True Membership Count and who's listening?

[killivolt]

So, many boast a head count this evening at 49 on another forum and the honest head count here at well, from what I"m seeing a mere 5 to 10. But, it does not show the true count. I was for the most part not showing my online status, until I set it so others new I was online.

The true story of online or community of ETO is not seen by the new changes in the software, cool. But it won't truly reflect the people who are here and present and listening to the online post.

If anyone on the net comes, they will think we are not here and it's a Ghost Town.

Yet I know many senior members are here and listening, however not seen by guests.

So, set your online status in your "Preferences" and make yourself known!

 

[JoeJester]

killivolt,

That members online is just that, at the time you logged in or refreshed the page. Also remember that most cookies expire after an hour of no activity, so that could play into the figure.

Remember, members can lurk or participate.

The default, as I never looked into preferences till now, is show online status, and I know I didn't set mine.

Here are the stats as of 0100 on 9 Nov 2013 ... when I opened the link in a new tab ... Total: 318 (members: 8, guests: 228, robots: 82) Two members were listed as in chat.

If you looked at any of these forums that allowed you to view the membership community, you would find multiple pages, typically 30 listings per page, of people who haven't visited the site in a year. Unfortunately, here at ETO, the community isn't open to the membership, to look at the other members.

For those forums that you are a member that allow you to view the community, you can pick your favorite metric and rough guess the active membership by total pages of membership names.

For instance, the other forum where we both are members had 36870 members in their member list. Of that, 1397 have visited the site within the last month. I did not bother to extract the data further for a better classification of the membership. The front page at 0119 today, listed Members: 223,358. I can not account for the differences between the total list and the stats on the front pages. I did notice the numbers decreasing here and EM told me he was culling the forum, only because I asked.

There are reasons to leave that membership number high.

 

[ericgibbs]

hi kv,
You should include the number of Guests who are reading the Threads
Currently its:
Total: 361 (members: 13, guests: 265, robots: 83)

Over the last few years the ratio of Members to Guests logged in at any one time varies from around 5% to 10%

The total number of members appears to have fallen, this is due to pruning that Electomaster did to remove old inactive members, when the new site ware came online.

E

 

[jpanhalt]

What constitutes a robot?

I assumed they were, in part, indexing services, like Google, but the number seems kind of high to me. Are the NSA and other government agencies detected?

John

 

[ElectroMaster]

Every 'thing' that loads a web page from a site has a user-agent tag listed in the header of the request (ie GoogleBot) and thats matched to a database to identify them from a normal browser (ie Chrome).

You can view the robots:
https://www.electro-tech-online.com/online/?type=robot

Baidu seem to be a big hog!

 

[jpanhalt]

Thanks, EM.

It is nice to know the actual number of different bots is relatively small. Baidu, Bing, and Google are obvious. Had to look up Majestic-12. It has some pretty lofty goals. Should we sell Google yet? And Proximic, I guess, is for advertising leads.

John

 

[3v0]

A far better indicator of activity is the number of postings.

 

[JoeJester]

If one looked up "internet culture" at wiki, it would help explain the "guests" to "members" ratio. Guests are just non-registered lurkers.

The two electronic sites I visit listed:

Total: 1393 (50 members and 1343 guests)
Total: 459 (members: 30, guests: 340, robots: 89)

for a snapshot at 1128 9 Nov 2013.

This represents a membership to guest ratio of 1 to 26.86 and 1 to 11.33 respectively. I don't know if the first one lists the robots as guests as that discrimination is not made.

My question would be, what does one consider to be a good metric or metrics for a narrowly based forum. I lean towards participation as a metric to learn about the membership.

I'm sure the political forums have their own problems with metrics.

 

[tcmtech]

To me a simple indicator is how many and how long threads hang between posts .

There could be 500 members logged in but if only a dozen or so threads have had any posting activity in 12+ hours to me that says not a lot of people are here and talking. More likely they are like me and just have the site open on a back screen while doing other online things.

Anyone have access to a thread activity counter that says how many threads have been posted to and number of posts done overall since midnight last night or to similar effects?

 

[JoeJester]

To me a simple indicator is how many and how long threads hang between posts .

Interesting metric. That time between postings on a thread would be a function of the membership. How would you remove the "same" person on multiple threads? Conversations on a thread between three members could have hours between postings, with microbursts during the time the three were logged on. It certainly would be reflective of activity. Do you have an idea of what you think the results would be in a specific forum topic over a week?

