A version of this piece was published in Marketing in 2008
There are, I am assured, people who walk around wearing tinfoil on their heads because they’re convinced that aliens/the CIA/the government are beaming thoughts into their heads.
They are of course, barking mad.
As, we usually assume, are all the privacy activists who go on about CCTV, oyster cards and cookies.
Two years ago I wrote about AOL, who had in a rather spectacular goof, released data they held on the search behaviour of thousands of users. They claimed it was anonymous – but it took just one day for an enterprising newspaper to track down one searcher’s identity by deducing this from her searches.
The release caused a storm, as it turned out few people were aware just how much data was routinely collected about them.
Most search engines, for instance, keep a history of every search you make, with Google only deleting records over 18 months old. Websites keep logfiles, perhaps never deleting them, of which web pages you’ve visited, what you filled in on forms, what images you view.
And it’s this that’s now got Google into trouble.
Media giant Viacom, who have been in a long term dispute with YouTube over alleged copyright infringement, got a New York judge to order that Google (YouTube’s owner) hand over internet addresses, email accounts and a history of every video ever watched on the site.
The judge, Louis Stanton, dismissed privacy concerns as ‘speculative’.
But the consequence of this is that users of YouTube, which serves over 2.5 billion videos a month to 70m users in the US alone, are now exposed. Their personal media consumption is now something for Viacom to pore over, regardless of whether they had consumed content Viacom owned the copyright for.
And as some observers have pointed out, this would never have happened if Google hadn’t collected the data in the first place.
Meanwhile back in the UK, another, connected, story has resurfaced. I wrote back in February about the BPI’s efforts to get ISPs to spy on their customers on the BPI’s behalf, punish them for infringing the law, and provide evidence to record companies.
The BPI got quite upset about this, calling my observations “quite wrong”, and claiming I’d recycled the information from the Times. They wanted to set the record straight, as they expected this story to run and run.
I had in fact got the information from the BPI’s own website.
But I quite wrongly expected this story to go away. It seemed to me that nobody would be so daft as to think they could build a business by suing their customers.
It seems though, in this I was wrong.
Last week, Virgin Media started sending out letters to customers that the BPI had identified to them, telling them that filesharing copyright files is illegal. Virgin’s view is that they’d resist cutting off consumers, preferring an education campaign.
But at the same time, the BPI sent out letters to the same users, threatening “We don’t want you to face legal action, or risk losing your internet service”.
Though I don’t download music, I do expect my ISP to guard my privacy. Moreover, I don’t think it’s any of their business what I do with my internet connection.
Just as I don’t expect the post office to read my mail or BT to listen to my phone calls, I don’t expect an ISP to snoop on my internet connection without a court order compelling them to.
Carphone Warehouse told the BPI to sling their hook – Charles Dunstone described how the fax machine in his office, unused and forgotten for over a year, ground into life to receive a fax from the BPI requesting their cooperation. Good for him.
At polar opposite ends of the modernity scale, Google and the British Phonographic Institute. One cavalier with our privacy, the other trying to get others to invade it on their behalf. I’m reaching for the tinfoil…
Showing posts with label data. Show all posts
Showing posts with label data. Show all posts
Thursday, July 10, 2008
Thursday, April 24, 2008
Buzz, giving marketers ears
A version of this piece was published in Marketing in 2008
There is a research group out there that’s so large, you can’t get it in to the biggest viewing room. And it isn’t bothered about what you think, so it doesn’t flatter you or attempt to double guess you. And you don’t need to feed it crisps and coke, or travel to Watford to watch it.
Digital media are mostly regarded in terms of their capacity for carrying our messages – the £2.8bn that’s spent on online media, and the thousands of websites that this funds are a direct product of this.
But for all the effort that goes in to advertising on the internet, a tiny fraction goes in to using it as a research tool - perhaps a reflection of the fact that as marketers we’re often better at talking than we are at listening.
