Jump to content

Cost of localising a VN


phantomJS

Recommended Posts

Can anyone in-the-know share with us the cost of localising a VN (and VN only)? Doesn't have to be exact, but a accurate rough estimate will do as well (coz I think getting actual real data may not be easy) Tried to googled, but only came up with links such as this, which is not very useful....

I'm sure a lot of us are curious about it. Thanks in advance!

Btw one of my main reason to seek this info is to vivify Moenovel more (if it's indeed a fact that they don't spend much money of Tl-ing a VN, comparatively speaking) :P

Link to comment
Share on other sites

I'd love to comment on this... but my knowledge isn't extensive. 

The big one is licensing costs, though.  Originally, that was the biggest cost for a localization (this doesn't apply to MG or localizations by Western branches of Japanese companies).  One of the early rumors I heard, when they were trying to license Clannad, was that Key wanted one million dollars for the license.  However, I imagine that is at the higher range for this kind of thing, lol.

Edit: To clarify, I honestly don't know how much the licensing fee thing applies under modern conditions.  With the advent of Kickstarter and closer interactions between VN companies and localization companies, there is every possibility that this kind of cost is no longer valid or nonexistent.  I do know that some companies pay translators by the symbol and others by the line (symbol usually is a better deal for the translator)... and costs for the average game (just the translation) can be in the low to mid five digit range depending on length.  Depending on how the company in question handled the original game production, there is every possibility that royalties have to be paid out to VAs or the original team... *shrugs*

This kind of crap is why I hesitate to make a guess on the total costs.  When I first started getting into VNs, most Japanese companies didn't want their stuff coming over here, because they didn't want to have to deal with Western mores and companies (there were a couple of issues with eroge of the Illusion type at the time).

Link to comment
Share on other sites

Maybe kickstarter goals can be some indicator on how much companies feel they have to earn to cover the license and translation costs? If I remember well, Sekai Project's campaings for bigger titles have goals between 100 and 150k dollars, it was also 150k for MiKandi's Libra localization and I think even more for MuvLuv and Dies Irae. 

Frontwing, which obviously don't have to buy its own licences, set up a 35k dollar goal for Momoiro Closet, a medium-size game. If this is indeed any kind of clue, VN localization seems fairly expensive relatively to the size of the western VN market. 

Edited by Plk_Lesiak
Link to comment
Share on other sites

Additional info for those not in the know:

Valve takes a 30% cut from all games sold on steam (afaik. Kindly correct me if I'm wrong).

Just thought I'll add this on as my purpose of starting this thread is to let those who doesn't know yet have some idea of the business side of our beloved VNs

Link to comment
Share on other sites

44 minutes ago, Plk_Lesiak said:

Frontwing, which obviously don't have to buy its own licences, set up a 35k dollar goal for Momoiro Closet, a medium-size game. If this is indeed any kind of clue, VN localization seems fairly expensive relatively to the size of the western VN market. 

The Momorio Closet one is a kickstarter for physicals not for localizing the VN. They state in the kickstarter that they would localize it regardless, and that the aim of the kickstarter is the physicals. However if you look at Solpress they also have kickstarters with roughly that goal. I would assume that the licenses they got weren't too expensive considering that the ones they are working with aren't that large companies and the titles weren't well known in the west. However there is still a problem with using kickstarters as an indicator, firstly we don't know if the cost intend to cover all of the localization or if they cover some of it with their own money and assume they will get the rest needed through later sales. Secondly kickstarters also often include physicals, this means that some of the money will go to cover the production of these physicals, this will of course be taken into account by the ones making the kickstarters that make the goals.

Edited by bakauchuujin
Link to comment
Share on other sites

You can  get some information about publishing figures.  Just not anything about visual novels directly.  All of the information about licensing fees and licensing terms is very much secret.  But there are times when companies get forced to reveal these secrets: when they have to go to court.  If you want to sue over something, you have to file the contracts, including real dollar figures, with the court.  And court documents are public.  Some lawsuits in the video game and anime business have resulted in some real interesting info surfacing over the years.

