There’s been a lot of talk recently about how Star Wars: The Old Republic was going to be BioWare’s most expensive mistake, ending up consigned in the footnotes of history as a video gaming failure. It’s easy to see where this comes from – at a $200 million price tag this is one of the most expensive games yet made. Yet it will, in time, make a profit.
When looking at console games the main driver is to make back the cost of the game in boxed sales. An XBox 360 game that shifts 2 million copies globally equates to roughly $120m in sales, of which the publisher is likely to see roughly $30m. If your game comes in at less than this budget or you sell more copies than the forecast then you’re into profit territory.
MMOs break this mold by placing some of the budget for the game against forecasted subscription, micro-transaction or “in-life” revenues. The Old Republic pushed this further than most, assigning a huge chunk of the project budget against these in-life revenues. It’s a bit of a gamble, which is why building up preorders is so important. But in this case, it should also pay off.
Looking at the first month sales we know that over 2 million copies of SWTOR have been sold. What we don’t know is the proportion of people who bought the boxed copy, Collectors Edition or Digital Deluxe from Origin. This is important as a publisher only tends to receive about 25% of the retail RRP for boxed copies but can earn much more from an online store offering digital downloads. Assuming that EA earned a minimum of $15 per sale, earnings for the first month would be around $30 million at a minimum. That leaves EA with around $170m to bring in from subscriptions.
Last week I took a closer look at how many subscribers are likely to stick with SWTOR after the first month, coming up with an answer of roughly one million. Assuming these subscribers stick around for the whole year that’s a further $180m earned from monthly subscriptions at $15 a time. There are currency variations and multi-month discount deals to factor in, as well as people who prefer to use game time cards, so overall it should even out.
Offsetting against this income are operating costs. SWTOR uses three datacentres located in East Coast USA, West Coast USA and Ireland in order to provide players with servers near to them to play on. That said, it’s unlikely that these facilities would be dedicated to supporting the MMO, instead providing support for Origin as well as console-based multiplayer sessions. The entire EU server cluster is likely to squeeze into 10 to 15 racks, with each rack measuring 80”x40”x24”. A typical large datacentre costs $10m a year to run, but colocating in existing EA facilities should bring operating costs down significantly.
There’s also customer support requirements. Again these are likely to be an expansion of existing EA facilities rather than dedicated BioWare installations, although SWTOR support might be made up of a dedicated team. Even with a team of 100 staff on an average salary of $30k per year the overall support cost would be $3 million per year. All in, the annual run and operate cost for SWTOR should be significantly less than $10 million and probably be in the $4m to $6m mark.
To recap, a $200 million game that made $30 million in first month sales has $170m still to recoup from subscriptions. SWTOR is likely to retain 1 million subscribers, earning $180 million in the first year. Operating costs are likely less than $10 million.
From my crude napkin-maths, Star Wars: The Old Republic should break even by the end of the first year, with anything beyond that being healthy profit margin. It’s possible that the forecast was for the game to pay back within three years, in which case it’s easy to see that the game should make a substantial sum even if subscriber numbers drop.
As with any rough calculation there are assumptions. The first big one is not taking payment processing, debt refinancing, currency exchanges and other fiscal operations into account. The second one is the cost of ongoing development and maintenance. it’s not clear how much future content was ‘baked in’ to the $200m cost, or if BioWare will be looking to dip into some of their subscription earnings to fund further development.
It’s likely that we’ll never know the real numbers from BioWare, as these costs tend to be competition and marketplace sensitive. Either way, it’s clear to see that BioWare have produced a game that stands up to financial scrutiny as well as critical acclaim.
As someone who works in Capital Markets, it’s interesting to see a video game blogger use the words “debt refinancing” correctly.
Great job as usual Gaz!
I think having a million subs lasting a whole year is a bit of a stretch but never say never however you could easily say that new purchasers would help to contribute to sales of course. I’m also not convinced of the running costs of a dev team for a year or for producing new content, as said thopugh a lot of that you can include in the initial out lay of $200m but it’s really just pulling numbers out of thin air right now unfortunately.
The one glaring omission however is the licensing fee to LucasArts, which varies a lot depending on what you read and no doubt we’ll never know. Regardless though, Lucas would not let it away for cheap and there’s no idea if it’d be a lump sum per year or a percentage of profit.
Will SWTOR make profit? In the long run yes. But there’s a whole lot of variables really.
