Facial Recognition, Hacking, and Multivariate tests – The Monday Feb 21st Roundup

Not the busiest week on the Internet. In fact, most of my reader list was left without any updates at all. Where are you guys?

That said, after much searching we have a pretty full roundup, with news on new Google facial search technology, multivariate testing in WordPress and how Anonymous hacked HB Gary. Very cool stuff.

• Bronto Blog has a pretty cool article on progressive profiling in email marketing. This is a pretty cool example of how analytics can enhance your marketing: the idea is to watch what users do and progressively tailor your marketing to them.
• Get Elastic touched on an interesting topic last week: black hat tactics that aren’t really black hat. The central idea is that a lot of things that are thought of (and treated) as black hat by Search engines can be used in completely legitimate fashions. However, I’d think that feeding paywalled content to the Google bots to drive traffic to a page is EXACTLY what Google doesn’t want. Maybe if, like Bustos describes, you let the visitor read the whole article for free, then prompt them to pay. However, if you’re feeding content to a bot, only to have users click through and find that they cant access your content because it’s behind a pay wall, then that seems to me to be a bait and switch, and thus your content is irrelevant to the user. Hence why Google de-indexes sites that do this.

• We start this section off with a pretty hardcore topic: Google’s recent patent application included in it a part about facial recognition in visual search. Some of this technology sounds really awesome.
• Came across a cool little JSON tool today: tempo. Tempo is a tiny JSON rendering agent that crafts templates in in pure HTML.

• I’ve been thinking of writing an article like this for a while: a simple guide to tracking your conversion rates using Google Analytics. Cool stuff.
• WPMU has an interesting post on using Google Analytics and their AB Theme Testing plugin to perform multivariate tests. Honestly, it all seems pretty wrap around for what is otherwise a pretty simple set of A/B tests, requiring three different copies of your theme, Google Analytics, and a WordPress plugin. Wouldn’t it be easier to just use GWO? Then again, all of your data ends up in GA this way, which lets you do some nice analysis.

• UX Matters has a post titled “Usability Testing Is Qualitative Only If You Can’t Count“. There’s some real irony in this post, as it’s about how usability professionals need to communicate more clearly, and after a quick glance over it I am not sure what he’s trying to say. He starts saying that user testing isn’t necessarily quantitative, then ranting a bit about A/B/N testing.
• UX Movement has a pretty cool post on how users fill out forms faster if labels are positioned above the text boxes. I had never thought about this before, but it certainly makes sense.

• Is using celebrities in advertising a mistake? A new study suggests that it may be, showing that fewer than 12% of ads using celebrities showed a greater than 10% lift.
• This is the single best thing I’ve read this month: the latest of Ars Technica’s series on HB Gary and Anonymous, this story tracks how Anonymous managed to break into this security companies servers. The best part? It could all have been prevented by following best practices.