When anything new is launched on the mainstream, and it has a companion web site, it’s usual to be prepared to get a serious “spike” of people coming in. Planning ahead for that spike is not being “optimist” but just rationally and calmly prepare for the worst. Sure, it’s a hugely expensive move to do it in SL with LL (almost US$750,000 just in island setup costs!), but what would be truly chaotic would be to get those 50,000 extra users that registered for SL in just a few hours knocking at the doors and not being able to get in…

The difference is that on websites hardly anyone bothers to count how many servers are actually serving content, mostly because it’s not something that’s easy to do for a layperson. On SL, however, you’ve got statistics!

I still wonder why people claim that these “extra people” never logged in (LL’s statistics say otherwise!), and claim an infallible method based on “green dot counting” (which everybody knows that the “green dots”, like the “X avatars in sim” stats, are cached and do not reflect “instant” data anyway). Even so — some report that 800 people were in-world on the CSI:NY sims during the period shortly following the airing of the episode… and that is bad? I’ll say WOW! That’s amazing!

As I’m tired of repeating, 2-3 people in a sim per minute, during a whole month, is the equivalent of a website having 150,000 unique visitors per month. So if the CSI:NY sims had *800* all the time, that would be close to 50 million unique viewers on a web site! I’d go *gosh!* on that if it were true!

Alas, of course, this only happened during the “spike” after the episode was aired. Obviously those sims are not going to have so many visitors, not even close to it. However, my own “conservative” feature of a million new sign-ups after the last episode was shown, would not be too bad. The episodes tend to attract an audience of about 12-13 million viewers in the US:

http://www.csifiles.com/news/011204_02.shtml

http://www.variety.com/article/VR1117972892.html?categoryid=14&cs=1

So this means that about 1% of those people who watched CSI:NY did in fact register for SL in the 48 hours following the show. That’s not bad at all! That’s about the same percentage of effectiveness of ads on Google AdSense, and a pretty good average for most types of mass-market advertising.

After the spike, naturally, it doesn’t make sense to keep so many sims open. It’s like the shop that hires a hundred temporary attendants when a truly popular gadget (like a new PlayStation or the iPhone) is launched. You don’t need that many on the next few days — just on the launch date.

Tateru’s graphs on http://dwellonit.blogspot.com/ are sadly very tiny and hard to read, but still, you can notice there that the number of total signups had a slight rise in the past few days, the number of daily signups spiked wildly to three times the average (and kept at 50% above the average rate for the last few days), and the number of active accounts has found a new plateaux, instead of dropping like usual. So, sure, it’s not *dramatic*. But the difference will be felt on the long run, since signups *accumulate*, even assuming that the retention rate (just 10% of all users remain “active users”) remains the same over time. Is that “a failure”? Thanks to the CSI:NY episode (and possibly “The Office” episode before it), SL might hit 12 million users this year (a prediction made in January 2007 based on a an optimistic growth) instead of “just” 11 million, which would be the more accurate prediction before CSI:NY was aired.

Forseti, for the sake of the argument, I’m still a believer that there might be a million *additional* signups when all the CSI:NY episodes featuring Second Life are aired, not only in the US, but world-wide — which will take a year or so. Of those million, another 100,000 will become “regular users”, according to the current churn rate.

So I can’t possibly understand how people can deem this “a failure”. What did you expect? A million signups in just a day? The number of active users doubling after a week? 100 million users by the end of 2007? Hardly! Remember that only 12 million people watched the show — barely more than SL’s population! — and only a tiny fraction of those would sign up for SL anyway. Also, why should these people behave differently in SL than the rest of the users? They’re newbies like all the rest; their experience in SL will be slightly better (they have a better orientation island; a more user-friendly viewer; better defined goals and purposes; quite a lot more Greeters/Mentors/Helpers on duty), but overall, the whole SL experience will not be overwhelmingly different. Sure, I might concede that the “clueless CSI:NY newbies” might have a lower churn rate than regular SL newbies. Maybe 20% remain active users instead of only 10%. I don’t know. It would be nice if that happened, which would point out several of the reasons why people leave SL, but that’s not the point.

