New World Notes

VRChat Hits New All-Time Peak User Concurrency Record: Over 55,000!

Data via VRChat Status

Last weekend VRChat hit what looks to be another all-time peak concurrency record for the platform, crossing over 55,000 users online at the same time (see above data from VRChat Status, a user-run concurrency tracker).

Much of that record seems attributable to a VRChat-based furry conference promoted by VRChat the company. Speaking of which, this furry influx pushed VRChat's weekend concurrency over that of Second Life, which peaked last weekend at 53,326, according to SL Grid Survey. Second Life was once the key immersive home of furries online, but to to judge by these stats, much of that fur has flown.

Also notable:

Second Life Wesbite Traffic Down Nearly 50% Since 2018, According to SimilarWeb, Likely Reflecting Strong Decline in New User Sign-Ups

Surprising stats from SimilarWeb, especially given recent Linden Lab reports that Second Life was enjoying a quarantine upsurge:

• This February, SecondLife.com earned 9.4 million visits, and in the five months prior to that, peaked at 12.7 million visits in December 2020, and averaged visits in the 11 million+ range in other months. 
• In December 2018, however, monthly visits to the site were nearly 21 million, and earned similar traffic numbers in the previous five months.
• As a more distant comparison point: In 2014, SecondLife.com was averaging monthly visits in the 16 million range.

Unlike other social virtual worlds, most of which have a mobile app and/or run on consoles and/or Steam, Second Life is still heavily dependent on its website for shopping, account management, community activity, and -- perhaps most notably -- new user sign-ups. So web traffic on the official site provides a pretty strong indicator of overall user activity, including new users.

This doesn't, however, necessarily mean in-world user activity is also down:

Second Life Usage Now Higher Than Its 2007 Hype Level Period?

Interesting comment from Amanda Dallin, reflecting on SL's recently reported "pandemic boost": 

If the definition of a monthly active user has remained the same then the comparison to the peak in 2007 is valid. Actually there were likely more than 350,000 new registered users a month in 2007 when SL hit the peak of 1.1 million MAU so the number of returning users could be about the same or higher than in 2007 using your math and logic.

This sounds right. 2007 saw a lot more churn due to massively promoted events like a cross-over experience for the television show CSI, mostly from people who were unable to install the program or get past the orientation. However, the definition of a monthly active user has definitely changed since 2007 -- back then, Linden Lab would mainly refer to "Total Residents", i.e. anyone who signed up for an account, whether or not they even installed the program and went in-world.

The definition that hasn't changed is the amount of concurrent users, and that is indisputably much larger now than it was in 2007:

Second Life Sees Significant User Concurrency Growth of Old & Returning Players, Post-Pandemic Declaration

Data courtesy Tyche Shepherd

While newer virtual world usage so far seems to be flat in the wake of recent coronavirus news, legacy world Second Life has indeed seen some growth in usage. As renowned Second Life grid analyst Tyche Shepherd puts it to me today:

"This weekend there seems to have been more activity above expectations -- Median Daily Concurrency was up by 10% on Saturday 14th (45,000) and Sunday 15th (44,000) over expectations. There was also uplifts in Maximum and Average Daily Concurrency."

Notably, the World Health Organization designated the Coronavirus a global pandemic on March 11, while the White House declared a national emergency on March 13th. 

"However," as Tyche cautions, "this effect has only been in the last 4 days of data (13th onwards). We'll see what happens over the weeks to come."

In terms of actual users, ten percent growth in concurrency translates to about 5,000 people; not a lot, but not nothing. And to be sure, correlation does not prove not causality. (Except in the sense that writing about a correlation causes Internet commenters to angrily write, "Correlation does not prove causation!") But common sense suggests at least some of this growth is the result of more people staying indoors longer, looking for new (and classic) social outlets to play in.

Linden Lab CEO Ebbe Altberg did just confirm this growth, and also associated it to the coronavirus:

Surprising SimilarWeb Stats: 1 in 4 Visits to SecondLife.com Are Mobile; More Visits from YouTube, Than Facebook

SimilarWeb's report on traffic and usage of Secondlife.com includes some surprising stats on top of the visits by country I posted last week. While most of the traffic (66%) is direct and Second Life is a PC/Mac-only client, a large chunk of the visitors are on mobile -- take a look at the deep dive data at right, generously shared with me by SimilarWeb Marketing Solutions: Roughly 1 in 4 visits are via mobile, running between low 20s to low 30s by percentage and country -- with the largest audience by far, from the United States, visiting via mobile 1 out of 3 times.

This is a major reason that Second Life usage has been stagnant: Since the account creation and installation process is so desktop-focused, some 25% of potential new users are likely to see their efforts confused or frustrated right off the bat. It also contributes to the attrition of established users unable to easily access their account via iPhone and Android -- the lack of an app just becomes another reason to switch attention to other, easier to access interactive experiences. (It's all the more reason it's stunning that Linden Lab is only now creating an iOS "companion app" for Second Life -- roughly 12 years after the iPhone launched.)

For SL merchants and other content creators, this data stresses the need to have a mobile strategy, at least in terms of having a web page optimized for iOS/Android users.

