Copyright © 1995 Robert Seidman (robert@clark.net). All rights reserved. May be reproduced in any medium for non-commercial purposes.
Last week I wrote about TimesLink severing its relationship with Prodigy in favor of moving to the Web. Unfortunately, I didn't clarify exactly what TimesLink is. This was mainly because I've mentioned it several times in the past, but I forget that we're always adding new readers-sorry about that!
TimesLink is an online offering by the Los Angeles Times newspaper. It is an area that contains the newspaper, bulletin boards, chat rooms, an archived database of newspaper archives and more. Despite losing money, TimesLink is a very nicely done service. Unfortunately, not many people want to pay extra for it...
It's been a while since this particular "headline" has graced the pages of "In, Around and Online". Too long! But we're making up for lost time. Oh the places we will go...
*How Many Subscribers*
The first stop is last Friday's (10/6) Wall Street Journal and a story by Journal reporter Jared Sandberg, entitled, "On-Line Services' User Counts Often Aren't What They Seem".
In Sandberg's story, he covers many of the same issues that have been covered in this newsletter-that CompuServe counts 900,000 Nifty-Serve accounts (for new readers, Nifty-Serve is a Japanese online service that CompuServe helped design. CompuServe collects royalties for Nifty-Serve accounts). Also, that number has apparently gone up, 2 months ago, CompuServe was reporting 850,000 Nifty-Serve accounts. Then again, CompuServe is now claiming 3.6 million (up from 3.4 when last we checked), this means 25% of their numbers are made up of Nifty-Serve accounts. AOL has a "royalty" arrangement with Apple's eWorld. Apple licensed the software from AOL, but AOL does not count the 90,000-100,000 or so eWorld customers in their subscriber totals.
Sandberg's story also covered that Prodigy counts members, not accounts-if more than one person or "name" is signed up on an account, they count them all, and as a result, they claim 2 million. In Sandberg's story, Prodigy spokesman Brian Ek tries to justify the practice of counting "users" rather than accounts. "Advertisers are interested in eyeballs," claims Ek. True, the problem is, if a Prodigy account has 5 "users", there is really no way to know if those users represent 5 sets of eyeballs or one. So, the number is meaningless. Even to advertisers, who care about eyeballs. America Online also allows multiple users (up to 5) on a single account, but they don't count screen names, but single accounts.
America Online, as it turns out, does count accounts that are in their 30 day trial status. Sandberg's story quotes an AOL official as claiming they count the numbers because 85% of the people that take the trial accounts become paying subscribers and spur new revenues. Unfortunately, AOL won't say how long the newly converted members stick around.
I have already received several letters from people assuming that this news means AOL counts all the trial DISKS that are floating around as customers. It's not that bad! They just count the people who actually sign-up for the trial. We don't know exactly how many AOL disks are out there, but we're very sure it would account for a figure a LOT larger than the 3.5 million accounts AOL currently reports!
To be fair, I wasn't aware that America Online counted the trial accounts. I was curious as to how long this practice had been in place, so I asked America Online President Steve Case.
"We have always counted subscribers the same way (for 10 years) so what we report is apples-to-apples basis," said Case.
Apples to apples versus themselves, yes, and it isn't a stretch to imagine that the other services also count their trial customers. CompuServe and Prodigy have taken a very AOL style marketing approach recently, scattering disks from the mountains, to the prairies, to the oceans, well, you get the picture.
While the AOL counting practice may not be new, I confess, it was news to me. While I never specifically asked Case, or AOL spokeswoman Pam McGraw whether they counted trial customers, I had been told by people at AOL that they didn't count an account until it was 45 days old. This is symptomatic of the numbers game played by all the online services, because if you ask several different people the same question at ANY online service, you're quite likely to get a different answer.
