(+11 rating)
Loading ...This is a topic those working with data are going to have to get their heads around. It’s a big topic, so best get that thinking in early. Behavioural targeting is finding initial use in the marketing space, where the ability to observe a user’s activities and adapt the advertising they are presented with increases returns quite dramatically. However, there are other applications too, from identifying children at risk and applying early intervention, to spotting employees with work-affecting issues.
Make no mistake, it is an area that is both murky and controversial. One of the shining lights in the space is Phorm – or perhaps “kicking boy” is more apt. They have had a far bit of churn in their senior management team of late. However, they were prepared to participate in a recent debate. A little while back I headed in to London for the NMK Behavioural Targeting event (very many thanks to Ian Delaney). The very distinguished panel included Guy Phillipson, CEO, IAB as chair, and included Nick Barnett, UK Commercial Director, Phorm, Rupert Staines, VP Europe, Specific Media and Ian Brown, Research Fellow, Oxford Internet Institute. Baroness Sue Miller, Liberal Democrat Member, House of Lords was due to join, but was detained in parliament (it has been rather busy there of late, and I’m not talking about the Christmas Tree going up).
The session was titled ”Behavioural Targeting: The Fire and The Fury” – although I couldn’t help noticing that the event URL was actually “behavioural-targeting-the-filth-and-the-fury”. So, why is the marketing industry pursuing this technology and why is it so controversial? You could start with the Wikipedia Definition or this FAQ from Ottawa University. Let me summarise it this way: The Internet content industry needs advertising in order to continue. Today, most websites sell less than half of their advertising inventory. People are advertising less, because advertising is becoming less affective at driving sales. No purchases, no advertisers. No advertisers, no Internet content. Using user data to adapting the advertising gets users to read more and to respond more. That is good for business. This idea is not new. The Guardian has been using certain behavioural technology on their site for some time, as have many other sites.
Most implementations use cookies, stored via the users’ web browser, to create a ‘unique id’ for each visitor. This is a number to track where the user goes, when they visit and how regularly they visit. The technology needs volume and reach to make it effective. If I see that you have read five articles on fishing, it is probably a good guess that I’ll get a better response from showing you an advert for fishing kit than I would showing you an advert asking you to sign up to an anti-blood sport campaign. A little provocative, but you can see why people are getting hot under the collar about it.
Phorm, and others, are working with the IAB, which has set up a behaviour marketing workgroup to look at the many issues around the technology. As Ian Brown rightly pointed out, this sort of tarketed ad serving has been going on since the start of the Internet. However, Phorm, and particularly their rocky trials with BT, have thrust things into the spotlight. One significant difference with Phorm is that they can collect data based on all of the sites a user visits, rather just ones within the advertising network. More than that, Phorm potentially has relationships with 3 ISPs, giving them visibility into 70% of Internet users in the UK. That has privacy activists watching their every move.
Phorm provide an explanation of their technology on their website, although it would be nicer to see this coming via validation from an independent third party. It would also be good to see similarly up front information from others, including the likes of Google. According to the organisers, they were asked to join the debate, but declined to attend.
Many users, and even IT professionals, are not aware of how much data is handed over to websites behind the scenes. I’m not the snooping kind, but even the very standard stats package I use on this site keeps track of the site you came here from, pages you visit and where you went next. It is only collected anonymously and in aggregate data, so I have no idea who you are or what you did as an individual. Most other websites collect significantly more data. Are you sitting comfortably?
The site managers, and advertisers, would argue that using this data provides users with a benefit: we get to see more relevant ads. Alan Patrick and I both pointed out that most users probably would not see that as a benefit! The issue of user benefit is an important one. Remember, we live in a country where most of us will happily surrender aspects of our privacy in exchange for a few discount vouchers – how many of us have supermarket loyalty cards? Phorm now position their solution as offering anti-phishing and malware/spyware protection – ie a user benefit. Later the discussion drifted on to the topic of the government collecting user data and monitoring, but that is an altogether different topic, although it should be noted that it isn’t a completely separate one.
A key issue raised was the nature of the relationship between the monitored user and whoever is collecting the data. You are a customer of your ISP, so you can exert pressure on them, even if that is by leaving them if you do not want behavioural data used. That isn’t the case for all users of behavioural data. It is a complex area, and not well understood. All too often users end up clicking “accept” on privacy policies and end-user license agreements that are either unread or too complex to interpret, as witnessed in the recent launch and subsequent changes around Google Chrome.
A case of surfer (and IT manager), be aware.
)
. If it is offline, then yes, it must have a copy of the page. That wasn’t how it was explained at the event – there again, it wasn’t a technical discussion. 
You say:
‘Phorm provide an explanation of their technology on their website, although it would be nicer to see this coming via validation from an independent third party.’
I suggest you start by reading Dr Richard Clayton’s blog postings, and downloading his technical analysis PDFs, which were made with the co-operation of Phorm.
http://www.lightbluetouchpaper.org/?s=Phorm