Only truly dense people would believe that TOR is going to crash and burn like say, Tabula Rasa.The real question is whether or not TOR will be a great enough success to show the industry that the traditional big iron style still works, throw as much money at the project you possibly can, DIKU mechanics, subscription fees, the works.
Never mind that these things have only really worked out for a mere handful of games compared to the number of empty husks left on the wayside.
If the TORhog sees its shadow (or rather, the shadow cast by its large pile of money) be prepared for another 7 years of stifling bullshit from the MMO industry. Personally I’m hoping the forecast is overcast and rainy.
From my crude napkin-maths […] != […] that stands up to financial scrutiny 😉
Gaz, sheer curiosity… do we have any numbers to give indicators about subscriptions *other than* box sales? Also, no mention whatsoever about the cost to use the SW IP. Napkin math would be fine for any other title, I feel, but GL trying to turn a buck on G.L.’s product (oh, the puns!).
I don’t think it’s ever been a question of ‘how’ will the dive break even, but ‘when’. I can’t say that the devs weren’t taken aback by the complaints at launch, given the amount of post-release PR (and ‘unsub’ button going missing). Lastly, you use 200m as a figure, and while it might very well be the true amount, it is on the lower end of the projected cost gamut, not the mean. Without a doubt, SWTOR’s launch performance disappointed the backers, and with the exception of WoW/FFXI/EVE, MMOs don’t improve after launch, they decline.
OK, here goes.
The number of subscriptions was taken from my previous article about the SWTOR retention rate. There’s a model that was put together about six years ago that proposes a retention rate of 40% to 50% for any new MMO after 2 months, and there’s no reason to suspect that SWTOR will be any different. Have a read if you’d like to know more about how I came to that figure, as it includes mention of both the positive and negative effects of events leading up to renewal day.
There’s this article from the LA Times that refers to the development cost, but there are a couple of questions that it raises – is marketing included and does it include the LucasArts IP cost? Are there royalties on box sales or monthly subscriptions as well? It’s an unknown.
My personal feeling is that SWTOR was budgeted with a 2 year payback model, which makes things balance out with a much more flexible margin.
Also, I’m not sure if the box sales will disappoint all analysts. Some of the analyst feedback quoted in the LA times article says that SWTOR is still on track to meet predictions. Even looking at the VGcharts PC format sales chart has SWTOR landing in at 36. In the context of the current PC market, that’s pretty good going.
[i]with the exception of WoW/FFXI/EVE[/i]
And Lineage, and Lineage 2, and Everquest, and DAoC, and CoH/CoV, and UO, and probably some other things I’m forgetting. Assuming you’re speaking of # of users anyways. The whole “explosive release followed by an equally catastrophic nosedive into oblivion” thing is mostly a trend that started occurring after WoW distorted the whole western market into revolving around itself.
Also, 200m is the figure EA gave out just the other day when asked about the games budget. So Gaz ain’t guessing about that.
Actually, Nick Yee’s research was based on industry insider feedback checked against WoW subscriber retention rates, so it includes that game. The PDF is well worth reading, even though it’s a bit weighty. It’s apparently something that all MMOs seem to go through. The critical thing will be where they go from month 2 – there’s apparently a year’s worth of content waiting in the wings, but there’s a ton of player concerns about systems like the lack of LFD and the GTN (auction house) interface. If BioWare can deliver on these and keep the content flowing then it’ll be interesting to see how the subscriptions trend.
@ A.L.
For clarity:
a/ The three titles listed are the only three to have, in Western markets, continuously grown through their lifetime following the initial product sub fluctuation (discounting very recent history).
b/ Different sources give different numbers, and corroborating is important. 200m being all-inclusive, just the games [development] budget, does it include the ‘over-budget’ sums etc?
@Gaz
Apologies, I had read that article but will revisit. I just remember other articles, namely the Morhaine one describing how 1/4 of their players play past lv10 and sub and how this was ‘a big deal’ compared to the rest of the industry. Perhaps the context is different, but again, asking for the sake of corroboration. I do appreciate the articles.
That’s a good point. I think that Warcraft’s acquisition model has changed. Previously it was a 10-day trial then buy the box, then decide if you want to subscribe. Now there’s an indefinite trial, with the box unlocking extended content and the subscription keeping that content available.
How you recruit new subscribers once the game has gone live and what influences them to take the trial and ultimately subscribe is an interesting topic in itself. Definitely one worth looking at in further detail.
