Second Life 2012 Statistics

It is hard to find good statistics on Second Life™ that tells us what is happening. Nor have there been any exciting ‘statistical’ developments this year. So, I haven’t been writing much about them. I did want to summarize them for 2012. It took me some time to get around to it. But, here they are.

User Signup per Day

User Signup per Day

Signups per day tells us something. It is just debatable what exactly. We can’t separate re-signups, those creating Alts, from new users signing up. We can’t even get an idea of how many people may be going for an Alt. My belief is the majority users have at least 1 Alt.

But, we have never known and as far as I can see there is nothing that would have changed the rate of users creating Alts. So, the signup rates are hopefully consistently distorted.

The chart is too short to show any seasonal patterns. I marked December to April as we have to instances where signups slowed. That may be a seasonal thing. Give it a couple of more years and we might know.

2011 was a better year for signups than 2012. We cannot tell what the Lab was doing from the chart. Did they decrease or change advertising? Probably. I can draw a trend line (green) that shows a steady increase over the last couple of years. I could draw another (no shown) that runs from the beginning to December 2011. That would show a steady gain. From there I could draw a line to the end and it would show a steady decline. It depends on what one wants to show.

Regardless of what we want to show, the number of signups in 2012 is less than 2011.

Concurrency 2005 - 2013

Concurrency 2005 – 2013

This chart shows the number of people concurrently logged in. It runs from 2009 to January 2013. I chose this chart because it shows a consistent trend over a long period, which is important.

You may remember that Rod Humble said something about use-patterns for Second Life changed. I think that was the reason for the Lab discontinuing hours-in-game reports. Things were supposed to have changed when mesh was released and people spent more time in Blender/3DS/Maya. They were in creative programs creating things for Second Life, but they don’t show in the SL stats, sounds reasonable.

I know that I spend more time on SL projects outside of SL, not logged in. But, how many people actually do that? We have no way to know. But, what we do know is there is no sudden change in this graph. We got mesh in August 2011. I placed a vertical red-line about where I think August is.

Concurrency peaked in early 2010. It has been in a steady decline since then.

Those big negative spikes are typical of server crashes and network outages, which through off the mean and median vales. In the top image in the inset I have a concurrency chart corrected for those crashes.

Another look at Concurrency - 2012

Another look at Concurrency – 2012

My tracking of the peak user count cleans out the server crash numbers. I want to know how many people are logging in. Therefore I consider crash numbers anomalies. I remove them and use the next day’s numbers.

The data is a bit flaky this year, meaning my collection of it. I’ve revised the script I use to collect the info a couple of times. I made some significant changes this January. My hope is the 2013 data will be more representative of user behavior.

We can see that for most of the year peak concurrency was decreasing. The minimum number of concurrent users has remained nearly constant. While during the last part of 2012 there is an upturn showing, it is too soon for it to be considered change in the trend. It could be. I hope so.

Regions

Region Count

Region Count

The number of regions in Second Life has been steadily decreasing since 2006. While there is some stability in Adult regions and even some growth, PG and Mature regions are decreasing faster than Adult regions are growing. So, more are leaving than coming in.

Many think this is a reason to lower tier costs. I disagree. Second Life is the Cadillac of virtual worlds. It is developing and moving forward with technology… it is behind many games. But, it is old and is being updated. This year the hardware was updated to newer, faster models. The software is being updated.

If you look at the OpenSim side, many grids simply cannot keep up. In-Worldz has most of the features of Second Life. But, visiting In-Worldz is not like visiting SL. Linden Lab has the money to rapidly develop SL and it shows. Visiting OpenSim worlds often reminds of me of my first visits to Second Life.

Cutting off the tier source of income would end that development cycle and I believe make users even more unhappy.

I also do not see lower tier bringing in new users. If that were the case, In-Worldz with $75/month tier would be a boom town. Instead it is at a growth plateau of around 1,000 concurrent users…  if I remember correctly. Check Hypergrid Business for more information on In-Worldz.

I’m not slamming In-Worlds. It is pretty nice. I has greatly improved since I was last there. But, I am pointing out that it is Chevy world at a third the cost of a Cadillac AND it is not selling that well.