The point is that the CSI:NY episode *did* have a measurable positive impact on SL, no matter how little it might seem to many who were expecting, I don’t know, a *tsunami* of new users. Three times the number of signups on a single day sounds quite good to me, though! And getting 1% of all CSI:NY viewers to sign up for SL is *fantastic*, considering that SL is not exactly the easiest application on the world to use…

It appears to be complete bollocks.

Please try a pre-amble introduction sometime as we are not on the same planet as you.

thanks

I asked if the Herald was correct that she’d be losing her job. She was angry I asked. I asked what they were paying her. Silence. These were personal questions, and I shouldn’t ask!

Sorry, but I think the public should know whether this experiment, which was going to dramatically impact our shared world, had the numbers claimed or not.

http://secondthoughts.typepad.com/second_thoughts/2007/11/sheepish.html

I am a CSI fan (though of Miami, not NYC!), and I’m an SL fan, and I thought if *I* were watching CSI, I’d be just the type to hop onto SL and play the interactive game!

ESC made sure they had enough sims not to have to turn away people at the door, and that was wise. But I think it is safe to say they rather overshot their estimates.

And wisely, they cut way back on the sims once they saw what the numbers were.

In addition, the people who did come in were not evenly spread over the sims.

I kept an eye on the sims over the evening, and noticed the same pattern all night, with cluster 2 jammed (and later I learned, cluster 1), many clusters empty, and some with only one or two on them.

After the show had aired in California, I counted the green dots in the sims and found (in brief here):

My count, from 11:30 p.m. SL time, of all the people on the CSI sims: 732.

I couldn’t find cluster number 1, so when that was added via another person’s report of it as packed, I believe that brought the total to: 802.

(I can’t find the post right now to verify, but I remember that he said about 70, for a total of 802.)

_____

The breakdown of the 11:30 p.m. snapshot, with four sims per cluster:

Clusters 2-10: 202 people

Clusters 11-21: 77

Clusters 22-31: 143

Clusters 32-41: 70

Clusters 42-41: 48

Clusters 52-61: 24

Clusters 62-71: 16

Clusters 72-81: 11

Clusters 82-91: 34

Clusters 92-101: 34

Clusters 102-105: 73

41 of these 104 4-sim clusters had 0 people on them.
_____

In addition:

Clusters with 0 people on them: 41
Clusters with 1-10 people: 35
Clusters with 11-20 people: 17
Clusters with 21-30 people: 7
Clusters with 31-40 people: 3

(No clusters with 40-70 people on them)

One cluster with 71 people (that was Cluster #2)

And one cluster with some 70-odd people on them (Cluster #1, as reported by someone else)

= 105 clusters; 802 people
_____

coco

but I’ll give you credit for making it more readable, I wrote that with no sleep. lol.

Umm, what? I think you have your concept backwards here. Unless you’re touting the fact that a random noob can log in, and have a desolate, overmarketed, overhyped wasteland all to himself.

That’s rolling out 4 sims to attract one resident, who most likely, will NOT buy a sim.

Odds are, they’ll spend all their time on a camp pad while they’re afk watching CSI.

This project was a failure. For CBS, for LL and for ESC (but I’m beginning to expect that from all three)

We needed large numbers of sims for the evening of the first episode because that was when there was the greatest occurance of a spike, and you don’t want to invite people to an experience and then say, sorry, no room for you. After that initial time period, you expect usage to smooth out.

There was some speculation that this project was going to bring in ridiculously high numbers of people (like a million new residents) — and if those speculators are disappointed, well, there’s not much I can do about it, because they certainly didn’t hear those numbers from me. I am happy with the numbers we are seeing – they came in where I expected.

We say over and over, we need to take one step at a time until virtual worlds are mass market, and we’re laying the paying stones for each level. This was an important step, and it’s only the beginning of the CSI project. This whole project isn’t about one day.

I know, I know… doesn’t make as sizzling a story.