There could be 500 members logged in but if only a dozen or so threads have had any posting activity in 12+ hours to me that says not a lot of people are here and talking. More likely they are like me and just have the site open on a back screen while doing other online things.

This is consistent with the "internet culture" of 90-9-1, where 1 percent creates, 9 percent modifies, and 90 percent reads. This was first discussed at wiki, and there were others that experienced similar results. I would suspect something near that ratio at forums.

Using the numbers above ... 3.5% (members) creating or modifying, 96.5% (guests) lurking. at the first listing. The second illustrates, and I'll remove the robots from that data, 8.1% members and 91.2% guests.

 

[tcmtech]

Interesting metric. That time between postings on a thread would be a function of the membership. How would you remove the "same" person on multiple threads?

Not really. One person hitting a dozen posts in a short time period with marginally useful info would suggest that they had some time to spend here and words to burn up but if all the posts they hit haven't seen or do not see another response for a considerable period of time, most of a day to weeks+, to me that would suggest overall slow and direct hands on site usage.

As far as actual statistics go I had a number of classes in college and as far as I am concerned they are only used for lies, exaggerations and cover ups (or at least thats' what I felt was what they were primarily focused teaching in the classes anyway) so I do not use them nor believe them to have any true validity on average.

I go by gut feeling based on direct observation.

 

[JoeJester]

I've read an interesting book that published in the 70s titled "how to lie with statistics."

I am always curious when someone, as I am sure others are when I, quote some statistic.

Yes, the statistic provider can have an underlying reason to present their stats in a manner, i.e. public opinion polls that don't quote the margin of error.

I've seen some forums that have "active members" in their statistic line ... and yet, I have not found a definition of "active member". When I did that analysis awhile back about that one non-electronic narrowly focused website where I am a member, how to discriminate between the members was difficult because I could not find the definition of an active member. That website, members could be away from computer access for 90 days as a matter of routine.

I'm working on another 90 day review of a forum and should be able to compare the earlier pie chart with one from a different forum.

I personally agree that if the forum has a area that does not relate to the purpose or goals of the other sections, all posts in the non-germane area should not increase the post count.

 

[cowboybob]

Having myself massaged data (using statictics) for very modest personal gain , I am still amused/chagrined, pretty much daily, by Mark Twain's observation that there are, "Liars, damned liars, and statisticians."

No slurs intended by the post...

 

[jpanhalt]

I agree. The persona of Mark Twain is credited with many memorable comments. Nevertheless, his satire should not be taken literally. Our society depends on appropriate application of statistics to advance. Gut feelings don't cut it, and wholesale rejection of statistics, which I don't think you are suggesting, is a fool's gambit.

On the matter membership, using percentages is certainly useful, but that use implies that changes in membership follow first-order kinetics. In other words, one obvious assumption from JoeJester's data is that two forums with the same percentage of growth are equally healthy (or unhealthy). I am not sure that model is accurate. Maybe forum growth is zero-order based on the number of existing members? Maybe forum growth in raw numbers is most related to the number of hits each forum gets from search engines? Thus, a large forum with more hits may show a lower percentage of growth than a much smaller forum with only slightly fewer hits. From that perspective, I hope JoeJester will include raw counts as well as percentages in his updated data.

Taken a step further, I just did a Google search for "electronics forums." Here is what I found for "electronics forums" at 1045 EST on 11.11.2013 listed in order of appearance:

Electronics Forums
www.electronicspoint.com/‎

Electronics Forums - All About Circuits Forum
forum.allaboutcircuits.com/forumdisplay.php?f=3‎

Electronics Forum | Electrical Engineering Questions | EEWeb
www.eeweb.com/electronics-forum/‎

Electronics forums - CNET Forums
forums.cnet.com/electronics-forums/‎

Forum for Electronics
www.edaboard.com/‎

SparkFun Forum - SparkFun Electronics
https://forum.sparkfun.com/‎

Electronics Lab - Community
www.electronics-lab.com/forum/index.html‎

EEVblog Electronics Community Forum - Index
www.eevblog.com/forum/‎

Electronics Forum (Powered by Invision Power Board)
www.dutchforce.com/~eforum/index.php‎

Electro Tech Online
www.electro-tech-online.com/‎

Notice that ordering! ETO was listed last on that page. That certainly might affect our number of hits. So, why are we listed last? I am quite ignorant of what controls listing order, but maybe something as simple as changing the index terms sent to Google would help? Maybe adding the words "electronics forums" to what Google gets would get us further up the list?