Case in point, I asked several people at CompuServe about how they counted the Nifty-Serve accounts, and whether they counted them at all. At first, all I got were "NOs". Finally I spoke with Michelle Moran, an official corporate spokesperson for CompuServe. Moran confirmed that they were counting (at that time) 850,000 Nifty-Serve accounts.
The problem is, had I not begun writing this newsletter, some industry insiders would have never "informed" me that CompuServe is counting close to a million affiliate accounts. CompuServe is very up front about this, IF you ask them. But you'll never see CompuServe mention the 900K Nifty-Serve accounts in a press release, nor will you see America Online put an asterisk next to their subscriber totals to indicate that they're including trial accounts. You won't see Prodigy using an asterisk either.
It could just be that they don't want us to know...
Steve Case would not comment on specific numbers regarding how many of the reported 3.5 million plus accounts were trial accounts, but, if we figure they are adding about 50,000 subscribers per week (which based on recent growth, is about right), then at any given time, AOL's figures might have about 200K of "fluff" in them. If you then figure that 85% percent of the trial accounts go on to be paying subscribers (an astounding figure which Case confirmed), you're only talking about 30,000 or so accounts.
It's not a significant number, but it is still an overstatement. Even 200K or so is a small % of AOL's actual subscribers (about 6%). While that might not be significant, it is still more subscribers than GEnie, Delphi, eWorld, and probably Microsoft can claim.
Case, in fairness, is rightfully more concerned about revenue than about subscribers.
"The revenue growth has actually OUTPACED the member growth for each, of the past six quarters," he said.
"Reported revenue has consistently grown faster than reported subscribers, so obviously there is very, very strong underlying growth," Case added.
But Case supports efforts to standardize counting practices.
"...the online industry really needs standardized reporting so everyone can compare apples and apples. We've been pushing for this for several years with ISA and SPA, so far to no avail," said Case.
Next Friday (Friday the 13th no less), I'm scheduled to be the electronic guest speaker for an interactive chat session for members of the Interactive Services Association (ISA). While this chat is a "private" chat for ISA members, I wanted to use this opportunity to plug the ISA. Since all of the major online service providers are members of the ISA, the ISA has an excellent opportunity to drive the issue of counting standards. For more information on the ISA, please check out their Web site at <http://www.isa.net/isa/>
*How Long do They Stay*
Sandberg's article also touched on reporting of online "churn", which is a measure of how many people cancel the service. Over a month ago, AOL reported that the average AOL subscriber stays online 41 months.
About a week ago, I asked AOL spokesperson Pam McGraw to verify that figure. I received confirmation very quickly. Then I sent her an e-mail asking her if she could point me to someone who could explain the churn calculation to me. I didn't understand (and still don't), how the figure could be 41 months when over 2/3 of their base had joined in the last 24 months. I still haven't heard anything yet. In AOL's defense, I've seen some pretty ridiculous churn calculations in my life, and those were just the ones meant for INTERNAL review. Coming by churn numbers from any online service is difficult. Thankfully, we have surveys...
*Line 'em Up*
This information comes from Norwalk, Connecticut based INTECO Corporation which recently did a study of online services numbers, with emphasis on cancellations. This findings were based on " on large-scale interviewing in U.S. homes, with over 10,000 interviews completed through August of this year".
INTECO reports that as of August, there were nearly 5.6 million U.S. homes subscribing to one of the big online services. That's the good news. The bad news, if the study is correct, is that over 6.2 million homes have tried an online service and subsequently canceled it.
I've taken some numbers that come directly from the INTECO press release about the study which was issued on Friday, October 6 and plugged them into the little chart below.
Note: The study measures accounts in terms of HOMES in the United States-this was only a study on the consumer side. As a result, CompuServe, who has many business customers, over a quarter million International customers, PLUS the 900K Nifty-Serve
accounts in their 3.6 million figure, was reduced to 1.2 million homes (about a third of the overall number they report) as of 8/95. Also these numbers are approximations as the press release used a lot of phrases like "nearly", "over" and "more than". Since some "households" subscribe to more than one service, if you add up the number of subscribers below, you'll get more than the estimated 5.6 million.