Forum

Forum Usage Stats

Forum Usage Stats

I have not followed the forums stats. This is the only source for this information, that I know about, for the forum. I saw a link to it once on Daniel Voyager’s blog. I am not sure how the data is collected. But, it is being kept up to date.

Obviously it was created in February 2011. I am not sure why the high numbers in early 2011. Whether that is a data collection anomaly or actually reflects user activity, I can’t say. There appears to be a decrease in use about mid 2012.

In September 2012 the JIRA Change was implemented. By the appearance of this chart that has had little if any affect on forum use.

I looked back through my 2012 Review Article: Looking Back at Second Life 2012. Pages 6 & 7 cover June and July. This is when we first started to get extensive coverage of Cloud Party. It is also in this time frame that Google and Bing discontinued free translation services. I can’t see anything else that I think would change the popularity of the forum. It obviously decreased, but I have no clue as to why.

JIRA

I have not been tracking JIRA use. I would love to have had historical data on the JIRA for the last year. But, I did not anticipate the JIRA Change. We run about 200 to 300 bug reports per 30-day period.

You can click on the Projects drop down in the dashboard to get a page that shows the stats.

Analysis

The best I can say is Second Life is coasting along. Use probably is slowing. There is no doubt that it is down from the earlier 80k peaks. And there is no doubt that we have fewer regions than we once did. Whether those that are left are better quality or not is hard to say.

I think the RL economy is having an effect on Second Life’s economy. While the media is hyping the recovery, there really is no recovery. You can look at the numbers in this card by Allen West.

US Economic Facts - Jan 2013

US Economic Facts – Jan 2013

The reference sources for the numbers are here, just below the card on the linked page.

The result is that regions not paying their way are being dropped as people tighten their budgets and look for places to cut spending. I expect to see more of this during the next 2 years. Meaning we are going to lose more regions. But, that has good and bad aspects. As spread out as SL is, in some ways it is like living in the countryside. Fewer regions may make it more like a city with denser population, which would lead to more social interactions, which would help player retention.

As any business sees its revenue stream shrink, it will start looking for things to do to improve it. I am certain the Lab is no different. Obviously one step is to diversify, which we have seen with Patterns, Blocksworld, and other spin-offs by the Lab.

I am certain that the great popularity of social networks has not been lost on the Lindens. But, geeks are not known for being highly social. So, getting programmers to understand what is needed for good social networking is an uphill trek. Whether they will try more socialization or more gamification is hard to say.

There are rumors that the Lab may work on chat this year. I don’t mean just the User Interface (CHUI), but the chat services. They need to do something with them. Chat lag is back with a vengeance. There are a number of groups I am considering dropping out of because I can never comment in them. I type, hit Enter, and get and error… cannot establish a connection thingy. That could be a serious depressant on player retention as studies show it is social transactions that keep people in MMOG’s.

I do think Second Life needs the Materials System and a better avatar. While we have what I think are the best looking avatars around many users want to improve them. It is hard to know how important that is to user satisfaction and as a player retention feature. For most games improved graphics does little for player retention. But, improving avatars is not quite the same. Avatars are far more personal and improvements there will likely have more impact.

The Materials System will greatly impact the appearance of Second Life. That will attract those people that like a high quality visual experience. So, I have some hopes for that affecting player retention among new signups.

Gamification… I don’t know. I do think getting the Experience Tools released is important as that will allow game creators in SL to do more and require less of the users. If that does help, it will be long term ‘help’ as it will take time for developers to build the games. Loki is certainly looking forward to getting the tools.

All of these things could help player retention and they will be arriving this year. So, we may see the stats change direction this year.

Competition

I like my OSGrid. But… if you haven’t been using a new viewer and playing in the test regions where HTTP, Interest List, the new Render pipeline, and other features are in use you are missing out on some the performance improvements.

After you use the new viewers for a bit, try visiting OSGrid or In-Worldz. It is like rezzing into molasses. So, as much as I like OpenSim and will continue to use it, I don’t consider those grids serious competition for SL. They may become serious competition, but I doubt that will happen this year.

Cloud Party has the potential to become serious competition, possibly this year. It is a beta project for now. But, many of the tools we would like to see in SL are being built into Cloud Party. As always the last guy to update has the best tech. The Lab has to deal with legacy stuff. Only time is going to tell if the Lab can stay ahead of Cloud Party or not.

