GaiaOnline Sucks

Fred Kelson, student in mathematics/economy and Vice-President of Chaos Programming (www.chaosprogramming.com), has agreed to help us against Gaia Online by providing us information on Gaia’s financial state. Since Gaia does never releases any financial data, we asked him to estimate their relative income for us to verify if Gaia Online was indeed making money. The result is presented under the form of an interview, although, it was more his job. Again, I do not take credit for anything said here.

JIM BOOSH, administrator at GaiaOnline Sucks.

Fred Kelson, Vice-president of Chaos Programming

Jim Boosh : Hello Fred. We are here to discuss of Gaia Online financial report.

Fred Kelson: Correct. The first thing I’d like to say is that by remaining privately-held, Gaia Online manages to do two things:

1) They are capable of keeping all profits for themselves while still enjoying the benefits of no taxes.

2) They do not need to emit financial reports, meaning they can hide all their profits from the general public.

Jim Boosh: You mean there are publicly-held companies?

Fred Kelson: Correct. In fact, most companies (of that size) are publicy-held, meaning they do redistribute a (good) part of their profits to public holder. Stock holders, for example, or bond holders.

Jim Boosh: Why would a company do that?

Fred Kelson: Privately held companies are usually going for profits. This is normally utilized by small businesses that cannot afford to go onto the Stock Market, or for big companies that prefer to generate a lot of profits. Publicly held companies are here for one reason: to make a big annual income. 

Jim Boosh: Why so and why do they need to go public?

Fred Kelson: They do it because while they may make less profit, they can keep growing. The main problem of a company is and will always be to get cash for new project. By emitting bonds and stocks, they can keep the interest they have to pay low, while still having enough funds to continue growing. That means that Gaia Online does not intend to grow but to continue to make profits.

Jim Boosh: Are you saying that Gaia does not really care about their growth?

Fred Kelson: Precisely. Gaia has a very large community; and by becoming a public company, they could not only have much more free cash flow, but be able to hire many, many more employees and help serve their many customers much better. 

Jim Boosh: Interesting. Gaia Online states that it can barely pay for its hosting bills. What do you think of that?

Fred Kelson: That is total bullshit. They can not only pay their server hosting multiple times, but also have a large portion of profits.

For this analysis we are going to take pessimistic hypothesis. In the evaluation of discounted cash flows, especially when using Gordon-Shapiro techniques, there is always three types of scenarios: pessimistic, neutral, and optimistic. When a company compares a new project, it will always look at all the possible outcomes; however, I do not have the time nor the interest to make all three. You have to understand that whatever I am saying is pessimistic, meaning that things cost more in my planning that they should and that Gaia gets less than they should.

Jim Boosh: In other words, you take consider the worse possible scenarios.

Fred Kelson: Exactly. There are two types of Cash Flows (Cash Flow is the money that enters and leaves a certain portfolio). For now we are going to consider the negative Cash Flow.

Jim Boosh: Alright.

Fred Kelson: There is no technical support. There is no customer support. There seem to be very limited upgrades in Gaia Online if you consider that they only have to put two new items per month. There are no material costs. Gaia does not need to have offices, it does not need to buy materials to make its products (except for the store, but they are considered in our evaluation). They do not have music, bosses, levels or anything that would require man power. Their forums are on freely hosted PhP site. Their games were designed ages ago and are not considered in our analysis (they are not even complex to make). Also, most content is user-generated.

We are in the situation of a perfect project where costs and work have been reduced to a minimum. I must remember you that we are still taking pessimistic hypothesise. 

Now, let’s calculate their only real costs: the servers. On big-boards.com, Gaia Online claims that ‘’It has 80+ web servers, 11 database servers (4 dedicated to forums), 3 session database servers, 1 memory cache server’’. From the interviews, Derek Liu claims that these servers are

‘’ 1 x Celeron 1.3 w/512k RAM, single HD

4 x P4 2.6Ghz w/2GB RAM, single HD

1 x Dual Xeon 2.6 w/4GG RAM, 4 80GB HD running as 2 sets of RAID 0 arrays

1 x Dual Opteron 240 w/4GG RAM, 4 x 15,000RPM 16GB HD as 2 sets of RAID 0 arrays from Raid Array controller, 1 80GB HD for OS’’

These are quite old news, but we can safely assume that Gaia uses similar type of servers now, although they use more. This is a server similar (and even stronger since these are dual cores servers) than what Derek Liu states :

http://www1.ca.dell.com/content/products/features.aspx/pedge_6800?c=ca&cs=CABSDT1&l=en&s=bsd

The most expensive of these cost 7,316$. We are going to assume Derek Liu got no discount for the number of servers he bought (although a 15-20% discount would be reasonable for someone who buys 80+ servers!). This would make 95 servers at 7,316$ each. Let’s assume each server is 7500$ (a bit more) and that he has 100 servers. This makes it a 750,000$ investment.

Jim Boosh: Really? Wouldn’t the servers cost more?

