It actually stands for Completely Automated Public Turing test to tell Computers and Humans Apart, and while they’re a pain in the arse they do serve a purpose – helping sites work out whether you’re a human or an unthinking robot. If you’re filling out a reCAPTCHA, in particular, console yourself that you’re also helping recognise words that have defeated the computers processing real book scans.

Wasted attention

However you feel about CAPTCHAs, I guarantee that you have never, ever, seen them as a wasted opportunity to be advertised at. Have you?

Well, all hail Solve Media’s TYPE-IN™, a proprietary branded CAPTCHA “which places relevant brand messages where users are already engaged”.

I encountered one of these for the first time this morning, and felt that being forced to sit through a short video advertisement before seeing a conventional CAPTCHA was offensive enough to merit swears. Isn’t the whole point of a CAPTCHA to stop automated spammy adverts spoiling a website for humans?

According to Solve Media, the ads offer “Guaranteed engagement for advertisers”, which is true in a sense because one particular format of TYPE-IN asks you to parrot back a “Brand Message” to prove that you’re a human.

That’s right; obediently mumbling back some advertiser’s unique selling proposition goes to show that you’re a human. Rather than, say, some kind of empty, unthinking machine.

It only took an hour or two to get unhacked, but we spent considerably longer puzzling over 4li’s manifesto: “If you hate Islam, We hate anyone who does not hate you!”

4li, we don’t hate anyone, except those two blokes in the BMW 1 series adverts. And hacking isn’t very nice.

I love getting these emails. They almost invariably contain a codicil explaining that you’re receiving the email because you signed up to receive it. Almost invariably you didn’t.

I’m a particular fan of this work by RS, who, unless I’m mistaken, I once ordered from back when I was running Hand Painted Dog, perhaps seven years ago.

Translation:

We’re really good with data. Shit hot, in fact.

Here’s a way we use data that you specifically opted out of.

We know that you specifically opted out of it, but we assume you must be an idiot.

Here’s a big button you can press to fix it.

Which is total cock.

The story explains how BA responded to the first of two cabin staff strikes by publishing videos on YouTube which, the magazine’s research suggests, are “winning over customers”.



Leaving aside any consideration that the one video I watched was a dry exercise in corporate spin, the article’s interpretation of the primary survey results is flawed.

The survey asked 1,000 consumers about BA’s campaign. Of those who’d seen it, “nearly 19% had an enhanced view of the airline”. Sixty-four per cent felt the same about the carrier, while “only 17%” said they had a diminished opinion of it.

The margin of error

There’s not much between 17% and 19%. If your sample size is 1,000 it’s 20 people.

But more importantly, look at the story’s exact wording (my emphasis):

The survey of 1,000 consumers, carried out with market research company Toluna, found that the majority of those who had seen BA’s campaign have either retained or improved their perception of the brand.

Which implies that not all of the sample had seen the campaign.

I’m going to go out on a limb here. The video I link to above has around 22,000 views at the time of writing and BA’s total YouTube views run shy of 130,000. Unless people are gathering in stadia to watch BA’s channel uploads on a single PC, that suggests hardly any of the general public have seen BA’s YouTube videos.

Charitably assuming that, perhaps, 10% of the Toluna sample were familiar with the campaign, that two percentage point difference amounts to two people.

Statistics are bent and twisted all over the backwaters of the news to make convenient arguments for vested interests. I’d always assumed that the liberties people took came from a need to further their own agenda. Perhaps that’s the case here, but I find myself wondering if it’s simply that marketers and PR workers can’t do the sums.

Or that they can, and they assume the rest of us are too stupid to undo them.

I feel like they know me already.

I’m fascinated by this stuff, because it’s shit, but also because if Google News is going to allow it it drives home a depressing realisation: from day one, Charlie and I could have put all our ‘journalist’ crap aside and just rewritten optimised press releases as news – don’t think nobody wanted us to. We’d have hated ourselves and gained even fewer readers, but clearly they’re not the point.