The complete contract between Bungie and Activision for publishing Destiny (which originally contemplated that Destiny 3 would already be out and Destiny 4 would come out next year) can be found here. A writeup of the significant parts (a contract is rather dry and boring) is here.

Now, Destiny is a far cry from your average visual novel.  But it's the only time a full contract has been released that I could find.  A bunch of stuff was released in the lawsuits over Curt Schilling and 38 studios: you can find more here and here (note that the actual document trove is no longer online).

But even that's a bit far afield.  There is some very interesting license data from the implosion of ADV Films, though.  When ADV was falling apart, Funimation sued them over a debt they said they bought from a Japanese company.  That led to some dollar figures being released. We know how much ADV paid for certain shows.  The list is on ANN and while anime is more popular than VNs, of the three I found, it's the closest match. I'm not going to reproduce the list here (just check ANN).  But there are a few things I can say about it:

1) It's not clear what these numbers are.  Are these the up-front minimum guarantee prices?  Or is this the total amount paid, including the up-front minimum guarantee and all the royalties that were owed?  Did they even owe any royalties?  The Crunchyroll coverage seems to imply that ADV paid a flat $25K per episode for 009-1.

2) Does this include the costs of translating the show and producing the DVDs?  Almost certainly not.  $21,335 for a TV show with 39 episodes (UFO Ultramaiden Valkyrie) is not enough money.

3) The numbers vary widely.  The lowest price for a TV show is $21,335 for UFO Ultramaiden Valkyrie.  That's barely more than the $20K for the Air movie.  The highest price is pushing a million: $960K for Kurau Phantom Memory.  Four of the listed shows cost ADV more than three quarters of a million dollars, and one more is just under.  On the flip side, six others cost less than sixty grand.  The most expensive show cost ADV nearly forty five times the least expensive.

4) Licensing costs are probably paid by every company in the business.  Even MoeNovel (a Pulltop subsidiary) probably on paper pays Pulltop for the rights.  This is something I learned when I got a job.  I worked for the phone company for awhile, and I found out my department pays another department for the phone lines.  This seemed really weird to me; I was like 'why bother?'  My coworkers explained that because of the way budgets work, if that division provided the phones for free, their numbers would look terrible. The cost savings they generated would make everyone else in the company look good, but their bottom line would look terrible.  This would lead to them getting punished (for being 'unprofitable') for being successful.

Likewise, Moenovel is owned by Will.  Money that they generate goes into Will's pockets.  But they made that money off of Pulltop properties.  So the deals need to be structured in such a way that the Pulltop entity reaps the rewards of Moenovel's success.  It is probably safe to assume Moenovel pays something to Pulltop.

The only company that might not have to do this 'one part of the company pays another' thing is Frontwing, because they might not have set up a different entity to run the overseas business.  It's hard to say.

5) The non-licensing related costs are going to be relatively fixed and predictable.  They will be a product of the size (in bytes) of the script, with added complications for things like actual gameplay elements.  Scripting costs (engine work) are less predictable, but compared to a 45X variance factor in licensing costs?  That's not too significant.

6) Licensing costs vary widely.  Sol Press seems to be licensing on the cheap (i.e. not going after famous or hugely-successful Japanese projects).  For all but the cheapest licenses, however? Licensing fees likely dwarf all other costs associated with the release.  Witch's Love Diary's script is about 1.5 megabytes, or about 750K characters.  Now I know MG pays by the character (everyone does) but not how much: if they pay 1.5 cents per, that's about $11K.  If they pay 5 cents per, that would be about $37K.  [Edit: Apparently some companies pay by the character, some by the line.  That probably won't change this math too much.]

Both of these guesstimates are on the low end of the ADV price chart.  Visual Novel licensing fees are going to be quite a bit lower (they have to be, or there's no business to be done, sales just aren't high enough).  But still.

7) For most titles, the licensing fee is probably the biggest cost.  Since this is the part that's least predictable, it's extremely difficult to say what a given title might have cost.

Edited by Nandemonai
Link to comment
Share on other sites

  • 3 weeks later...

MMM: Murder Most Misfortunate

Saw a creator of that VN post on reddit on his team's experiences making the VN and after it went on sales. Thought I'll post it here as my purpose of creating this create is to give people ideas on the business side of making Visual Novels:

Spoiler

Well, it's been a year since I released my first game, MMM, so it's time I did a proper post-mortem of it. I think there are definitely some useful insights to be gleaned from it.

MMM: Murder Most Misfortune was released May 1st 2017 for Windows, Mac & Linux on both Steam and Itch.io. It is available in English, Spanish and Italian. MMM is a short murder mystery visual novel game, inspired by British murder mysteries by Agatha Christie (such as Poirot) and the Ace Attorney series. In it you play as Miss Fortune, a widow nine-times over. Miss Fortune is invited to a secluded mansion for a dinner party where a murder occurs, and she is framed. The player must then talk to the other guests, and explore the mansion to figure out what happened before the police arrive and come to the wrong conclusions.

The game had multiple endings, where you could frame any of the characters with murder if you had enough evidence. It took between 1.5-3 hrs of play-time depending on how many endings you do. Each line of dialogue is fully-voiced, and you are able scroll through rooms in a point-and-click style to find items of interest. We also had an optional timed mode, where the player had a limited amount of real-time to explore the mansion before being forced to make an accusation before the police arrived. MMM was made mostly between four people: I directed, programmed, my brother wrote the excellent script, and a friend made all of the art, and another friend did most of the audio engineering. It took ~2 years of part-time work to finish it.

Critical Reception: Although not having too many Steam reviews, the vast majority are positive, with 25/26 of them being positive. Many reviews mention the quality of the voice acting, and the writing itself. Unfortunately we did not get any large media outlets of note to cover or review the game (only a few extremely small sites). We did distribute keys through key-mailer, and had a few small youtubers play through the game. Here's a picture of the steam achievements 6 months after launch, which can be used as an estimate to how many people actually played the game, and how far they got. 57% of people had finished the first chapter (~20 minutes of gameplay), and ~37% had gotten at least one of the endings. When sales slowed to a crawl, we recently put the game on a bundle. This understandably drastically lowered the achievement rates. Overall, the game had a median play-time of 2.5 hours.

Sales: Although I can't be too specific about Steam sales numbers, overall they have been low (~1k sales). Sales followed the usual pattern of having the most purchases in the launch week, having almost no 'organic' purchases when no sales are on, and substantial spikes in purchases during sales. As can be seen from this regional sales chart, the majority came from the U.S., with Germany, U.K., Canada, China and South Korea following belatedly. Japan had fairly low sales. It is difficult to surmise how well it would have done if there had been a Japanese localization (which would've been very expensive), but I have heard that western visual novels are not so popular there. Surprisingly, only 38% of sales came from primarily English-speaking countries (U.S., U.K., Canada). A quarter of our sales came from 'other' countries. This is significant, and it is possible the Spanish and Italian localisations helped with sales from these 'other' countries.

Why the low sales? Although we had positive user reviews, overall sales were disappointingly low. We now enter the realm of speculation and list some possible reasons why:

  • Poor marketing. None of the team had experience marketing, nor do any them of like social media. The only marketing done was some forum posts and some Facebook ads. A larger emphasis needs to be placed on marketing as game market becomes more and more crowded each year.
  • Target audience mismatch. Do people who like well-thought out murder mysteries like anime visual novels? The anime aesthetic might have been a turn-off for many, with many outsiders to the VN genre thinking it might involve dating or romance. On the other hand, those looking for romance VNs won't find the game appealing. Of course there are some popular mystery-esque VNs (999, danganronpa, ace attorney sort of, etc.), but they often have more in-depth gameplay. It is a difficult balance to strike.
  • There are so many games. Steam has become bloated, with ~15 new games being released each day (at time of our release). It’s increasingly hard to get noticed as you vie for attention not only with other indie games, but AAA games as well. We released close to the end of Steam Greenlight's lifespan, but it's hard to say if that had any effect on sales.