Another surprising SimilarWeb datapoint is traffic via social media. While Facebook tends to dominate traffic to other media sites, with Twitter a close second, SecondLife.com social traffic a quite different:

SimilarWeb's Top 25 Visits to SecondLife.com By Country in 2018 - Guess Which One is #1 Right Now

SimilarWeb has a new report on traffic and usage of Secondlife.com, so I reached out to the analytics company to know more. Overall, user traffic remains stable (21 million visits in December 2018, somewhat more than July 2018, with 19 million), but the biggest surprise is where the visits are coming from: The United States is not the largest visitor in recent months.

As it turns, an analyst with SimilarWeb Marketing Solutions tells me, the highest traffic in November and December is from Brazil. This is based on comparing traffic from the first half of 2018 (January - June) and the second half of the year (July- December): "The direct traffic between the two time periods increased by 21%, along with significant increases in paid search, social, and referrals. All of these combined, likely increased the brand awareness, driving users to directly seek out the site. Display ads actually saw a 49% decrease between the two time periods." In other words, Brazilian aren't primarily visiting SecondLife.com after clicking on an ad, but going there directly. So it's likely we've seen an organic growth in Brazil users in the last several months.

Now here's SimilarWeb's Top 25 visits to SecondLife.com by country for the whole of 2018, which also contains some surprises: 

Second Life Website Data from SimilarWeb Shows Surprising Results - Much Younger, Male, Mobile Than Typically Assumed

Click to expand -- data courtesy SimilarWeb

We know that Second Life's active userbase has been about 500-600,000 for over 10 years, but visits to the official website SecondLife.com tell a different, more interesting story. Courtesy top web/mobile analytics company SimilarWeb, take a look at these charts. Some key takeaways:

• Visits to SecondLife.com from the United peaked in July 2016 at about 8 million/month
• Monthly US visits to SecondLife.com have been trending down since 2016, from around 6-5 million/mo since 2017 to now
• Current US unique visitors: 770,000, from 6.9 million total visits
• 1 in 3 of US visits to SecondLife.com are via mobile

First question that jumps out: Why the spike in July 2016? The only major US/SL news item I can find from the archives is an SL-based fundraiser for victims of the Pulse nightclub massacre in Orland, Florida.

Followed by: Why has Linden Lab refused to create (or just go ahead and buy) a smartphone version of Second Life when 1 in 3 of its own website visitors are on mobile?

Now let's look at Similarweb's demographic data from the USA, because that's also surprising:

Is Second Life Not Growing Because Its Most Dedicated Users Keep Dying Off? (Comments of the Week)

Pictured at right: SL demographics as reported by Linden Lab in 2008

Why does Second Life keep getting hundreds of thousands of new user sign-ups, while it's monthly userbase refuses to grow? Longtime SLer David Cartier makes a point I hadn't quite considered:

I've been seeing a fairly large number of new accounts, lately, where the driver clearly doesn't have a fricking clue what they are doing, so we are seeing at least SOME new people. I've also known quite a lot of people who've simply died over the years. That has to be affecting the numbers just as much as getting tired of receiving the friendly - and yet dreaded - "Wow you're really OLD!!!" IMs. Given the generally older user base I'm guessing that is going to trend upwards in the future.

That is true. The last time Linden Lab reported SL user demographics in 2008, 15% of the population were over 45, and 23% were between 35-44 in age. (See chart above.) Ten years later, any of those users who are still in Second Life are in their 40s, 50s, 60s, even 70s and 80s. And the reality is, a small but significant percentage are, sadly, going to pass away every year. (Including pioneers like Lumiere Noir.)

Chic Aeon makes a related point from her perspective, starting out in Second Life over a decade ago as a student at an Ivy League school:

Most Second Life Users Self-Identify as Roleplayers & Other MMO Game Types -- Very Few as Educators

Canary Beck has a very interesting survey of over 3000 Second Life users (one the biggest data samples I've ever seen), asking them to describe their main activities in SL, and whether they recommend Second Life to others. Respondents could choose three categories, and by far, the largest preferences were Roleplayers, Socializers, and Explorers. These are very familiar and core types of activities in most other MMOs, while other top preferences like Decorator, Fashionista, Creator, and Photographer are very much consonant with activities in sandbox-type games like The Sims franchise and Minecraft.

Just as notably, "Educator" was among the very least designated categories. Taken together, this should finally put to bed a long-running controversy among the SL user community and Linden Lab itself: "Is Second Life a game?" As I've argued before, most people who use Second Life use it for game-like activities or explicitly play it as a game, and this survey data seems consistent with that. 

So why do so many SLers insist Second Life isn't a game? More key, why does Linden Lab keep insisting that, even saying so on national television? Canary's survey provides a strong hint of that too:

Watch This: The Size of Second Life Even Now

Luca Grabacr made this pretty great machinima infographic illustrating the size of Second Life as it exists today:

All the data seems fairly up to date, but I haven't seen it presented as dramatically as here. Why no, I didn't now it would take 23,175 days to walk across the entirety of Second Life's land mass (if all the sims were laid end to end) -- so that would make SL's total length like 772 miles? (Assuming you walked 30 miles a day.) Checking with Luca on her sources now.* Also love Luca's mention of the first Linden-made object in Second Life: