News, research and discussion on virtual goods, currencies and economies globally.

Update on Chinese Gold Farming

The following represents some notes and comments building on my earlier research report on gold farming. It draws mainly from a couple of interviews kindly given by Anthony Gilmore, who has been filming in China for his forthcoming documentary, Play Money: Also from Rowenna Davis' earlier article that interviewed both myself and Anthony.


A. Overall Prevalence:

Estimates of 400,000 people working in China on gold farming/trading are seen as very realistic, and the number could be 500,000 up to 1,000,000 (the latter figure is also estimated in Nick Ryan's recent articles on RMT, which claims total turnover for Chinese RMT operations of US$10bn per year). There are many brokerages in all major cities in China (Ryan's article estimates 60,000 in total).

B. Brokers and Farms:

Brokers (i.e. those doing the trading of currency/accounts/levelling) employ English-speaking graduates, and so locate within main cities (those graduates would not go to work in more out-of-the-way places). These are seen as legitimate customer services businesses.

Gold farms employ those with just high-school diplomas or below, and so the farms are put where they can get space cheaply; typically some way outside a city or in smaller towns/cities. Of course, there are limits to this: they still need a broadband connection, so this goes down as far as medium-sized towns but not yet to small towns and villages.

There is typically an n-to-n relationship: each gold farm ("workshop") serves multiple brokerages; each brokerage sources from multiple gold farms. (Though there are also cases where brokerages own their own gold farm.)

The relationships to date have been almost exclusively in-China. However, some brokerages were starting to look to other locations outside China for gold farms, such as VietNam and the Philippines. It is not clear if this is to reduce labour costs further, or to link up with sources for games that are particularly popular in those countries.

C. Typical Brokerage Size and Division of Labour:

In Changsha, one of the main brokerages (with a turnover of around US$1.5m per year) employs more than 130 staff: 20 people doing ads in game and email spamming; 4-5 on customer services with whom players place orders. These numbers are doubled because they work two shifts in a day. There are 5-6 in the IT department sorting out servers and Internet connections. Then there's the delivery department who interact with the gold farm and deliver the money in-game. They contact the gold farms to find out how much money they have banked on particular servers, and will take amounts until they can fulfill the customer order.

There are some gender divisions of labour: most customer services staff are women; almost all gold farmers and technical staff are men.

D. Gold Farmers and Other Workers:

The gold farms Anthony saw were full of young guys aged around 18-20 with beers and cigarettes, hanging out playing games and having a good time. They realise it's not a long-term occupation but they also appreciate that they are part of something that is rather new and different. His estimate is that they'll stay for maybe 1-2 years before moving on to another kind of job.

Typical earnings, etc are in line with earlier estimates: around US$140 per month with food and accommodation provided, working c.10 hours per day. It seems there are two 10-hour shifts per day, with some down time allowed for IT maintenance, though the notion of a strict shift period isn't present: start and end times are a bit fuzzy and the farmers generally have a target of gold to earn rather than a set number of hours; that target varies depending on market conditions. Some of the gold farmers are killing mobs or gathering resources, but others are making money by arbitrage in the auction houses or trade channels on WoW.

Customer service staff in the brokerage are paid more like US$350 per month; they typically deal with 100 or more customers per shift. Managers earn significantly more (unfortunately no figures available).

E. Customers:

Brokers have a mix of Chinese and Western customers. Some brokers are wholly World of Warcraft-dedicated; others work a mix of Chinese and Western MMORPGs. Typical purchases are for around US$25-worth of in-game currency. (Highest reported purchase was for more than US$3,000.)

F. Bots/Automation:

There is quite a lot of use of automated software in the gold farms, to automate fighting or resource gathering. At least some of this is written in China, bought by the gold farms and then the code is adapted by gold farm staff.

Photo credit: Anthony Gilmore

Average Wage

To put this in context, if we accept this report into wages on the following link:

As reliable (if).

Then the average Chinese wage for a city worker is about $190 a month and the average wage for a rural worker is about $58 a month.

Given that the average wage for a gold farmer is reported at $140 including accommodation (generally a substancial part of ones wages) and food bills, it would seem that the average gold farmer worker isn't too badly off at all in comparitive terms.

The customer service representatives (who, I'm assuming are employed because of their stronger language skills) being on $350 is earning almost double the "average" chinese worker salary.

I have issue with these averages though (as I'm sure anyone who works with statistics will tell you) in that they are too broad to really give an impression of wages in a country which is as vast and diverse as China, and thus the "average" hides probably high levels of regional variation as well as the city/rural divide.

In all though, gaining the "average wage" for sitting around doing 10 hour shifts using a computer as an office worker has to be slightly more appealing to these young men (greater ultility) than the alternative of 10 hour shifts in a factory to gain about the same wages.

I'd be interesting to see what the opinions of the managers is on the effect of the recession on gold buying is though. So far most commentators have said that the MMO subscription market is expected to be highly resilent to recession. If an MMO product is seen as a luxury expenditure (and thats up for debate to start with!) then gold buying must be the ultimate in un-necessary entertainment expenditures that we could do without (again, up for debate) and so it is of interest to see if this "shadow market" if you like is effected at all.

David Grundy

vw biz and recession

If an MMO product is seen as a luxury expenditure (and thats up for debate to start with!) then gold buying must be the ultimate in un-necessary entertainment expenditures that we could do without (again, up for debate) and so it is of interest to see if this "shadow market" if you like is effected at all.

Thank you david, good points.

There has actually been a significant amount of articles on how virtual asset sales and "virtual business" are thriving (or at least not declining) during recession.

Granted, people might spend less on luxury goods, goods that have a lot of added surpluss value on top of core functionality. You hypothesized that gold is an ultimate luxury good, if MMOs were a luxury product. I disagree with this notion to some extent. How gold is spent in performance oriented MMOs, such as WoW, does not correspond well with the nature of luxuries goods, as the gold is commonly spent on highly functional items, like high-end raiding. Gold is also relatively cheap, being "manufactured" in the east.

As you said, when looked at from "real world" perspective, playing an MMO and buying gold is still in a niche and might be perceived as luxuries to some extent. For this niche though, it might not always be very luxirous. Take for example power-leveling services. One might be seeking to buy these services to have more time to earn real dollars.

In online-hangouts though, purchasable items tend to be more "hedonistically" oriented or have social functionality (see Vili's post about virtual items properties), lacking explicit functionality. Thus intuitively these could more easily be perceived as luxury goods, but the core argument for virtual business still thriving is that, people are spending less on real world luxury goods and actually substituting them with (luxuries) virtual goods.

A couple of entertained arguments:

- escapism

- virtual worlds and goods are relatively cheap entertainment

- virtual goods are cheap substitutes (at least for products lacking explicit functionality)

- companies reducing meeting costs by using virtual worlds


Some articles:


Gold Farming/RMT and Recession

It's about as ropey as most data on gold farming but the firm that Anthony focused on has been growing not shrinking during the recession, and the same claim is made in Nick Ryan's article.