Growth5 Blog

Wednesday, October 28, 2009

An eCommerce Experiment: Pay What You Want

On the one year anniversary of World of Goo, the popular physics-based puzzle game, the game's developers decided to let consumers pay whatever amount they wanted (the hugely successful game's retail price is normally $20). A minimum of one cent was required. The experiment was to last from October 13 - 19.

After seven days of the "pay what you want experiment", the developers posted this blog. Initial data:
1. Sales increased dramatically due to the coverage the experiment received in the media and blog universe.

2. 57,000 people bought the game for an average price of $2.03 ($116k).

3. 13% of the seven day revenue went to PayPal as their transaction fees increase the lower the price of an item is. On transactions of 30 cents or less, PayPal kept the full amount.

4. Almost 17,000 of the purchases (30%) were for one cent. 40% of all purchases were less than a dollar.

5. The developers saw a dramatic increase in sales of their other full-priced games due to site traffic.
In the October 19th post, the developers announced they would be keeping the "pay what you want pricing" going for another six days. This past Monday, they posted this update on the two week experiment:
6.After two weeks -- 83,250 birthday copies were sold for an average price of $3.25 ($271k). Not a bad two weeks.

7. The developers attribute the Week Two sale price jumping by almost $1.00 to their Week One post announcing that so many people were only paying $.01 and that they received nothing for purchases of $.30 or less.

8. Linux users paid the highest average price - just over $3.50. Windows users paid the lowest average price - just under $2.00. Linux users commented that they are appreciative of any developers who take the time to develop for them. Windows users, I suspect, have limited funds due to constantly replacing sledgehammer-smashed computers.

9. The developers created the following "generosity" table by taking the average sale price of each country and normalizing that amount by using the country's per capita GDP. They admit the "normalization" model is flawed, but the table is interesting nonetheless.


Thanks Sean Carton for emailing this story.

Labels: ,


Post a Comment

Subscribe to Post Comments [Atom]

<< Home