Bioware has already opened a call center in Galway, Ireland to handle customer support issues. It is a 400-seat call center.
http://www.bioware.com/careers/galway/living
http://www.joystiq.com/2011/09/26/bioware-ireland-opens-as-400-person-customer-service-center/
Now, as far as those call center costs go … I’m a business analyst for a 200-seat call center, a 400-seat call center, and a 600-seat call center. Your costs are grossly undervalued; it’s an absolute money sink. Now I just did a bit of quick research that may or may not be entirely accurate and I’ve found that Galway’s cost of living is very similar to a large city in America. Ireland is in the top 15 most expensive countries to live in Europe/Asia with a median tax rate of 21%. They currently use the Euro with higher prices; but with a favorable Euro-Dollar conversion rate of 0.7691, the differences seem to mitigate themselves. In large American cities, call center employees (at least for my company) can expect to make $11/hour and work a minimum of 120 hours per month. The Galway call center is currently hiring, so let’s assume that they are at 90% capacity for max headcount (being 30-40 heads short of cap plan can decrease a call center’s performance by as much as 10%, so it is ideal to staff up to your capacity as soon as possible). So, if we do 360 agents at $11/hour for 120 hours per month, we come to an average monthly straight-time wage cost of $475,000. Now, in my experience, approximately 10% of all wages are overtime, so we’ll toss in an extra $50K of OT costs, even though Ireland doesn’t have a mandatory law saying that x1.5 pay will be activated during overtime hours (http://www.citizensinformation.ie/en/employment/employment_rights_and_conditions/hours_of_work/working_week.html).
So, you’re looking at $525,000 month in wage costs; you haven’t even gotten to insurance and other expenses, and that’s just for the employees. You’ve still got rent costs for the buildings, expenses for supplies, janitorial, support staff (supervisors, operations managers, workforce analysts), depreciation of equipment, etc etc etc. The costs just keep piling up.
Long story short, if I can compare it to my American 400-seat call center; our operating profit usually comes in around 1.3 million per month, with expenses totally 1.1 million per month. As the Galway facility is a first-party facility and it’s a customer care center that means there is no profit to be made (there’s no client to bill). So, if apples are apples (and I know they’re not, but it’s close enough to estimate), then the cost of operating the facility each month is 1.1 million dollars.
Whew, I think I just nerded myself out!
Interesting, although there’s a couple of challenges I’d throw into the mix if that’s OK.
First up, I don’t think that all 400 seats in that call centre are going to be dedicated to SWTOR. According to the VGCharts info we’re only looking at 600k copies sold in Europe, and 400 feels on the high end to be supporting that number of users. They’re probably also going to be supporting Origin, Mass Effect 3 and the new Warhammer League of Legends style thing.
Secondly, I don’t think the building’s just a call centre. Unless EA have a separate datacentre building nearby, they’ve probably split the setup to have servers on the ground floor and the call centre above. That should bring the relative rents & rates down.
You’ve got some good info there though – I’ll need to bump up my estimates a bit, but I’d be interested to know what your take is on the number of support staff you think will be needed.
From all indications, the call center does indeed appear to be devoted entirely to SWTOR; they have opening for French and German speakers as well (the two other TOR localized languages). From the articles in local papers (the best source for on-the-ground info), the entire call center seems to be aiming to support TOR, at least in the short term. Other projects may be settled in later.
http://www.irishjobs.ie/Recruiters/BioWare-Ireland-Ltd-A-Divison-5493.aspx
http://www.irishtimes.com/newspaper/finance/2011/0927/1224304799848.html
As for dividing the space into different sections; I don’t entirely know if that will work. The server farm for TOR in Europe and the customer care team may not be located in the same area for simple reason; efficiency. In the call center world, you generally want very specified buildings because it makes the data connections that much easier to maintain. For customer care, you’ll need voice, data, and internet all piping in at a pretty high bandwidth; do you want your care work to then utilize the same trunk that your servers are using? I’m not a server admin so I can’t really answer that, but it doesn’t seem to make sense, and none of the call centers in my company house anything other than call centers in their buildings.
Lastly, for the VGChartz info. Look, that’s a great tool to estimate with, but VGChartz is fundamentally flawed in that it only supplies information for retail, physical box sales. We will likely never know the true amount of copies sold, but some prominent analysts have speculated, when adding the digital copies, that 1.5 to 2.2 million copies were sold. My company is a third-party supplier to a major client; we currently have 15 sites running customer care for this client. The client also has 10 other sites with other vendors, as well as 5 sites run directly by them. In all, there are ~15,000 employees handling customer care for a company that has over 50 million customers. So if we use these as guiding principles, that means you’ll need approximately 1 customer care agent per 3600 customers. If you take the optimistic angle and say that Bioware currently has 1 million subscribers, you will need approximately 275 agents. So, I’m guessing that a 400 seat call center would be right in line with their needs for this game at this time, with room for growth.