Itch.io: Sales stats. The game is available on Itch.io as pay what you want (which we think is the future of digital purchasing). We’ve had close to 2.5 times the downloads compared to Steam, but only 16 purchases. We don’t believe this affected the sales on Steam, as the audience on Itch.io are quite distinct than the Steam audience, but we can’t confirm this. Our goal is to make people happy by playing our games, so we’re not disappointed when people who would otherwise NOT buy our game have a chance to try our game.

Takeaways & other tidbits:

  • It's easy to get wishlist adds, but even with sales it's difficult to convert those wishlist adds to purchases.
  • North America and Western Europe accounted for 64% of sales. High, but not as high as we anticipated.
  • Although difficult to ascertain exact numbers of people who bought the game it was in Spanish or Italian, these translations were probably not worth the cost and effort of producing them. Sales from Italy is only 20, while Spanish sales are more difficult to ascertain (there are many Spanish speaking countries). It could be argued that the Spanish translation is worthwhile, while the Italian one isn’t. Why did we choose Spanish and Italian? Well, the Spanish one was due to a friend of a friend being a native Spanish speaker. The Italian one was offered to us online for a very low price.
  • Everybody plays on Windows. Linux activation's account for 7% of sales, and mac for 2.5% of sales. We made the game using Unity, so luckily it wasn't much effort to export Linux and Mac versions. I would have been very disappointed if it had been much work making mac versions.
  • Itch.io doesn't make money. It has some great ideas (pay what you want, great community tools), but barely pays for a few meals. Itch users appear to expect mostly free games.

Lessons learned:

  • More marketing. Marketing needs to be done throughout development, not just at the end. Although I did the occasional screenshot post or forum post in visual novel dev forums, the overall marketing effort was extremely lacking. This is difficult, as I personally dislike social media. Finding someone who can do this, or finding a more entertaining way to do marketing will be important in future endeavours.
  • Art is all-important. Although a game is made of many parts, the only way to get people to try your game is through incredible visuals and screenshots. We felt our art was decent, but it perhaps could have been improved. Several reviews mentioned not liking the art style, which could be due to the different art style from anime-looking visual novels. It was also sometimes difficult to pick out objects from the background when exploring rooms.
  • Voice-acting is expensive. We had promised our voice actors a percentage of the profits (up to a limit), but unfortunately with our initial sales these percentages amounted to very little. I felt bad about the work they put in, so I decided to pay them the maximum amount that was promised out-of-pocket. This maximum amount was still below what is considered industry standards voice-acting fees. Only after a year have we recouped that expense. Although the voice-acting was of excellent quality and definitely improved the game, I would be very hesitant in including it in the future. Profit margins are already slim, and voice-acting can be incredibly expensive. It would not have been possible to make the game any longer and still have each line be fully-voiced.
  • Everything takes longer than you think. As mentioned earlier, it took us ~2 years of part-time work to make a 3 hours game. We thought that making a short high-quality story-based game would be fairly easy. Turns out, nope. The voice-acting, and different endings took significant amounts of time. I now have new respect for TellTale games, and now know how much effort it takes when making branching story-lines. Making games with branching stories is a big risk, since it takes a lot of effort, and you're making content that many users won't see.

Only after a year have we finally covered all outside expenses (not hours spent writing the script or coding), making the game a financial failure. Overall though, we're proud of what we made, and had a good time making it. Despite the financial aspect, we've learned much from making this game, and are considering a sequel (involving Egypt, a mummy, and of course, murder).

 

Link to comment
Share on other sites

37 minutes ago, ittaku said:

Indeed they are but I prefer to see -ise which makes it clear it's not American (unless it's a typo.)

Sorry, I was reading your post with my eyes closed and thought that you were critisising the -ise spelling. I am totally on board with you!

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...