Thanks PlasmaJohn, ehye_khandee, and Lazarus Magni for giving us some idea of what your hosting costs run! With some idea of the range of hosting costs (anywhere from $30 to $500 per month at the extremes), we can rough out some idea of a revenue stream for the idea, just making a few assumptions.
I am just pulling numbers out of the air, though I suspect they are pretty generous. Let's say there are 1000 PWs with a consistent member base - that's significantly more than there are currently, but we'll assume a high number to represent Bioware's continued involvement keeping interest up. (BTW, I am assuming it's Bioware who gets into the PW hosting business, but it could be someone else.) And we'll further say that 500 of those PWs are willing to pay for hosting and they can come up with $250 per month for Bioware hosting, which is a little higher than the folks who posted here say they are paying, but it's the cost of convenience and the Bioware name means something. That's $1.5 million/year in revenue. Given the assumptions made, that's likely an upper bound, but it's a number that entices one to look further into this.
Of course, that's just revenue, not profit. This kind of hosting is a hands-on, customer service business. The whole point of a PW paying Bioware to host their module is that Bioware would take care of the constant tech support, hardware maintenance and upgrades, network issues, etc. Running a server farm for 500 PWs is going to require several trained tech people to keep it running and keep up with normal support tickets. And, this is NWN PW hosting, so they have to be able to assist people setting up databases, upgrading modules, creating backups, and generally getting things running. And there would likely have to be a crew of people to keep the operation running 24/7, since you can't very well tell someone whose PW has crashed Friday night that tech support is out until Monday. I am going to guess it takes six employees to keep someone on hand to run the server farm and deal with support tickets around the clock. These guys are going to cost 75 to 100 grand a year each in total compensation (salaries, taxes, benefits, training, etc.). I will say one senior tech support manager at $100k and 5 staff at $75k. BTW, one could argue that it's possible to get away with fewer people and/or they can be doing double duty elsewhere and so on or even that they might work for less money. All possible - but keep in mind that if these folks feel like they are working like hasted gnomes and not getting sufficiently compensated, then turnover shoots up and you quickly lose whatever money was saved in added training costs for new people and low productivity.
Then, there is a need for someone to deal with billing, new accounts, closing accounts, and so on. That's probably a whole person, since 500 PW customers can generate some paperwork and there will definitely be flux in the customer base as PWs come and go, though staffing for that doesn't have to be 24/7.
So, far we are looking at over $500k in compensation costs and we haven't bought any hardware yet nor rented a building to put it and our personnel in. And, of course, if we have 500 PWs running, maybe averaging 25 players online during peak time (the bandwidth has to handle the peaks), we will need a somewhat decent data pipe. I am thinking those 12,500 players are soaking in the neighborhood of 3 kbps each for a rough bandwidth requirement of almost 40 Mbps, but one of the PW admins here may have better numbers on that. Ultimately, that's not the biggest expense and it may be best to outsource the actual rack space itself, but that means less control over the system hardware, which may or may not be a disadvantage. And, it has to be paid for regardless of where it sits.
Of course, ongoing advertising/promotion is needed and so on, but I won't try to estimate that and we'll just hope that PW players who become interested in building their own PW will be aware that Bioware can provide the hosting for them. There's probably a standard assumption for scaling overhead in this business, but I don't know it offhand. It seems likely there's another one or two hundred thou in personnel costs I haven't looked at. Also keep in mind that none of that is providing more content or development of game or server features, which require different people with different skills...
All in all, this is still too rough to judge, but I'd say the idea may well merit looking into by someone who has a solid handle on the numbers. Keep in mind that the revenue figures are pretty optimistic and the couple of costs I looked at may be low. Unfortunately, it wouldn't take much of a erosion of the revenue figures before the scenario loses viability.
Modifié par MrZork, 28 août 2012 - 12:27 .