Fred Kelson: 7,500$ is already too much. Hell, if I was realistic, I’d say that his 80 servers cost 2,000$ each. The server linked above could probably run a World of Warcraft server alone without lag. And I must remind you that this is a forum and some flash animation, not a full-scaled, MMO developed game. I am only taking the worst scenarios; I highly doubt Gaia’s server costed 750,000$. 400,000$ would be reasonable.

Jim Boosh: What about the server management?

Fred Kelson: If by that you mean making sure the servers are always connected and restart the servers that bug, then it’s mostly 500$ per month tops (300$ in electricity and 200$ in maintenance). Hell, the most expensive maintenance is probably to clean the top of the servers. Don’t believe rumors of servers costing millions.

Today, you can get a pretty solid dedicated server for 299$/month, at least one similar to one like Gaia Online has. That is, in my opinion, the same cost if you choose to lend a 12,000$ car (such as Honda Civic). However, the 299$/month counts all the money spent into engineers, handling, insurance, profit margin, marketing, meaning that if you purchase the servers and handle them yourself, they will indeed be much, much cheaper. On the 299$/month, maybe 125, 150 dollars go to pay the servers. Finally, you must remember that like cars, servers have no value after 10-15 years simply because technology will have evolved so much. This would be like running a MacIntosh today, it’s completely outdated. 

Also, but by buying the servers, Derek Liu saves a good bunch of taxes.

Jim Boosh: Interesting. Continue!

Fred Kelson: Let’s go back to our 750,000$ investment and let’s assume it is true (I have just shown it was way overpriced but let’s assume it was not. Also, one count argue all the servers were not bought at the same time but rather over time, leading to even more taxe saves). Assuming MACRS 5 years depreciation (any good company will use this, especially on this short-handled project), Derek Liu will be able to save a lot of taxes. First, since he gets commercial taxes, he only get 30% tax rate and can also deduce all servers cost from his tax report. Second, he does not have to pay the whole 750,000$ and can use MACRS depreciation over five years, leading to tax credits of around 40,000$ per year. To conclude, this makes HUGE taxes saves.

Now, let’s go on positive cash flows. To proceed, we need to establish a few hypotheses. Here is every single of them with a description of whether or not it is acceptable.

H1: We are going to assume everything said by Gaia Online is true to their last word. That seems reasonable because they already say very little about how much they make, so why would they lie?

H2: We are going to assume that Gaia Online is a medium-risked company and use a company internal rate of return of 10%. This is a bit too much since risk-free rate is around 4%, but for the sake of argument, let’s stick with this one

H3: We are going to assume every item is purchased at the end of the month to simplify the maths. This is wrong but it decreases the positive cash flows and thus doesn’t go against our calculations

H4: We are going to assume Gaia online lives forever. There is no reasons why it should close and this is a good approximation if Gaia remains at least 10 years. This seems reasonable because nothing indicate they are going to crash

H5: We are going to assume servers have no residual values and are simply thrashed at the end of their life. This seems a bit underrated but again goes again decreases positive cash flows (since we ignore what happens with them).

So let’s begin with the main source of income: advertising. There is an average of 50,000 players logged at any time of the day and night. Because of the large quantity, the IP redundancy is ignored, meaning each page loaded loads a different page.

We are going to assume a 0.10 CPM (cost per 1000 impressions), a VERY low price. If I could advertise on such a popular website at this price, I wouldn’t hesitate one second. I calculated this after evaluating the position and number of advertisements. 

From the 50,000 players, we are going to assume 20,000 (only) are in forums talking, others being playing games. We are going to ignore publicity placed on PMs. 

We are going to say each of those 20,000 players loads one ad per 5 minutes. This seems pretty low because I highly doubt anyone would spend 5 minutes looking at a thread doing nothing. Finally, we are going to say 70% of these players have nothing to prevent ads loading (that’s the real percentage in the population). This is a bit low since most are children and have no idea how to block ads, but oh well.

So this makes : 20,000 x 0.75 x 0.10/1000 x12 (number of ads a regular player load per hour) x24 (one day) x 30 (a month, all months have 30 days)

=12,960$ per month through advertising.

Now, for the monthly collectibles. A bit harder to estimate, a correct formula is:

6,750,000 (current players) x 0.10 (still playing) x 0.01 (one person out of 100 will buy a monthly collectible, very low estimate, I can hardly see a company selling 1 items for every 100 customers, but oh well!) x 2,20$ (cost of the donation after the handling fees of from donations

=14,850$

Gold selling: 

6,750,000 (current players) x 0.10 (still playing) x 0,001 (one person out of 1000 will buy gold which is kind of stupid, but we are going to assume less people buy gold) x 9,56$ (We assume average order is 10$ minus a 0.44$ fee, which makes a 20,000g order, low again. This was a bit harder to get and is a very rough approximation. I suspect average gold order is most likely at 20-25$, who would take this much time for only 20,000gold?)

= 6,453$

Again, there is more than 600,000 players playing, but we are taking pessimistic hypothesises.

