Thirty Days in the Sun: Search traffic for FFXIV:ARR

Thirty Days in the Sun

Search Traffic for Modern MMO’s

Note: This article is a companion piece to my MMORPG.com column – A Realm Rising?

In my MMORPG.com article, I observed that FFXIV may be experiencing a bit of a rebirth.  In a Google Trends analysis I showed that, while search traffic remains low, there is a definite uptick starting at the end of February, 2013.  In this column I want to continue that analysis by depicting how there are predictable search patterns for MMO’s.

One of the observations I made in the MMORPG column is that MMO’s experience a Google Search peak right around their launch window.  As evidence of this, I present the launch window for FFXIV, The Secret World, SW:ToR, RIFT and GW2.  This figure was prepared using Google Trends examination indpendently for each title, so all of them have a 100 peak.  The correct way to interpret the changes from week to week is to think of them as a percentage of the peak week.  You can’t make comparisons between titles (other than the general shapes of each trend), since the actual peaks are being scaled to put them on a common level.

Further, I constrained the search window to the month prior to and the month following the launch.  Thus, launch week is week zero.  The weeks leading up to launch are numbered -4 to -1 repsectively.  The weeks following launch are similarly named week 1 to week 4.  Finally, I only report the highest Google Trend value for each week of observation.

There are two observations I want to make regarding the first figure.  The first observation is that MMO’s, unsurprisingly, peak in search traffic during their launch week.  With the exception of FFXIV (more on that in a momnet), MMO’s ramp up each week prior to launch, peak at launch and then rather quickly dropoff from their launch week traffic.  Indeed, fourteen days from launch, search traffic sits at about 50% of launch traffic.  This tracks fairly consistently with the game industry as a whole where short shelf-lives are the norm.

The second thing that stands out is that FFXIV had a double peak.  It peaks initially three weeks before launch (-3 on the x-axis) and then reaches its second peak at the launch window (x=0).  FFXIV had a very troubled launch.  It was plagued by set-backs and, deservedly, had bad press.  Week -3 was actually intended to be a launch window, but it was also the point of the public beta.  That public beta released a wave of bad press, which undoubtedly hindered sales.  If we assume that week -3 should have been just a normal per-launch ramp up week, this suggests that the week zero (launch) search traffic should have been two to three times higher than it turned out.  You never get a second chance to make a first impression and the state of the game at public beta did not help subsequent search traffic.

Sustaining Your Buzz

It is unsurprising that search traffic peaks in the launch week of an MMO.  Marketing teams exert immense efforts supporting the launch of a game.  This includes a flurry of interviews, advertising, and media support to fansites and game publications.  Much of that effort tapers off after the first few weeks and indeed, as illustrated in the first figure, search traffic drops off noticeably a few weeks after launch.

In the next figure, I depict search traffic in the six months following the launch window (period 0).  Each interval along the x-axis is a month following the game’s launch.  Thus, month 1 for the Secret World would be August, 2012 while month 1 for Final Fantasy XIV would be October, 2010.

As with the prior figure, each graph is built around traffic for that game.  So, all traffic peaks at 100, which does not mean that they all had the same traffic.  Rather, the figure should be read that subsequent months were x% of the launch month for that game alone.  The purpose of commonly scaling each title is to observe how the patterns are similar across games, even though the volumes differed substantially (more on this later).

By the third month after launch, most titles have hit 20% of the launch window traffic.  The 20% trough seems to be one with little chance of escape (other titles follow similar paths).  There are two exceptions to this trend.  The first exception is RIFT, which has actually done an admirable job retaining search traffic.  You can also see this in the MMORPG article, RIFT’s traffic to this day is above that of other launched MMO’s.

The second exception is the Secret World surge in month five.  That happens to be the shift from subscription to Buy to Play.  While that move did not produce a new launch window level of momentum, it doubled the search traffic for the Secret World.  Further, the Buy to Play move shifted the vector of its trough upwards.  Subsequent months indicate that B2P served as a somewhat sustained 10% bump in traffic, relative to their launch window.

Measuring Up

Now its worth questioning whether looking at Google Trends is worthwhile.  Google Trends simply shows search traffic and search traffic isn’t sales.  But, search traffic may well be a leading indicator of sales.  One way to check that is to compare the relative search volume of different titles.

Up until now, the graphs all tracked search traffic for a title by itself.  The correct way to interpret these graphs was to read them as percentage changes from that titles peak.  There was no comparison between titles, other than to note that the activity patterns were similar.

The next figure changes that.  Here, we will look at the peak search level for each title with a twist.  Now, each titles peak traffic will be defined as a percentage share of the largest title’s search traffic.

Guild Wars 2 is, unsurprisingly, the big fish.  Remember, each title hits its peak during its launch window. So, the correct way to interpret this figure is that RIFT’s launch window produced about 35% of the search traffic that Guild Wars 2’s did.  Or, similarly, Guild Wars 2 had about three times the search traffic of RIFT or Star Wars.  While I don’t have the sales figures for each available, it seems to reason that GW2 outsold RIFT and SWTOR by several multiples.   Notice that TSW only had about 10% of the traffic of GW2.  We know TSW sold about 200,000 units in its launch window which would put GW2 in the 2mm range, which is actually not too far off the mark.