Number of Cancellations Current # Subscribers Service (Historical Total thru 8/95) (as of 8/95) ------- ---------------------------- --------------------- America Online 2,900,000 3,400,000 CompuServe 2,100,000 1,200,000 Prodigy 3,500,000 1,700,000 Delphi 496,000 93,000 GEnie 372,000 46,000 Internet Services 573,000 1,500,000
America Online comes out looking the best in this, and that's not a surprise because their interface lets them reign supreme as the king of easy to use services. Also, their pricing, while expensive by the "flat-fee" Internet model, is at least very easy to understand.
Still, AOL only keeps about 54% of those who try their service, so they probably want to improve on this very badly. It's much bleaker everywhere else though.
CompuServe retains only about 36%, Prodigy about 33%, Delphi about 19%. As for GEnie, ALL the other services can boast that they've LOST more subscribers than GEnie's ever had. GEnie comes in at about a 12.3% retention rate.
I know what some of you may be thinking-"See, the Internet is winning!" In a way, you're right-they retain 72% of their subscribers. 72% is an amazing figure, especially compared to the online services.
But when you consider how long the ISPs have been "popular" versus how long the major online services have been around, and that a lot more people have tried an online service, that number isn't quite as significant. Don't get me wrong, the number is still significant, we just can't yet be sure of how significant it is.
This is some pretty ugly news for everyone folks. I'm not sure that this was a very scientific survey, but I wouldn't quickly dismiss these results. They seem reasonable, and I hope to speak with INETECO's Rob Rubin next week to follow up with some questions about the study. One thing is for sure: we've come a long way, but we still have a long way to go.
"While the on-line industry continues to experience excellent growth, our study clearly indicates that there is little brand loyalty and that there is a significant amount of dissatisfaction with all of the services," said Rob Rubin, vice president of INTECO in the press release.
"What we found was that many home PC users switch from one service to another to take advantage of free trials or to try out new features and they eventually cancel the less satisfactory of the two to avoid paying two monthly minimum charges," Rubin added.
*HIT Me With Your Best Shot*
Over 50 people forwarded me a blurb that ran last week in the New York Times (or the abstract of that story that ran in Edupage) which summarized a report from the Coalition for Advertising Supported Information and Entertainment. The report in essence said, that hits are a misleading measure that imply "that the user has been exposed to the deeper material, which may never have been selected for display on the screen."
It was gratifying for me to see these stories, not because they echoed my own sentiments, but because this is the first real step towards awareness by a critical mass. As more and more become aware, standards for measuring numbers will become a necessity. The national press these issues are beginning to receive, ensure, at least in my mind, that we're not too far away from an apples to apples world where the numbers actually mean something.
*Still MORE Fun With Numbers*
Summary results of a survey commissioned by O'Reilly and Associates and Trish Information Services were released last week. Billed as the first "statistically significant" survey, the study concluded that 5.8 million Americans (people, not homes) or 3.7 percent of the U.S. adult population, had direct Internet access. An additional 3.9 million people, according to the study, subscribed to a commercial online service like AOL, CompuServe and Prodigy.
The survey, "Defining the Internet Opportunity," found that over 2/3 of those with direct Internet access were male. While this shows that the Net is made up of mostly males, it wasn't so long ago that the estimates were 1 female out of 10 Net users. That it's up to 1/3 is encouraging. The study also found that 55% of Net users are between 18-34 and the 18-44 age range accounts for 80% of Net users.
According to the survey, the household income for the largest group of users was $35K-$50K per year, accounting for 22% of all users. An additional 21% of the users are in the $50K-$75K/year range. Of Net users, 65% have a household income of over $25,000, keeping in mind that 15% of the respondents refused to answer the question.
The survey resulted from three phases. Phase one was an e-mail survey to people that registered for the GNN Web service (which was owned by O'Reilly, but is now owned by America Online). They mailed 2,052 people who'd registered at the GNN web site a survey. We're not sure how many responded. Then, they dialed 200,000 phone numbers randomly and screened over 32,000 to come up with 1,000 Internet users to survey and 500 users of online services to survey.
To find out more, check it out on the web at <http://www.ora.com/gnn/bus/ora/survey/index.html >
There's even more studies on the way, so stay tuned...
Those who like to read this on the Web have noticed that the Web site has lagged behind the LISTSERV version significantly. To be honest, the Web traffic has been a lot less than those who read it via e-mail (unless you count ALL hits), but I received several impassioned pleas which basically said, "PLEASE update your page, it's soooooo much easier to skim your newsletter that way!"
Fair enough. Starting this week, I pledge to have the Web site updated by Tuesday morning each week. Yes, it's true that the LISTSERV version will still go out first, but I'll be adding some interesting nuggets each week to the Web site which won't be in the newsletter. Like what is AOL's procedure for determining whether a GIF violates its Terms of Service (TOS), pictures from France and Israel (as soon as I get the newest version of software from Seattle Filmworks <http://fine.net/sfw/>, which will run on Win 95 hopefully, next week-they get major points off for not making the software available via the Web site, but other than that, I'm very pleased with their service), and other fun stuff that won't really take me too much time.
So, if you care to know that bare buttocks in a GIF don't violate AOL's TOS, and want to find out what does, keep your eye on: < http://www.clark.net/pub/robert >.
For the many that have asked-I haven't forgotten about you or the coffee mugs. Here's the scoop, I wanted to find a company that would take on the whole process of making the mugs, collecting the money for them, and process the orders, including shipping. They all wanted the money UP FRONT. Hey, I can't blame them. And hopefully, you can't blame me! And no, I am not asking for you to contribute!!
(Seriously!)
Moreover, the one complaint I get about the newsletter above all other complaints is:
It's true, when I meet people, they say things like, "Oh, you write that out, about and around online newsletter!" "In, Around and Online," seemed to work pretty well when I thought of it, but I was thinking more in terms of "tens" of people reading it, not thousands.
So, I have been thinking about other names. I ruled out anything with NET or CYBER in it immediately. I'd go with "The Bennett, Case, Kurnit, Massey and Siegelman" Report," but that doesn't exactly roll off the tongue and with the exception of Case, nobody seems to stay in one place very long.
And some of my esteemed online counselors have advised me in fact that I SHOULD put my name in the newsletter's banner and have offered up suggestions like:
Seidman Online or Seidman's Online World. Actually, my ego isn't that big, but this isn't an issue of ego so much as pronunciation. I pronounce my last name "Sideman",
but there's a fairly reputable bunch of bean counters who spell their name the same way, but call themselves "Seedman". It's bad enough that people think of me as Robert Seedman, I certainly don't want to deal with people thinking it's "Seedman's Online World".
But, soon, I will change the name of this newsletter, and it just might be Seidman Online. Actually, I like "Inside Online". Who knows? When that happens, the newsletter will have a new name, so it will need a new logo. And as much as I appreciate all of the "grass roots" entries for the logo contest, I'll probably seek something professional.
Either way, the coffee mugs aren't dead. I like the concept that was originally posed which mimics some of the magazines that offer coffee mugs in return for a good leads or information. And yeah, they'll be for sale, too. And no, I still won't want to make any money on them...
STOP THE PRESSES! Seriously, if you're interested in news about news and newspapers online and the net, you'll want to read Steve Outing's daily (Monday-Friday) column called Stop The Presses. Available exclusively on the Editor & Publisher Web site at: <http://www.mediainfo.com/edpub/ >. Focused information that's free! Check it out!
Three times a week, Educom comes into your mailbox with brief news abstracts on all that is going on in the high tech world, with a heavy emphasis towards Internet and online. So, if you have Edupage, in weeks like these, where NewsWorthy notes gets squeezed, you'll still be in the know. It's FREE, what are you waiting for?
To subscribe to Edupage: send a message to: listproc@educom.unc.edu and in the body of the message type: subscribe edupage Steve Case (assuming that your name is Steve Case; if it isn't, substitute your own name).
I had so much fun with numbers this week that I didn't get to fit in MSN's future plans for off-line reading or a summary of what's going on in the International marketplace. Hopefully that info, along with NewsWorthy notes will return next week. Maybe even a little discourse on Apple's plans for the Internet and eWorld.
See you next week.
Shades of Octobers Past? Bad week for AMER, PSIX-Meanwhile, Microsoft and Spyglass settle down-SPYG had an especially good week the week before on the news they'd be the developer and prime licensee of Microsoft's and Visa International's Secure Transaction Technology (STT) for Windows. Maybe UUNT was just a little too close to Microsoft this week.
Bonus: The American Stock Exchange has teamed up with Ziff-Davis' Inter@ctive Week magazine to come up with the "Inter@ctive Week Internet Index" which will also be known as @NET. Options on the new index will begin trading on the AMEX beginning October 18 under the Ticker IIX. From the press release:
" The internet Index consists of 37 companies, including Internet service providers, commercial online services companies, Internet tool companies, multimedia publishers and networking companies. The index also includes providers of technology expected to shape the future of the Internet, such as video conferencing and interactive television."
I'll add IIX to the list below once it starts trading. You can also follow it on Inter@ctive Week's Web site at: <http://www.zdnet.com/~intweek >.
This Last 52 52
Week's Week's Week Week
Company Ticker Close Close High Low
------- ------ ------ ------ ------- -------
America Online AMER $63.75 $68.75 $74.50 $14.94
Apple AAPL $35.69 $37.25 $50.94 $33.63
AT&T T $64.00 $65.75 $66.38 $47.25
Bolt,Beranek & Newman BBN $36.25 $37.38 $39.38 $12.63
FTP Software FTPS $25.38 $27.75 $35.50 $20.25
General Elec. GE $62.88 $63.75 $64.63 $45.38
H&R Block HRB $40.13 $37.13 $47.25 $33.00
IBM IBM $94.13 $94.50 $114.63 $68.00
MCI MCIC $25.44 $26.06 $27.13 $17.25
Mecklermedia Corp. MECK $14.00 $18.75 $24.38 $ 2.13
Microsoft MSFT $85.88 $90.50 $109.25 $54.38
Netcom NETC $42.00 $44.00 $47.63 $16.75
Netscape Comm. Corp NSCP $63.25 $62.00 $75.00 $45.75
NetManage NETM $22.00 $23.75 $27.25 $10.38
News Corp. NWS $21.38 $22.00 $25.13 $14.38
Performance Syst. Intl PSIX $16.63 $21.50 $25.50 $12.00
Sears S $35.88 $36.88 $37.63 $21.50
Spyglass Inc. SPYG $38.25 $45.75 $54.00 $26.50
UUNET Technologies UUNT $39.63 $46.25 $51.75 $21.75
I began writing this newsletter in September 1994, at the time I was working for a technology company that is now owned by MCI. In March, I began working for International Business Machines Corporation. As of July, my management has agreed to allow me to do some work on the newsletter during business hours (probably about 6-8 hours a week). I speak for myself and not for IBM.
To subscribe to this newsletter by e-mail:
Send an e-mail message to: LISTSERV@PEACH.EASE.LSOFT.COM In the
BODY of the message type:
SUBSCRIBE ONLINE-L FIRSTNAME LASTNAME
Example: Subscribe Online-L Robert Seidman
If you wish to remove yourself from the list please DO NOT reply to this message -- send an e-mail message to:
LISTSERV@PEACH.EASE.LSOFT.COM with the text SIGNOFF ONLINE-L in the body of the message.