Store (counting the ads):

This is a bit harder to estimate, but I am going to go roughly

6,750,000×0.10×0.001×5$ (average profit for an item sold, meaning a plain t-shirt would cost 10$ to make and ship) and their “ad system’’ giving a plain 5$ profit, transit fees excluded)

=3,375$

Which makes a total monthly profit of : 37,638. This is where it gets fun. We are going to estimate the NPV (Net present value) of Gaia Online as it is:

Note that we chose 0.10 ANNUAL RM rate (company rate of return); this gives monthly interests of:

(1+0.1)=(1+ i(12)/12)^12 è i(12)= 0.7974%

Using the perpetuity formula:

NPV = -750,000$ + 37638/0.007974 = + 3,970,007.17$

Or, if we compute the IRR (internal rate of return) 

0= -750,000$ + 37638s_12[1,04^(1/12)-1 ]/(i)

Where 4% is our risk free interest (meaning money is reinvested at 4%, which is odd because it’s easy to reach 6-7%, but again pessimistic hypotheses), s_12 the accumulated value of a 12 months annuity, 1,04^(1/12)-1 = 0,327373978% the monthly-calculated interest corresponding to a 4% annual rate of return and ‘’i’’ our IRR. 

We get i=61,3170299%

Meaning that a person putting 1,000$ into Gaia Online would get 1,613$ at the end of the year. This is simply crazy and is from far one of the strongest IRR I’ve seen.

Jim Boosh: Are you saying Gaia Online pays a lot?

Fred Kelson: A lot isn’t even the good word! I used the worst hypothesis ever made, such as 10% of players active (this would make it the less active forum ever made), and only 1% of people buying monthly collectible (this would make it the worst company ever) and I STILL comes to a NPV of +4 millions! Also, I calculated super expensive servers, meaning that the reality is probably close to 1.5x this, possibly even double!

Jim Boosh: According to you, how much does Derek Liu has?

Fred Kelson: Easily over 10 millions dollars.

Jim Boosh: What do you think of: http://mashable.com/2006/06/19/gaia-online-gets-893m-for-anime-community/

Fred Kelson: If it’s true, then this would be the biggest joke ever. 

Jim Boosh: Do you think advertising could pay server upcost alone?

Fred Kelson: We can see advertising brings about as much as donations (although I suspect more people than 6,750 buy collectibles), but I believe that 10.000$ is already enough to pay for servers. If you use Gordon Shapiro again:

NPV: -750 000 + 15,000$/0.007974= +500,000$

We still get a very positive NPV. That’s easily enough to pay for all future needs Gaia might have in server maintenance. So yes, advertising does pay for servers.

Jim Boosh: Why does Gaia shows so little advertising?

Fred Kelson: What do you mean so little advertising; they have it at the entrance of every forum and in each PM form. They could hardly put more without having customers pissed at it and moving to a new forum. Plus, it’s a commonly known fact that the more ads you put on a page, the less CPM each one pays. Advertisers love to be alone putting their messages on a page. Otherwise, the user will simply ignore it. Plus, the ad is very strategically placed. 0.10$ CPM seems very low considering these are banners and not text links and considering their targeted audience. Hell, for each 1000 persons that see an ad, you pay ten cents. If advertising was really like this, chances are that people wouldn’t even bother putting ads. On Chaos Programming, I used to get 1-2$ CPM, and it is much, muchhh smaller, with text ads, and with a single ad at the top.

Jim Boosh: Very interesting. So would you say Gaia Online lies?

Fred Kelson: They do not lie since they do not say anything, but even with pessimistic hypothesises, we can easily see they do gain hundreds of thousands every year out of little kids wasting their parent’s money.

Jim Boosh: Thank you, mister Kelson.

Fred Kelson: My pleasure.

Fred Kelson is working towards his CFA while still studying for his MBA. He is also studying to get a master in economy and works as CFO for Chaos Programming.

NOTE (1): In this project I spoke of the IRR only in a comparative manner. What I meant to say is that if someone spent 750,000$ today and got the popularity of Gaia Online, he would get an IRR of 61%. Note that most IRR are at around 15-20% maximum. However, it must be noted that no one knew how Gaia would grow before it started and that these are very hard to estimate, therefore our calculations are used in a retrospective manner. This means that Gaia is indeed making a lot of money but that it was very hard to foresee three years ago.

NOTE (2) I’ve tried to calculate the value of Gaia Online Stocks assuming it was emitting Stocks. Since it does not grow anymore, it would need to distribute every profit in dividends. If they were going on Stocks, they would do a lot of tweaking to increase Gaia Online rentability even furthermore (like World of Warcraft did). However, my calculations ended there, with a g= 0% and the obvious difficultly of estimating the number of stocks emitted. Also, where would the new funds go? Gaia could indeed improve its services a lot and also grow a lot, purely throught advertising, but Derek Liu does not seem to be interested in it. This would, however, be a good project to invest money in. I know I would buy Gaia Online Stocks if they did emitted stocks. However, being privately owned, they prefer to keep all profits for themselves, at the risk of not being able to grow. The money wins over the stuomers

NOTE (3) It should be noted that I’ve used pessimistic hypotheses all over this analysis. In reality, as I suggested, the cash flows are probably 1.5x or even 2 times what I said, giving an even higher IRR.

JOIN OUR FORUMS!