In the last two years we’ve understood something new about the internet. That it’s true power comes not from the ability it gives brands and companies to speak directly to consumers, but from the power it gives consumers (or people, as they like to regard themselves) to connect to each other.
And as these billions of conversations have unfolded, marketers have started to understand that there is value in listening in.
Social networks, blogs, forums, twitter (a mobile social tool) and review sites are bulging with conversations that people are having about brands – often referred to as buzz. Sometimes they’re saying nice things about you, often they’re brutally slagging you off. But the challenge for marketers is to make sense of what’s being said – to understand who’s talking, and the significance of those conversations.
There are lots of tools available – paid-for ones like Onalytica and Buzzmetrics that give us breadth of coverage, but also dozens of free tools including Blogpulse, Technorati, Icerocket and Tweetscan.
Between them, these systems allow us to build up a picture of what people are saying, who’s saying it and how much. But knowing what to look for isn’t enough. The sheer volume of data out there means we have to know what to ignore too – someone criticising a brand in a blog that only two people read is probably not a priority.
So typically, a researcher is looking at three dimensions of buzz – Influence, Popularity and Sentiment. Flemming Madsen from Onalytica explains the difference between Influence and popularity well – in the area of childhood obesity, Jamie Oliver is popular. But if you want him to reflect your views, you’ll find it hard to get to him. He gets his information from the National Obesity Forum – in this context, it’s the forum that are influential – get to them, and you might get to Oliver.
Sentiment is harder. Although there are tools that measure this, their results can be unreliable, because at the heart of it, they’re measuring humans – and humans aren’t consistent. A teenager describing something as ‘bad’ can mean the exact opposite, and when you get a post like this (genuine one):
THIS W3B SYT IS GUD BCOZ T3ER3 IS LWDS OF P3OPL3 THT R G3TTING BULLI3D SO B3AT TH3 BULLI3S ND I AM GUNNA DO A PAGE ON MY W3BBY ABOUT IT CYA
Any machine is going to struggle to make sense of it. So you also need humans to comb the data, and distill the sentiment from it.
Put this in place, and you’ve got a thermometer of great sensitivity, which you can use for long-term projects like NPD and brand tracking. But the real power of this technology comes from its immediacy – the almost real-time feedback you can get from the world.
Gauging the impact of new TV campaign, an early-warning system for PR outbreaks, a customer service listening post – these are uses that buzz marketing techniques are already being put to.
So despite the £2.8bn we spent last year on talking to consumers online, we may yet discover that the real value of the internet to marketers is not in the voice it gives us, but the ears.
There is a research group out there that’s so large, you can’t get it in to the biggest viewing room. And it isn’t bothered about what you think, so it doesn’t flatter you or attempt to double guess you. And you don’t need to feed it crisps and coke, or travel to Watford to watch it.
Digital media are mostly regarded in terms of their capacity for carrying our messages – the £2.8bn that’s spent on online media, and the thousands of websites that this funds are a direct product of this.
But for all the effort that goes in to advertising on the internet, a tiny fraction goes in to using it as a research tool - perhaps a reflection of the fact that as marketers we’re often better at talking than we are at listening.
In the last two years we’ve understood something new about the internet. That it’s true power comes not from the ability it gives brands and companies to speak directly to consumers, but from the power it gives consumers (or people, as they like to regard themselves) to connect to each other.
And as these billions of conversations have unfolded, marketers have started to understand that there is value in listening in.
Social networks, blogs, forums, twitter (a mobile social tool) and review sites are bulging with conversations that people are having about brands – often referred to as buzz. Sometimes they’re saying nice things about you, often they’re brutally slagging you off. But the challenge for marketers is to make sense of what’s being said – to understand who’s talking, and the significance of those conversations.
There are lots of tools available – paid-for ones like Onalytica and Buzzmetrics that give us breadth of coverage, but also dozens of free tools including Blogpulse, Technorati, Icerocket and Tweetscan.
Between them, these systems allow us to build up a picture of what people are saying, who’s saying it and how much. But knowing what to look for isn’t enough. The sheer volume of data out there means we have to know what to ignore too – someone criticising a brand in a blog that only two people read is probably not a priority.
So typically, a researcher is looking at three dimensions of buzz – Influence, Popularity and Sentiment. Flemming Madsen from Onalytica explains the difference between Influence and popularity well – in the area of childhood obesity, Jamie Oliver is popular. But if you want him to reflect your views, you’ll find it hard to get to him. He gets his information from the National Obesity Forum – in this context, it’s the forum that are influential – get to them, and you might get to Oliver.
Sentiment is harder. Although there are tools that measure this, their results can be unreliable, because at the heart of it, they’re measuring humans – and humans aren’t consistent. A teenager describing something as ‘bad’ can mean the exact opposite, and when you get a post like this (genuine one):
THIS W3B SYT IS GUD BCOZ T3ER3 IS LWDS OF P3OPL3 THT R G3TTING BULLI3D SO B3AT TH3 BULLI3S ND I AM GUNNA DO A PAGE ON MY W3BBY ABOUT IT CYA
Any machine is going to struggle to make sense of it. So you also need humans to comb the data, and distill the sentiment from it.
Put this in place, and you’ve got a thermometer of great sensitivity, which you can use for long-term projects like NPD and brand tracking. But the real power of this technology comes from its immediacy – the almost real-time feedback you can get from the world.
Gauging the impact of new TV campaign, an early-warning system for PR outbreaks, a customer service listening post – these are uses that buzz marketing techniques are already being put to.
So despite the £2.8bn we spent last year on talking to consumers online, we may yet discover that the real value of the internet to marketers is not in the voice it gives us, but the ears.
Thursday, March 20, 2008
Privacy, not just a worry for people wearing tinfoil beanies
A version of this piece was published in Marketing in 2008
From Whitehall losing CDs containing the child benefit records of 25 million citizens, to customer databases being hacked, every week seems to bring another story of incompetence or criminality in the data world.
Everywhere we go, we leave a data cloud behind us. From shopping habits to tube journeys, little bits of information track our behaviour, whilst an estimated 4.2 million cameras watch our indiscretions.
When we move online, we bask in the assumed privacy it brings us. People act out fantasies, masquerading as members of the opposite sex. Vast amounts of porn are watched, and otherwise responsible people download illegal copies of Hollywood movies.
But we also value our privacy for legitimate purposes. We talk to our friends online, and prefer of these conversations to remain private. We give our credit card details to websites when we buy stuff. And as Sir Tim Berners-Lee, inventor of the Web said this week in an interview with the BBC,
"I want to know if I look up a whole lot of books about some form of cancer that that's not going to get to my insurance company and I'm going to find my insurance premium is going to go up by 5%”.
Because once we get online, the data cloud gets denser. Every search we make, web page we view, someone is tracking our interest. When in 2006, AOL released the search behaviour of 650,000 users, it caused an outcry. Few had realised just how much information was routinely kept by online companies – every search made by these users was suddenly revealed, and it made fascinating reading (if you’re a bit nosy and had time on your hands).
Berners-Lee was responding to a question posed about Phorm – the behavioural targeting company that’s just signed a deal with Virgin, Talk Talk and BT.
Behavioural Targeting (confusingly also know as BT) is the practice of gathering data about what people do online, in order to make decisions about what sort of content then to display to them.
So a media owner tracks visitors to their site, and this enables them to serve ads for gardening products to people who have visited the gardening section of the site, even when those people are subsequently looking at the finance channel.
Until now though, BT has been limited to tracking only behaviour on websites that had the tracking code installed – and with competing systems that gives each quite limited coverage.
Phorm is different though. Through a deal with ISPs, Phorm watches everything you watch, and takes notes. If you look at gardening content, it adds you to a gardening segment of its users, and enables interested advertisers to target you later. But Phorm differs from other behavioural systems in one other major way. They don’t keep any data.
The search you make, or site you look at, allows them to segment you as a user. But Phorm discards the information it scanned to conclude this – it doesn’t need it anymore. So in some senses, Phorm is a lot less worrying than Google, Yahoo and Microsoft, who keep your search history for between 13 and 18 months.
Most of us value our privacy, guarding it both from crooks and unwarranted intrusion from government. But privacy is also an area where you find a lot of people wearing tinfoil on their heads, convinced the CIA is reading their brains using microwaves, and this can make it harder for the rest of us to tell what’s really important to worry about.
So Phorm have opened a can of worms. It may be that what they do is in reality less invasive of privacy than many of the practices its critics object to. But this is ignored because Phorm are a good target, and they’ve helpfully stuck their head above the parapet. And in doing so they may have successfully spread consumer concerns about privacy beyond just the tinfoil lobby.
From Whitehall losing CDs containing the child benefit records of 25 million citizens, to customer databases being hacked, every week seems to bring another story of incompetence or criminality in the data world.
Everywhere we go, we leave a data cloud behind us. From shopping habits to tube journeys, little bits of information track our behaviour, whilst an estimated 4.2 million cameras watch our indiscretions.
When we move online, we bask in the assumed privacy it brings us. People act out fantasies, masquerading as members of the opposite sex. Vast amounts of porn are watched, and otherwise responsible people download illegal copies of Hollywood movies.
But we also value our privacy for legitimate purposes. We talk to our friends online, and prefer of these conversations to remain private. We give our credit card details to websites when we buy stuff. And as Sir Tim Berners-Lee, inventor of the Web said this week in an interview with the BBC,
"I want to know if I look up a whole lot of books about some form of cancer that that's not going to get to my insurance company and I'm going to find my insurance premium is going to go up by 5%”.
Because once we get online, the data cloud gets denser. Every search we make, web page we view, someone is tracking our interest. When in 2006, AOL released the search behaviour of 650,000 users, it caused an outcry. Few had realised just how much information was routinely kept by online companies – every search made by these users was suddenly revealed, and it made fascinating reading (if you’re a bit nosy and had time on your hands).
Berners-Lee was responding to a question posed about Phorm – the behavioural targeting company that’s just signed a deal with Virgin, Talk Talk and BT.
Behavioural Targeting (confusingly also know as BT) is the practice of gathering data about what people do online, in order to make decisions about what sort of content then to display to them.
So a media owner tracks visitors to their site, and this enables them to serve ads for gardening products to people who have visited the gardening section of the site, even when those people are subsequently looking at the finance channel.
Until now though, BT has been limited to tracking only behaviour on websites that had the tracking code installed – and with competing systems that gives each quite limited coverage.
Phorm is different though. Through a deal with ISPs, Phorm watches everything you watch, and takes notes. If you look at gardening content, it adds you to a gardening segment of its users, and enables interested advertisers to target you later. But Phorm differs from other behavioural systems in one other major way. They don’t keep any data.
The search you make, or site you look at, allows them to segment you as a user. But Phorm discards the information it scanned to conclude this – it doesn’t need it anymore. So in some senses, Phorm is a lot less worrying than Google, Yahoo and Microsoft, who keep your search history for between 13 and 18 months.
Most of us value our privacy, guarding it both from crooks and unwarranted intrusion from government. But privacy is also an area where you find a lot of people wearing tinfoil on their heads, convinced the CIA is reading their brains using microwaves, and this can make it harder for the rest of us to tell what’s really important to worry about.
So Phorm have opened a can of worms. It may be that what they do is in reality less invasive of privacy than many of the practices its critics object to. But this is ignored because Phorm are a good target, and they’ve helpfully stuck their head above the parapet. And in doing so they may have successfully spread consumer concerns about privacy beyond just the tinfoil lobby.
Subscribe to:
Posts (Atom)