Interesting. I’ve got to admit that my own call-centre knwoledge is limited – I end up designing some of the systems that they interact with and how important process streamlining is, but I’m not involved with their day to day operations. But in terms of sharing datacentres and call centres, the best answer I can give is “it depends”. For some scenarios it makes absolute sense, while for others it’s a more difficult judgement call.
You’re right though – if we keep digging we start getting into average call handling time, after call working and so on, and from there it just gets messy. Would you say 400 staff for global support or just EU support though? I’d probably go with 100 in EU top-end, perhaps 200-300 in US split between timezones. What do you think?
I’d say that’s a realistic split; 250 US and 150 EU. It all depends on whether Galway is the only call center operating for TOR right now or if they have another center secreted away somewhere through a outsourcing company. Also, my call center works for a telecommunications company, so the AHT goals for the site may be far larger than the 5 minute TAT our agents are required to attain. I’m fairly confident in my estimate of total costs for the site being around 13 million per year but it also depends on how many of these agents will be voice agents and how many will be chat-only agents. Chat-only agents can, theoretically, be offshored to save on expense (and judging by the pics I’ve seen of in-game GMs, I’d say they’ve chosen this option).
I cannot believe we are having a discussion about call center operation as related to TOR. I believe we have now just bored your entire reader-base.
I believe we have now just bored your entire reader-base.
Not a chance. I personally am far more intrigued as to the numbers game behind SW:TOR than I am about the game itself. With huge budget claims there’s a lot of theorycrafting as to where the money pipelines lead.
I had sent Gaz offline correspondence for research I did regarding the 200m number (released as PR) compared to other claims of 300m. Had I known others were doing the same for the ongoing budget, I’d have posted here.
I enjoyed the read and insider knowledge sharing, Targeter. Gaz, I wonder if you subscribe to the 300m number and if further evaluation on maintaining costs has altered your outlook on how this product will find the black.
A small copy about the retention rate which is dramatic these days … Like I said earlier this game is still in its launching weeks (80-100K sold copies per extra week, a big PLUS on players) and it still is not getting growth anywhere.
EU: http://www.torstatus.net/shards/eu/trends
US: http://www.torstatus.net/shards/us/trends
You can see straight away that the amount of full servers have bottomed out to nothing, the heavy servers are on a massive decline and the amount of light servers are going up through the roof.
ON TOPIC:
Small count: Production cost $ 200 M = running costs $ 24 M per year (programming new content, support centers, MARKETING & publicity at $ 2 M per month is the bare minimum).
35 dollars per sold copy on revenue (retail, distribution) and 12 dollars per subscription card (minus costs for retail, distribution)… and you would see that around 700K subscriptions is needed to recoup it all until … mid to late 2013.
With THOSE stats above, I don’t think they can hold out @ 700K subscriptions and GW2, D3, are not even out yet.
The decline in activity is not as dramatic as AoC but certainly on par with WAR in its first 4 months. WAR still had around 300K subs after 4 months (coming from 800K).
The only factor in favor of SW TOR is its SW franchise which could lead to a slight better retention in NEW sold boxes over the months. But free to play is the new market, as WOW and EVE showed decline of around 20% over the last year.
Still I expect the game in its current state of decline to hit 400-500K by summer time.
Which would result in non profits until deep into 2014.
As I’ve replied elsewhere, using server activity to predict subscriber numbers is a logical fallacy.
Other than that, I pretty much dispute your costs. Marketing at $2m a month seems incredibly high – do you have a source for that figure?
Do you have a source for the amount of revenue per subscription card, or the split between game time card subscribers and directly billed subscribers?
Do you have a source for the data for WAR, either in terms of boxed copies sold at retail or the number of subscribers at the 2 and 6 month mark?
I understand that a lot of what you’ve provided is opinion, but I’m curious if you have any data to back your figures up.
Nice commentary good sir. However look at the server list 75% are in light as of this moment, any theory that a purchaser will retain subscription for a year is off the mark. SWTOR is bombing fast, this game is good for 2 months play only, as a player, I was excited at first but now there is little point. We’ll not go into those mechanics, however elements that keep those playing that are not hardcore star wars fans are completely missing.
rip SWTOR, I really hope $200 million wasn’t the budget, it sure doesn’t look like, play like or feel like that amount.