Data Driven Localization

Data Driven Localization:  Achieving Maximum ROI

Andy Johnson, Principal

Selecting which languages to localize your game to can be a daunting challenge. Not every language will give a positive Return On Investment (ROI), so which do you choose?  Every language selected represents a significant investment in time, talent, and cost.  Do you commit valuable resources and budget to shoot for an elusive but promising international community, and if so what languages do you target to support the most paying players?

The decision of which languages to choose for localization is often made in one of the following ways:

1. Go with the Industry standard – FIGS
2. Localize for potential sales
3. Localize to languages player request
4. Support the language selection from a previous release

While these four options are the most commonly followed, there are fundamental issues with each of these routes:

1) The Industry Standard – FIGS

FIGS, or French/Italian/German/Spanish, has been the traditional (console and PC) gaming gold standard language set to localize for many decades. Originating back in the 90’s, FIGS were a dependable language set for return for several reasons, primarily:

• Readily available localization, with localizers all close to the same UTC time zone
• Relatively efficient from a language standpoint, as most alphanumeric characters from these languages are grouped into the same Latin character sets (other than the troublesome German sharp S – ß), reducing relative impact when implementing
• Companies tend to have offices in countries where these languages are spoken
• The languages can be easily exported to additional regions such as Latin America, although there are local language differences, to spread costs

When FIGS were standard operating procedure, the global economic landscape was very different. The world has not stood still since the early 90’s and some more traditionally emphasized countries and regions have seen a sharp decline in recent years in regard to relative ROI, while others have gained in importance.

2) Potential Sales

Sales goals are often set by local sales teams, which are then communicated through channels to drive profit and loss expectations.

Unfortunately, these goals are not always set in a rigorous manner. It’s possible that sales teams will inflate sales estimates to force localization, as a localized product will sell more in their territory and help them hit their goals (reach their targets and get their personal bonuses).  Having a localized version will have a significant impact on sales.

These decisions are often made in a silo and maybe in the best interest for the local office/sales team, but may not be ROI positive when the entire costs and true revenue potential are considered, especially relative to other countries. As noted in Breaking the Fear Barrier by Tom Rieger[i], pressure to reach local or individual goals can lead to this exact type of behavior, and often does.  This can lead additional language versions which impact development teams (an unseen cost) and are ultimately not profitable.

3) Player Requests

Anyone who’s been involved with any gaming discussion group, forum, or social media thread will know that there is a very vocal minority that are extremely active.  Other than the vocal minority most other player’s interactions with online discussion groups are minimal – the majority of players either don’t know about, don’t visit, or don’t read discussion groups.

Since most studios cannot support every language, you are most likely ignoring player dialog for the languages you do not support, especially dialog that is happening in unofficial local-language discussion groups.  Those unofficial channels, however, are the forums where players may be requesting localized content in their language.

Consider the following, based on observations working in the industry over the past couple of decades:

• Of the active players let’s assume 10% participate in some way in online discussion groups.
• Of these players, most are looking for help, tips or solutions to technical issues, and therefore are passive and fleeting visitors.
• Of the 10% of visitors, maybe only 10% will post.
• Of that 10% there are likely only a handful of engaged, vocal players who are regularly actively posting on a variety of topics, and at frequent intervals.
• Of this small base, in almost all cases there are very, veryfew non-English (source language) speakers posting, as discussion groups are usually on non-localized forums / not in their native language.  As a result, Customer Relations team members are often English only, as this is the supported language.

This all equates to a troubling result.  Potentially less than 0.01% of your player base (optimistic) are vocal international players, likely who are English literate (to engage in the discussion groups and player your, non-localized to their language, game), asking for localization in their native language.

This is a very small, very select group of players.

4) Legacy Decisions

Another option is to stay the course and continue to only focus on previously localized countries and regions, which most likely, used one of the three other methods discussed above.

Releasing a sequel supporting only a previous language set could mean you are building your next game on legacy decisions backed by outdated rationales, leaving millions of dollars of potential revenue on the table. Economies change and develop.  Player preferences may shift over time.  As obvious examples, FIGS would ignore Japan, China and Korea, some of the highest revenue generating countries in the world.  Other countries such as Russia and Brazil are rapidly developing, and India is an emerging market.  And, different genres have different appeal in different cultures.

A More Disciplined Approach

Language choice decisions can greatly benefit from the insight given from data and behavioral sciences – factors such as demographic trends, industry penetration, player segmentation, player journeys within your current base, infrastructure, and competitive activity are but a few examples of the meta factors that can influence how you prioritize your localization decisions.  Understanding not only where your players are but also which players have a higher conversion rate (spend) will help inform your localization decision.  In addition, there are “soft” factors to consider, that may require input from cultural anthropologists or other behavioral scientists with experience in those markets.

Using these and other methods, our team has experience in teasing out which languages and territories to support, with proven success.  By viewing player activity and interactions we can help support your international release by ensuring your future language decisions are made on solid ground, supported by data.

Contact us for more information on how we can help.

 

[i] Rieger, Tom (2011).  Breaking the Fear Barrier:  How Fear Destroys Companies from the Inside Out and What to Do About It”.  New York:  Gallup Press.