Online Mind Mapping and Brainstorming

Create your own awesome maps

Online Mind Mapping and Brainstorming

Even on the go

with our free apps for iPhone, iPad and Android

Get Started

Already have an account? Log In

Metrics for Virtual Goods businesses by Mind Map: Metrics for Virtual Goods businesses
5.0 stars - 1 reviews range from 0 to 5

Metrics for Virtual Goods businesses


Daniel James

Three Rings

Andrew Chen

intro & terms

some key terms explained

e.g. LTV = life time value

introduction to Whirled

rooms, items, trophies etc

soft currency = earned with gameplay

hard currency = bought with real money

third currency = earned by content creators when content is sold for hard currency

Puzzle Pirates

roughly 4 M$ revenue

started in 2005

male/female 50/50

wide age range (7-70 year olds)

key metrics, see flickr photo of slide,, conversion rates vary a lot of depending of the source of the user, top 10% of paying customers generate 50% of revenue

what factors drive LTV

games are rare as they make money from end customers directly

so you can actually buy traffic

see flickr photo of slide

cost per acquisition

the key figure to understand

factors that drive it, flickr photo,, ad networks, plenty of sources, diversify, user requirements are crucial, installs may drop conversion from 20% to low single digits, flash trumps java hands down, 90% of people lost in Puzzle Pirates (Java), almost none lost in Whirled (Flash), audience/theme: horizontal vs vertical, horizontal (e.g. a socnet) works fine, vertical (niche) challenges in finding in audience, buying ads, scaling, funnel design, must be testing all the time, A/B & multivariate, all steps, viral marketing, gotta have it, especially for Facebook games, potential to really drive down your cost / acquisition, not just growth, focus on retention, otherwise your newly acquired customer's LTV might really low, A/B testing, pretty much the most important thing, test your hypotheses


understand this and optimize it

profit = LTV - cost/acquisition

customer retention metrics

sample dashboard

if your LTV > cost/acquisition, just go and buy traffic

puzzle pirates

CPA campaigns in march, looks like the traffic wasn't really worth anything, e.g. $4 CPA led to $1 LTV

have weekly figures for all metrics


cohort analysis, how fast do customers drop off,, at Whirled haven't done a lot to prompt retention, e.g. missing notifications

revisit rates, track, day-by-day basis, how many times do they come back, factors that drive it,, product depth crucial, need to have it, community, social ties get you back there

conversion / retention

whirled metrics


Whirled, three currencies, aggressive repricing needed, user's generating content do this well at Whirled, money booking unclear

Puzzle Pirates, has item decay, as no money payout, book the money as it comes in, no liabilities, players value their time at 25 cents / hour

spending / earning

at first gave out too much money out, changed this, try to balance the in and out flows

creators / purchases

item listings by type

item purchases by type

by money spent

avatars, furniture & toys

item purchase metrics

total coins spent, shows low end and high end products all earn well

purchases vs sales

puzzle pirates

coin shop purchase, privilege items, e.g. "captain's badge"

item sales distribution

different view on the same data as in previous slide

160-170 k$ a month in product sales, mostly vanity items

billing payment breakdown

some figures are obfuscated due to confidentiality agreemetns



Target card

use somebody else's infrastructure





if you go for kids & want lots of small payments

must have prepaid card

older people with big onetime payments

credit cards


always do a refund, otherwise problems with credit card companies

verify big purchases, e.g. fax in signatures


sample model of virtual goods LTV

search vs. display

people who are searching for you (in some way) will convert a lot better, use google adwords for search

analytics tools

typically own built tools, lots of data, hard engineering problem

facebook specific, SoMetrics, developer analytics, Cotigent??