30September
The headlines
size accounts 13924752 217075 1.58% size accounts_active_1 408368 -99752 -19.63% size accounts_active_30 1726643 -14080 -0.81% size accounts_active_7 1082151 6182 0.57% userinfo total 13910417 217464 1.59% userinfo updated 8611454 95834 1.13% userinfo updated_last1 150495 -46254 -23.51% userinfo updated_last30 979579 -1384 -0.14% userinfo updated_last7 569383 5472 0.97%
New accounts slip back down to their usual 1.5-2% range, but accounts_active_30 continues to fall.
Sex
gender F 3933953 94627 2.46% gender M 1929085 48292 2.57% gender U 2207609 34245 1.58%
Total sex declared: 8070647 (+177164, 2.24%) (58% of Accounts, 94% of Updated (+1).) Males are still closing the gap on females, and it's slightly faster this month, the F/M ratio stands at 2.039 (-0.002).
Age
Here's a list of the ages with at least 100,000 people:
age 15 190430 -1448 -0.76% age 16 341323 -5807 -1.70% age 17 534597 2658 0.50% age 18 637538 8774 1.38% age 19 680319 15777 2.32% age 20 649152 17798 2.74% age 21 591379 15721 2.66% age 22 520185 19416 3.73% age 23 414379 16057 3.87% age 24 342622 12076 3.52% age 25 282149 11430 4.05% age 26 237896 7196 3.02% age 27 235500 12715 5.40% age 28 153838 7187 4.67% age 29 128654 5566 4.33% age 30 110810 4799 4.33%
Total declaring an age: 7119188 (+197092, 3.75%) (88% of Sex, 83% of Updated (+2%).) Modal age remains 19. Quartiles come at 18.9, 21.7 (+0.1), 26.2 (+0.2). The creep older continues.
Top 20 Countries
country US 3298762 -104638 -3.17% country RU 506125 6307 1.25% country CA 284167 -5941 -2.09% country UK 250214 -4137 -1.65% country AU 118641 -1772 -1.49% country UA 62176 1291 2.08% country SG 48841 1381 2.83% country PH 47161 -1061 -2.25% country DE 44815 -176 -0.39% country FI 36166 97 0.27% country JP 29647 -92 -0.31% country NL 24138 -282 -1.17% country BY 20298 380 1.87% country IL 18781 45 0.24% country NZ 17362 -215 -1.24% country BR 16962 -95 -0.56% country ES 16747 -133 -0.79% country FR 16263 -46 -0.28% country SE 11851 -73 -0.62% country IE 11048 -140 -1.27% 224 other countries: 269184
Total countries declared: 5149349 (-111751, -2.12%) (72% of those declaring an Age (-4), 64% of Sex (-3), 60% of Active (-2)). No change in the order of the top 20. It's clear that there's been some sort of clear-out of dead data, affecting the old accounts more than current ones. The gaps beneath the 20: Ireland's lead over Italy 60 (-134), over India 216 (-37), over Malaysia 745 (-109).
The section labelled usertrans_upgrades_to_plus has been removed from the live statistics page as quietly as it was added.
Other news this month
Six Apart has a new Chief Scapegoat, Chris Alden. He replaces Barak Berk O'witzless, after seeing his Rojo creation bought up by Six Apart, and promptly killed at the first decent opportunity.
It was alleged that Six Apart began automatically charging credit cards without the explicit consent of the customer. Staff claimed that this was due to user error, but many customers stated that they had explicitly turned automatic debits off. Discussions amongst customers suggested that the payment page was confusing; we agree that the default is to create automatic payments, but this should not allow Six Apart to force payments back on.
Remarkably, such abuse of financial information would not be an obvious breach of Six Apart's proposed privacy policy. It would, however, represent a flagrant breach of law for all users in the EU and Canada, and we assume in other parts of the world as well. Customers who have been charged incorrectly may wish to seek redress (specifically, chargebacks) from their credit card issuers.
Like a flock of magpies in a Shiny Things factory
Six Apart also introduced a completely useless negligible thought waste of space: thumbnail graphics for external links. These are delivered by a third-party company, and can be blocked: add spa.snap.com/* to your Adblock or LMHosts lists.
Regular defender of the indefensible Anil Dash asserted, Regular people on the web *love* Snap previews. It was clearly a shortage of time that prevented Mr. Dash from linking to any of the many replicable, peer-reviewed scientific studies that confirm his (literally) bold claim. And it is clearly a failure in our ability to co-ordinate seventeen reliable search engines that we've not been able to find any of these studies ourselves. We've been able to find plenty of people telling how to block it, and some >arguing the philosophy, but no actual evidence that the product is, in any meaningful way, actually popular. Even the makers could only cite a paper that summarises as: this is an advert for our product, and did not adduce anything that might pass muster as replicable or peer-reviewed scientific evidence.
In an act of supreme pettiness, Six Apart removed the logon box from its precious navigation strip. In theory, this forces customers to use a page containing advertisements when they wish to logon; in practice, customers with an ounce of sense will have bookmarked a permanent member's user information page, and will use that. This, of course, negates many of the reasons for the navigation strip's continued ubiquity.
Six Apart introduced XFN™ data to the Userprofile pages. The implementation is more miss than hit, assuming that everyone listed on the customer's Reading list is a friend. The XFN™ definitions require:
friend - Someone you are a friend to. A compatriot, buddy, home(boy|girl) that you know. Often symmetric.
Two other definitions in this category are available: acquaintance - someone with whom you have exchanged greetings; and the still weaker contact - someone you know how to get in touch with. We would suggest that no stronger relationship than acquaintance should be assumed for the customer's reading list.
There's a greater philosophical problem here: XFN™ background document states,
When you annotate a hyperlink to someone with rel='friend', you are saying that you are a friend to that person. Whether or not you consider yourself to be a friend of someone else is something under your control, something which you yourself are the authority over, and thus it makes sense that you be able to specify it.
Yet Six Apart is making that decision for its customers. Simple accuracy and good manners require that people must be able to define their own relations, perhaps by using a simplified version of the demonstration XFN™ creator. We wonder why Six Apart has rushed to publish inaccurate data without allowing the submission of correct data; a cynic would say that the company wished to impose its view of the social graph over that of its customers, something we've previously argued against. That attitude riles us about the company's policy of Identity Theft Through Open ID.
Six Apart has introduced its own non-standard attribute, group; this is not defined in the specification, and our reading that such one-to-many relationships are explicitly outwith the XFN™ scheme. Well-built parsers will correctly ignore this proprietary extension. The "also friend of" list provided in each profile (though sometimes hidden using CSS) does not have an XFN™ relationship defined; we submit that this should be tagged contact.
