Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
August 21, 2017
arrowPress Releases






If you enjoy reading this site, you might also want to check out these UBM Tech sites:


 
In-app purchase pricing strategy: Do bundle discounts work?
by Mark Robinson on 09/21/15 02:00:00 pm   Featured Blogs

1 comments Share on Twitter Share on Facebook    RSS

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.

 

Premium currency bundle discounts, i.e. incentivizing players to purchase more currency by offering a better per unit price, is a commonly used strategy in F2P games. However, there doesn’t seem to be any consistency to how they are used. Some games offer only token discounts of <10%, while others can offer discounts of 100% and more. IAPs are virtual goods that have no unit cost, and so any discount can be applied without incurring loss, however, give players too much, too cheaply and they will not have to spend again.

Given this, it is worth asking the question, do bundle discounts even work? Or, put another way, do they encourage players spend more real currency to grab a virtual bargain?

Using statistical analysis to determine if bundles work
To investigate this, we looked at a small selection of 19 F2P games on our deltaDNA game analytics & marketing platform. These games were hand chosen to ensure that they used premium currency bundles. They were games that:

Did not have complicated starter packs and timed deals that might distort the results
Did not have an obvious virtual item that would require the purchase of a large amount of currency

By restricting the games in this way, we can be sure that the distribution of bundle purchases is primarily motivated by the discounts applied.

Normalizing currencies
The games chosen all used different price points and ranges, with the base price ranging from $1-$8 and the most expensive bundle from $50-$200. In order to put them all on a level playing field so they could be compared, we normalized all the virtual and real world amounts to factors of the cheapest offering. This means that if the base bundle was 100 coins for $2 then a $5 of 300 coins would be 2.5 RC and 3 VC, where RC stands for normalized Real Currency and VC stands for normalized Virtual Currency.

Calculating whether discounts work
The effectiveness of bundle discounts is not just a function of how big the discount is, but of how much more you have to pay to get it. A 10% discount for paying an extra $1 is likely to work better than a 100% discount on a bundle that costs $100 more. As a result, we need to define discounts in terms of %Discount per RC.

The aim of discounting is to increase the average transaction amount, so if bundle discounts work then we should see a trend of increasing avg. transaction amount for larger discounts. For each of the 18 games we calculated the avg. transaction amount (in normalized RC units) as well as the avg. Discount% per RC across the price points offered. The chart below shows the results of this.

Avg. transaction amount (in normalized RC units) & the avg. Discount % per RC across the price points for 18 F2P games

 

There is a clear trend of increasing transaction value for games that offer larger discounts.

Offering a discount of 2% per RC results in the average transaction being worth 2 times the base price, while offering 6% per RC boosts the typical spend to 7 times the base price. Putting this into a real world example, if my cheapest bundle was $2 for 100 coins, a 2% discount per RC means that my $5 bundle would contain 263 coins (5% extra), while a 6% discount per RC would make my $5 bundle contain 288 coins (15% extra). Meanwhile the avg. transaction value would be $4 dollars for a 2% discount per RC, and $14 dollars in the case of the 6% discount per RC case.

At first this result seems remarkable, but it is worth noting that most of this ‘boost’ to transaction value is from increasing the number of purchases of the most expensive bundles. If we class bundles that cost more than 10 RC as ‘high value’, then the least generous games only get 5% of their transactions from these, as opposed to 10% for more generous games. In others words, a game with <3% discount per RC is twice as likely to get a spender buying a high value bundle than those with less generous discounts. Again, taking our real world example, a high value bundle would be $20 or more and have a 30%-60% discount applied.

The verdict: Bundle discounts increase the value of individual transactions
It is clear from our analysis that bundle discounts are extremely effective at increasing the value of individual transactions. There may be negative impacts which we have not considered, for example, offering large discounts reduces the need for future spending, as well as discouraging first time spenders who simply cannot afford higher price points. For these reasons it is likely that the games we have studied do not contain extreme discounts, with the largest discounts seen only around 100% for most bundles and which can cost over $100 (50 RC).

Putting aside these concerns and acknowledging that the gains from discounts seen here across games may not be repeatable within a single game, spenders do clearly respond to IAP bundle discounts by increasing their spend value. Using them effectively is another valuable strategy in the quest maximize revenue from F2P games without negatively impacting player experience.

This post first appeared on the deltaDNA blog.

 


Related Jobs

Wargaming | BigWorld
Wargaming | BigWorld — Sydney, New South Wales, Australia
[08.21.17]

Scrum Master
Wargaming | BigWorld
Wargaming | BigWorld — Sydney, New South Wales, Australia
[08.21.17]

Software Engineer, C/C++, Game Engine Development
Phosphor Games Studio
Phosphor Games Studio — Chicago, Illinois, United States
[08.20.17]

All Positions
Phosphor Games Studio
Phosphor Games Studio — Chicago, Illinois, United States
[08.20.17]

Programmer





Loading Comments

loader image