John

 

[JoeJester]

Search Engine Optimization.

Google has a tutorial. search for google search engine optimization guide

However, **broken link removed** has ETO listed first.

 

[JoeJester]

John,

Since you asked ... here is the info I used for those two graphs concerning the electronics forum sites ....

 

[Vizier87]

ETO still rules. We have the old members of wisdom. Although not as glitzy as the ones moderated by the younger ones (like me) but I prefer being here.

 

[JoeJester]

ETO still rules. We have the old members of wisdom.

I'm sure the membership of all the forums feel the same way. Even knowing there is a common pool of members visiting a few forums.

Although not as glitzy as the ones moderated by the younger ones (like me) but I prefer being here.

And other members visit a variety of forums. The nature of the beast, internet forums, is a constant flux. Members come and go. Every forum tallies the numbers "coming" but few, if they don't cull the membership and adjust the number published, maintain that membership. The same applies to the number of posts and number of threads.

I know that some forums would be lucky if they retained 2% of their published number. Personally I look at it this way ...
a. A member who hasn't visited the forum in over a year probably isn't coming back.
b. There are a lot of people who just ask a question or two, get their answers, and vanish. A lot means a lot. They typically visit once and disappear ... common traits are join date and last visit date are the same, and it's been more than a month since the last visit.
c. There are prolific posters, who haven't visited a site within the last six months. That maybe a self imposed sabbatical or simply lost interest. In some cases it could be a personality conflict with other members.

I've seen that most forums do not allow access to the membership rolls. If there were a way to disable the email link for the community listing, they might be more inclined to allow it being shown.

Why some don't change their membership number? Simple. It gives the illusion to people who advertise on these sites of a higher visitor rate, however, the click through rate will not reflect what the companies feel it should achieve.

 

[tcmtech]

Every forum tallies the numbers "coming" but few, if they don't cull the membership and adjust the number published, maintain that membership.

Sort of like religion/religious following numbers. If you added up all the numbers of what each religion claims it has for members there would need to be something like 100 billion people on the planet.

There are at least a dozen churches I have been to as a guest and signed their guest books as to which I suspect every single one counts me as a member even though I have not set foot in any of them for years to decades let alone ever contributed to their activities.

 

[JoeJester]

Here it is another month down the road ... and another growth report.

Do not take this graph the wrong way. It illustrates growth of four forums. It does not call any of them the best. If you think, one or the other is the best,that is your opinion and not the conclusion I want to illustrate.

If you were a business owner, and saw you and three of your competitors with this growth, you would have questions, if you weren't the top dog. All four of you would be trying to improve your position in the market place we call the internet.

On the internet, fresh content is the key. If one looked at the SEO, Search Engine Optimiztion, they would see that everyone recommends fresh revelent content and by the mere nature of the forums, there is fresh content daily. This does not address the quality of the content. The end user will decide that. Also we learn about bounce rate in google analytics. Keeping the bounce rate (people leaving the site after one visit) low is the goal of any website, and with a forum, it is the lifeblood.

The websites listed have some baseline data taken on 12/4 and you can see their growth next month.

The excel spreadsheet of the data can be found at **broken link removed**

One thing to consider in the basic growth graphs, is the sudden loss in the data could very well be the owner of the site thinning out the database.

In the second set of graphs, concerns the Homework Help section, which may not be the focus of the four forums offering that area. I included that because it is the one common area and it could be a check against the other graph.

In the final graph, I illustrate a large forum and a small forum. I happen to belong to both of these forums and have access to their membership database. This graph illustrates on widely know fact ... you only have a short period of time to get the attention of a newbie and there is a considerable drop in membership percentage after the first month.
In both forums, I chose a three month period to illustrate the data.

The final graph illustrates the number of members posting during a specified period. I chose day, week, two weeks, three weeks, one month, three months (one quarter), six months, and finally greater than one year. This frequency of posting is a quick look at what percentage of the membership is active, by whatever you decide is active. Personally, if one doesn't make a post per quarter, their activity is low.

It was an interesting exercise ... to keep my one brain cell in orbit.