Cloud Party also has to overcome the Blue Mars stigma. Several people got burned investing time and effort into Blue Mars. I think that is making uptake of Cloud Party slower. Still I do see it as a future competitor.

While Second Life doesn’t really have any competition right now, at least from our perspective, non-SL users have a huge array of games competing for their attention. We will have to see how Amazon and Steam affect the signup and player retention rates.

Summary

The stats are certainly not rosey. But, neither are they all that gloomy. The potential is there. If someone comes up with the right spark at the right moment SL could take off.

My hope is by the end of 2013 we will have several new key pieces in place and SL users will be able to build things that will capture people’s interest. With any luck we will find a way to improve SL and get it growing.

Thanks to:

23 thoughts on “Second Life 2012 Statistics

  1. Thank you for putting these information together. Your blog is one of those where it is not constantly whining “SL is going down” (which can turn into a self fulfilling prophecy if the whiners are not careful). Instead, you do the work, compare the data and also clean up statistical errors.

    Great Job!

    One thing about the “sim loss” also: LL tried to make regions “without setup costs”. For a week or a weekend? Can’t remember. But i would not be surprised, if this also has lead to some market saturation, with its backlag still showing.
    Do you have numbers on that?

    • I have heard that most of the regions that came online during the ‘No Setup Fee’ offer have gone offline. I don’t have any data to support or contradict those claims. It would be possible to dig up the dates of the offer and use Tyche Shepard’s grid survey to find the region names that come on during that time. Then check them against his list of existing regions.

      Sounds like a lot of work to me, so I’ll pass.

      Thanks for the kind words.

  2. re. signups: Pussycat Catnap has an interesting hypothesis http://catnapkitty.wordpress.com/2013/01/02/new-theory-on-sl-signups-that-never-seem-to-join-theyre-forum-spam-bots/ It is hard to say just what % of new signups this accounts for but it is possible it is significant.

    re. RL economy: While you and I seldom agree on politics this time we are at the same place. The RL economy in the US has not improved significantly for the “us” who would have time, money and inclination to join SL.

  3. The stats have other issues behind them, for example the number of private sims boomed during the Open Space/Homestead debacle. The decline has been steady largely since those heady days.

    Concurrency dropped when LL started clamping down on traffic manipulation and removed traffic as a ranking factor, making like for like comparisons difficult.

    Second Life is where the people are, it’s the same reason World Of Warcraft prospers but newer MMO’s struggle to grab a foothold, despite being technically superior. However if tier in Second Life was USD$75 a month for a full sim, there would be a massive explosion in the number of sims on the grid, that would certainly turnaround the decline in regions, although I doubt it would be healthy for the long term existence of the product.

    The no setup fee deal was successful in getting new sims on the grid, however that mini boom was more than swallowed by the losses, whether the sims that came online during the promotion are still here, would take a hell of a lot of working out, even with the aid of Tyche’s stats.

    • For the Lab to break even on $75 tier, they would need to see the region count triple. I doubt that would happen. In-Worldz is a pretty good indicator that it is not price that is making a difference.

      So, I can’t see anything to suggest that there would be an explosion in region leases if tier cost dropped. You and many others think that is the case probably based on basic economics; decrease the price and sales go up. But, with cheap knock-offs not selling well it will take some convincing information to make me or the Lab believers.

      The boom regions from no setup cost may be the main part of the decrease.

      The Homestead debacle was/is a mess and I suspect did hurt region sales and has resulted in a loss of regions.

      • There aren’t enough people in Inworldz, that’s the same for games that come along that aren’t World Of Warcraft, cost alone is not the only factor, people want to do things in Second Life and socialise.

        The reason there would be an explosion in private sims in Second Life if tier was USD$75 is because existing owners would increase their land mass. However, as I indicated, I don’t think this would be good for the long term future of Second Life as many existing owners would triple their land holdings, but would be paying less each month than they currently do, in that example USD$225 instead of USD$295.

        I’m in the camp that says Linden Lab need to find alternative revenue streams before cutting tier, I’m certainly not in the cut tier and do nothing else camp.

          • Well first of all current landowners trebling their holdings would still be a huge revenue hit because Linden Lab would have additional costs but the evidence comes from what happened with Open Spaces.

            Originally they had to be purchased in packs of four, when LL changed that to being ordered as singles, albeit with a slight tier increase per unit, the product boomed. LL also increased the number of prims they supported so a tier per month rise of less than USD$2 per unit for existing owners was easier to swallow.

            Instead of costing USD$295 for 4, the price was USD$75 in tier for 1 and they sold like hotcakes. There was the other change too of allowing them to be placed in any available space on the grid, so the product changed, but the basic principle that made them popular was not having to buy four at a time and pay tier for four at a time.

            • Selling like hotcakes is not very definitive… and I doubt it is going to change Linden management’s minds.

          • Tateru Nino made a great point about why LL would never cut tier prices, specially never so dramatically. Putting it simply, in the short term, it would mean financial collapse.

            Let’s imagine they drop all prices from one month to the other to $75. That means that effectively LL would get 1/4th of the revenue on that month. Sure, they can handle a month of that — they have cash in reserve. But they would expect every single landowner in SL to at least buy 3 times more land as they had the month before. How long would that take? If it takes more than a few months, LL would be ruined.

            On the other hand, the same principle would hold for land barons. They would have to reach out and grab three times more tenants — in one month. Finding tenants and having them “stick” (using the word popularized by Rod Humble) is extremely hard. There might simply not be so many residents around. So land barons would be reluctant to treble their land; they would just have to figure out if they can survive on just 1/4th of the income (but also 1/4th of the cost).

            On the other hand, running an infrastructure with 4x the number of servers doesn’t cost 4x as much — there are savings all the way — but it might cost, say, twice as much as it does today. So even if LL recovered the overall income from the tier cuts, they would have twice the running costs, and a much shorter margin. Sure, we all know that the biggest cost of running Linden Lab is not infrastructure, but staff. And these days, the same tiny team can easily manage 4x the number of servers — it’s all automated. They would, however, need to hire 4x the number of Concierges. So, at the end, even on a “best case scenario” — growing from a 30,000-region-grid to a 120,000-region-grid, LL would make less money, not more.

            Or, alternatively, they could reduce their staff. This would mean less development, less new features, less bugs fixed, more shaky infrastructure. In fact, it would make SL look much more like InWorldz… again, “back to 2005”. And that would only give OpenSim grids the much-needed edge they want to compete.

            So, no, I don’t see how on the long term this could work.

            The only alternative has to come from “alternate income streams”, then LL can think on tier cuts. It might be the new games. It might be the ads. It might be a 10% “tax” on all in-world merchant transactions. It might be a different class of servers (I’ve suggested that they copied Kitely’s business model, but it would need to be very carefully evaluated, as they might kill their market for “always-on” sims very easily). It might be introducing a new type of Premium account that people are willing to join (if every active resident — about a million — paid LL $10/month, then they could give regions away for free: the subscription service would neatly cover the costs).

            But “selling like hotcakes” is simply not good enough. Not at this stage.

            On the other hand, reducing tier costs by 10%, 20%, or 30% will simply have absolutely no effect, specially at the lower end. If I rent a parcel for L$1000, renting it for L$800 or even L$700 doesn’t make any difference — I won’t get a second parcel of the same size because of that. I think that’s the main reason why LL is unwilling to charge less for tier: ultimately, if the lower cost is not truly significant, the results will not have any effect in landmass growth — while cutting deeply into LL’s income.

            • Thank you for taking the time to write your comment. I agree with the points you have made.

              It would be nice to have better stats on what is happening with SL signups and paid accounts. Not long ago I wrote about the massive financial losses players suffered from a player-war in EVE Online. There have been lots of discussions about whether subscription model games can make it or are a thing of the past. Between EVE and WoW it is obvious there are a significant number of people willing to pay to play.

              People think lots of things will sell like hot cakes. But, that is their opinion and I have yet to see anyone support their idea with sales-data-facts.

  4. Hello,
    I receive your blogs and enjoy them, thank you.

    On the stat of SL – I have been in SL for 5 years and never had or will have an alt, I have polled some long standing friends, all tell me they have never had an alt.

    Sadly some ppl use alt’s for devious reasons from my observations.

    • Thanks for saying you enjoy.

      I and several of the people I know use Alt’s for their businesses in SL and a number of other legitimate uses. Often inventory management is an important reason for using an Alt.

      You are completely right that some do abuse Alt’s and use them for devious reasons.

  5. A great summary – and I agree with most of your conclusions. I am hoping that Valve might offer a surge (and Amazon is definitely a plus), and I am expecting Second Life to be one of several free games offered on the new Steam games machine.

    I do think LL needs to continue working on the ‘new user’ experience, and give users who want it an answer to the question “what is the point?” maybe by some gamification that improves on Linden Realms.

    • Thanks for mentioning you agree.

      On the new user experience, I agree with you too. We have heard very little on how that is going or any updates on what people are doing. Answering the ‘What the point?’ question is a bear that runs of many a user.

  6. I don’t know who made the graph about signups, but it is obviously biased: the green average line should be horizontal and around the 12000 figure, because really, this is what maths would give you and what is the actual tendency over the past two years. A flat line would also match exactly what we see with other graphs: a stable or even slow decreasing activity in SL.

    I also want to stress out how irrelevant is the signup figure anyway: the fact there are many signups doesn’t tell anything about the users retention: LL “eased up” the signup process by removing the need to choose a family name for new avatars, but it didn’t do anything to increase the new residents retention.

    What LL forgot, is that if someone is not ready to invest enough time to fill up a sign-up form with two fields for their future avatar name, then they will never invest enough time to follow the steep learning curve that using SL entails !

    In fact, removing the family name for avatars only pissed off people ready to follow the learning curve, because their avatar look like second-zone ones in SL when compared to old timers !!!

    Finally, my own conclusion (which I came to since 2008 already: another “I told you” for me to say to LL, I guess…) is that SL is on a slow downward slope and will die in the middle to long term if LL doesn’t take a serious, very needed measure: dramatically lower the (currently completely* insane*) (is)land tier prices !

    • I laid the green line on. What that line should look like is totally up to what one wants to show and which beginning and ending points one uses.

      I tend to agree with you signup and retention being mostly unrelated. But, if there are no signups it doesn’t matter if they fix player retention.

      I completely agree with your thinking on the last name. Making the signup easier is a basic tenet of web design and sales. It does not consider filtering the audience to those likely to stay. But, we really don’t know which people will stay or how to filter them. So, any filters could be counterproductive. Without that information I can’t see how to implement our thinking in an intelligent way.

      I disagree with you on the tier cost, but I’ve written on that elsewhere.

  7. Pingback: Sliding, but not yet dead | Living in the Modem World

  8. Pingback: Second Life 2012 Statistics | Amazing WWWorld

  9. Pingback: Second Life 2012 Statistics | Persephoneia Ferina

  10. Pingback: Second Life, Second Attempt | Bex Ferriday's Edutechy Wonderland

  11. I really enjoyed the blog. Not everyone is privy to LL planning meetings and hearing thoughtful insight is welcome. I agree that dropping tier would hurt over all as it reduced a revenue source but I’m curious if land sales in the past have reaped any long term benefit, such as the weekend when sim start-up costs were suspended so you could acquire land just for tier or if it just benefited land barons. I know educators but perhaps a graduated tiers that benefits more groups like students, especially those interested in related information research or a random thoughts the disabled or veterans. Countless on SL are disabled, it’s partially why their time is spent online increased encouragement to those on SSDI or veterans who also are commonly in same boat might encourage so many new demographics to investigate SL. Honestly, I think SL needs staff skilled in sociology or anthropology to consult on LL sponsored events. The possibilities for types of events are so numerous but even the listing process could be better organized. I started about 5 yrs ago and saw the end of Burning Man which had amazingly creative displays. The only consistent event is Relay; I’m not why of the 100s of non profits this ones is so widely endorsed in SL. Perhaps similar classes similar to learning to build classes are needed to teach how to build community since this is what I feel creates longevity. If people struggle with this in RL, in SL it can seem even more daunting. Tks again all the info!

    • Thinking SL should do more THINGS/services really doesn’t fit with the popular idea they are incompetent… I don’t believe that, but they are primarily programmers and geeky. In general those types of people have a hard time working with those in the fuzzy sciences and certainly should not supervise them.

      By the Lab not providing a wide range of services it allows the users to provide them. Those that think they can do something well or provide a better service take on providing that product or service. Its the free market at its best.

Leave a Reply to Henri Beauchamp Cancel reply

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