So I’ve been thinking more about exactly why companies want psews on their websites. Make no mistake that it’s all about the optimised links, but a quick chat with @randipdhesi helped me better understand what’s going on. They’re not getting link equity from Google News, of course, but they are getting a degree of traffic, and there’s always the chance that people linking to the news story will – like I did – forget to add a no follow.

Linking 101

By omitting this, and thus conferring its (limited) authority to another site through a link, Hackbash is suggesting to Google that it rank that site more highly for search terms similar to the anchor text used. By way of an example, I know of a jolly good B&B. At the time of writing, search results for that exact term all relate to a pub in Hampshire. Within a day or two, now I’ve written that, the B&B closer to my heart should rank alongside it.

So that’s part of it, although of course you can’t control the text with which people link to your site*. But third-party sites that link to yours confer a more general sort of authority on it, and by careful management of links within the receiving site this ‘Googlejuice’ can be distributed in a controlled fashion.

By ensuring that any part of the site that attracts inbound links contains only optimised links that point to the site’s own product landing pages, a company can channel some of its incoming link love to its products. This helps them rank more highly when people search in Google for product-related terms. Like prestige car hire, commercial vehicle hire or fucking car lights.

Randip makes the point that a newsfeed is particularly attractive, because there’s no shortage of third-party sites who’ll syndicate, or simply rip off its content. In some cases the links stay intact, resulting in an optimised link to a product page from a third-party: the best of all worlds.

Tamsin’s written on the iCrossing blog about how Google’s useless in a crisis. To an extent that’s unsurprising; something as fast moving and serious as the Chilean earthquake is always going to generate a challenging amount of information and, as Tamsin says, much of this came via sources created specifically for the job. Google’s search algorithm wasn’t designed to reach an instant value judgement, and yet increasingly it needs to.

But news aggregation relates to established and ongoing sources, and as such it’s an area where a search engine has the time to fettle and refine its service. By now, Google News should be shit-hot. It’s tempting to say it’s half-way there.

* Search agencies, including iCrossing, do ask, and some webmasters are happy to oblige.

I’ve seen this advert on the site a couple of times now. It’s easy enough to ignore but… Wait, hey. I’ve got two messages? Well I’d better click.

The site you get if you do is another hawker of individual voluntary arrangements as a financial cure-all. What, no messages?

Phooey.

UK Debtors note: there are many companies purporting to be a ‘National Debt Helpline’ or similar – a google search will show you the ones who’ve spent the most on optimising their sites.

I’m only aware of one charitable example: the National Debtline, run by the Money Advice Trust. Call them.

We use Google News among our sources. I’ve written before of my concerns regarding its apparent lack of judgement, but I still find it useful – particularly its new(ish) custom section feature.

Recently its results have begun to include the peculiarly car-headlight-bulb-focused Autobulbs Direct News, providing such fetishistic takes on motoring news as:

Bentley launches new vehicle, with gloss black finished car light casings

and

Car lights of Spyker supercars to be fitted in Coventry (where, in fact, the whole cars are to be assembled).

We’ve seen this kind of deal before: take a press release, rewrite it, crowbar in the term you’re trying to optimise for, publish and submit to Google News. Even so, a recent story went too far, mis-attributing a focus on headlights to a Virgin Money spokesman:

[Grant Bather] suggested that this could include cracked car lights or other damage caused in a traffic accident or claims that result from breakdowns or theft.

This is just naughty. There’s no mention of headlights in Bather’s quotes, or indeed anywhere else in the Virgin release.

Taken in isolation, tucked away in a news feed that’s written for Google, perhaps this kind of search-term loading hardly matters, but it depresses me all the same. I expected Google to become more wise than it was in 2007 to the quality of the news it aggregates.

Back then, in a few months of optimism, it was tempting for Charlie and I to believe that we were part of a potential new direction for journalism. Like us, we assumed, other journalists would want to maintain decent standards when writing news for the websites of corporate clients, and that Google would help by favouring those who bothered.

For me, our two motoring news feeds are all that remains of that belief.