Is search traffic a perfect proxy for sales?  Of course not.  Does the general traffic level lend itself to reasonable speculation?  It seems so.  Oh, one comment about GW2.  Their search traffic is about 50% of World of Warcraft’s launch traffic!  We know that WoW sold and sustained five million accounts in the west (and about the same again in their eastern model).


Which brings us to FFXIV.  Tracing sales figures for a game is never easy, particularly for one which fared this poorly.  At least one source I found indicates that FFXIV has sold over 500,000 titles since its launch.  This makes its search traffic seem quite off relative to its sales figures.  One would have expected FFXIV’s launch window to have been about double TSW with those sales, yet it’s only 50% of TSW’s sales.  What gives?

I think the answer lies on figure one.  Remember that double peak for FFXIV?  The first peak was their public beta.  Their second, lower, peak was launch.  I speculated in the opening section that the public beta was so damaging that launch week search traffic for FFXIV may have been only 30% of its “proper” level had FFXIV followed the normal ramping up for search traffic.  Figure three and the sales figures seems to reinforce that conclusion.

Branding Power

FFXIV hurt itself in its launch window.  Right now, Square Enix has pinned immense hopes onto Yoshida’s team to do something never before accomplished in the industry, resurrecting an MMO.  Rather than start over, they have sunk years into rebuilding and relaunching.  Conventional wisdom might suggest this is either impossible or not worth the effort, but Final Fantasy is a bit different, it’s got a pretty powerful brand.  In the next figure, I depict the average search traffic for the brand name for the past twelve months.  So, rather than searching for FFXIV (and its alternates), I am just looking at Final Fantasy, Rift, Guild Wars, the Secret World and Star Wars (not the Old Republic).

Final Fantasy and Guild Wars are very powerful IP’s.  They are roughly as powerful of brands as Warcraft which earned a 20-share over the same period for those scoring at home.  When you consider the number of product entries for Star Wars (movies, cartoons, toys, etc.), it’s pretty impressive that a video game commands about a third of the brand strength (as measured by search traffic) of such an iconic franchise.  Additionally, for Final Fantasy, the heyday of its franchise predates Google Trends, it’s quite possible that at one point FF was in the ballpark of Star Wars!

The Final Fantasy franchise has had a rough decade.  The 12th, 13th and 14th iteration of FF were each flawed in their own ways.  While each has its own fanbase, none of them approach the franchise benchmarks set by FFIV or FFVII.  It’s probably unfair to expect a subscription product to reach such levels, but if FFXIV can at least approach the success of FFXI, the gamble on A Realm Reborn is likely to pay off nicely.

Bookmark the permalink.

7 Responses to Thirty Days in the Sun: Search traffic for FFXIV:ARR

  1. I’m unclear, are you plotting search results for FFXIV 1.0 launch or FFXIV ARR beta? How do you discriminate between the two? The problem I see from a data perspective is they share many of the key terms. In any case I really like this article.

    More germane to viewing trends, in my opinion, would be to analyze social media interest. Judging from the trouble history of v1.0, and the track record of poor advertising my SE (in general), I think a lot of buzz about the ARR will come from the active player base. I don’t have much time to look into it how, but replicating this study using appropriate hashtags would go a long way on measuring the social engagement of the title.

    • Ryahl says:

      Thanks for the read and question.

      In this article, I’m entirely focused on the 1.0 launch.

      We aren’t at the 2.0 launch and, by my best guess, not even near the one-month window prior to launch. The MMORPG article next week will pick up on some of that, I’ll spend a bit more time there comparing the last two months (which is clearly indicative of ARR) and other upcoming MMO’s.

      While I agree that social media (I’m thinking you are referring to Facebook likes, Twitter and even things like subscribers to subreddits) is a good benchmark today, many weren’t available (or at least the data wouldn’t be available) for some of the older MMO’s. Even Google Trends doesn’t really track back past 2004, meaning it’s virtually impossible to evaluate launch windows for Everquest or Dark Age of Camelot. I suspect that a hashtag analysis is a really good benchmark, though for the most recent crop of MMO’s. You just can’t really do historical with it. Nonetheless, I agree that the two approaches should corroborate each other.

      Now, one problem with a hashtag analysis is that it’s (possibly) more manipulable than Google searches. It would be difficult to do enough Google searches to manipulate the data on a global level. However, your marketing push is going to get a lot of retweets from game sites and the like. To some extent, that can be influenced by the media push itself.

      That said, the two separate examinations (Google Trends and Hashtag) become a possibility of estimating the marketing process. Hashtags should be a reflection of the success of the publicity wave converting media arms while trends is a reflection of the influence of the media arm on consumer interest.

      • Thanks for the follow up! I didn’t even think about how old some of those titles are, so you’re 100% correct cross analysis for social media would be impossible. Thanks for the insightful read! I found this on Reddit, i’d encourage you to post your articles on Google+ as well. I’m a part of one of the larger unofficial communities there. I always enjoy your work, here and at MMORPG.

        • AelaAela says:

          We both actually have very little experience with Google+, and I wouldn’t even know where to get started (I do most of the posting of the site). However, you are welcome to post this (or other articles) on Google+, I can usually follow the links from our viewers (via google analytics) to get ideas of where folks are coming from and follow up with them there.

        • AelaAela says:

          I just saw that “Google+” was a link, so I will look into it further :)

  2. Sorry for the typos ^.^b

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Anti-Spam Quiz: