph: Use Facebook Connect, Twitter, Gmail and other existing services to build a community quickly. There is a strong correlation between number of friends and ARPU in online games, so you get a double whammy of more users and higher ARPU
  • Build “invite a friend” into the gameplay: I know I said virality is dead, but there’s no harm in trying, right? The best invitations are personal, specific and give the recipient a good reason to accept them. For example, if a player designs an avatar for a friend in a game and send it to them, saying “I’m playing a great game, and here’s what you would look like in it”, the chances of a response are greatly improved.
  • Analyse, analyse, analyse: Build a conversion funnel in a spreadsheet. Track every step. Remove every unnecessary step. Don’t ask for data you don’t need at signup (Do you really need to know the sex of your gamers? Their exact data of birth? Even usernames might be optional if you just get an email address and a password.) Moshi Monsters improved conversion rates from 30% to 40% by spending three days tweaking their home page. There’s a good ROI in that kind of work.

Pull the levers you can pull

There is a difference between how much it costs you to get a customer to your websiite (sometimes called CPL, sometimes – confusingly – CPA) and how much it costs you to get a user to register, or pay, or whatever metric is your primary success measure.

  • CPL – cost per lead – is generally not in your control. It’s how much you have to pay to Google AdWords, or Facebook, or an affiliate for a user. It depends mainly on your advertising partner, and the competition for that advertising spot. If competition is fierce, CPL will skyrocket.
  • CPA – cost per acquisition – is a function of CPL x your conversion rate (to registration, or to paying, whichever you choose).

To improve CPA, you need to have an intense focus on improving conversion at every step of the way from initial landing to final conversion. CPL is much harder to tweak.

The equations

I have two equations which I use to explain and analyse Acquisition. Equation I is:

CPA < LTV = good business
CPA > LTV = trouble

where CPA is Cost Per Acquisition and LTV is Life Time Value

Not only is this common sense, but it suggests several potential business strategems:

  • If you can reduce CPA, you are more likely to have a good business. Virality, conversion, harnessing the social graph: these all help
  • If you can’t get CPA down, concentrate on getting LTV up
  • LTV consists on keeping users for a long time (i.e. Retention) multiplied by getting them to spend a lot (i.e. Monetisation).

So Equation I explains why, in a world of rising CPA, I focus my efforts on improving Retention and Monetisation, or the “unknown startup” strategy.

Equation II is about virality. (I know I said virality is dying. It can still be used to shave the cost of CPA, although it is more likely to reduce CPA by 10% than to 10% these days).

Viral coefficient = A% x B x C%


  • A% = Percentage of your users who invite a friend
  • B = Number of friends they invite
  • C% = Percentage of friends who accept the invitation

If you can get a viral coefficient of greater than 1 (and preferably greater than 1.3 to allow for churn), you have a truly viral business. Every new user will generate more than one additional user and your business will grow rapidly. That’s how Skype, Bebo, Facebook, Friends Reunited, Classmates and other social phenomena exploded in usage and value.

Virality does not equal spamming

True virality is not spam. A great example is Skype. If I have Skype, and I want to call a friend in the US, I email him and say “hey, Bob. I’d like to call you but don’t want to pay extortionate long-distance charges. Why don’t you download this piece of software and we can talk for free.”

This is truly viral because:

  • I benefit: I can talk to my friend Bob for hours for free, saving me hundreds of pounds
  • My friend benefits: Not only does Bob get to talk to me, but he can also talk to all of our other friends who are also on Skype, for free
  • Skype benefits: it acquired new users for free, and Bob is now a potential convert for its upsell to SkypeIn, SkypeOut and so on.

In contrast, Zynga was never truly viral. “Hey, come and be a friend on Farmville so I can get a bigger farmstead” benefited me and Zynga, but the benefits to my friend were less clear, particularly if they were not a gamer.

If you want to be viral, you need to make sure that both the inviter and invitee benefit from the invitation. Otherwise you are not viral: you’re a spammer.

The Acquisition game

Acquiring customers is critical to any business. It is also expensive. It can be done simply by throwing money at the problem (which is why I think it is the least important of the three aspects of the ARM yourself framework)

However, it is worth spending some time on it, because every penny shaved off CPA is either money on the bottom line or more money to spend on customer acquisition.

But I can summarise my thoughts on Acquisition like this:

  • Buying customers is expensive
  • Buying customers is necessary
  • Pouring customers into a leaky bucket is stupid
  • Ergo: Retention is the new Holy Grail
  • And the most forgotten

So next time, I’ll focus on Retention.

About Nicholas Lovell

Nicholas is the founder of Gamesbrief, a blog dedicated to the business of games. It aims to be informative, authoritative and above all helpful to developers grappling with business strategy. He is the author of a growing list of books about making money in the games industry and other digital media, including How to Publish a Game and Design Rules for Free-to-Play Games, and Penguin-published title The Curve: