Phorm revisited

Back in January 2008 I met with a couple of guys from the behavioural advertising company Phorm. Shortly after, I wrote a pretty rosy blog post about their proposition.

Positivity about Phorm has been, and continues to be, unpopular, but I haven't worked in display advertising for over a year now (I moved into paid search) and so I hadn't given much thought to my one-sided initial appraisal of Phorm since writing it. (Some will suggest that I didn't give it much thought at the time!) Last week, however, twitterfeed suffered a technical 'hiccup', and all my blog posts were accidentally automatically tweeted about, as if they were new posts. My old post about Phorm was brought back to the attention of my twitter followers, and to that of anyone who was searching twitter for Phorm-related tweets.

As you might imagine, I've since received one or two twitter mentions of criticism. I owe many thanks in particular to @zootcadillac who suggested I revisit the topic, and who sent me a link to this recent (October 2009) report of the All Parliamentary Communications Group.

In the report, evidence submitted by Phorm makes a forceful case for the legality of their system, and for their principles of collecting and storing as little data as possible (much less than Google, for example). Testimony from Gill Davison, on the other hand, suggests that allowing Phorm's proposed interception of ISP data would be akin to allowing one merchant to covertly listen in on a customer's phone call as they ordered services or goods from a competing merchant. Needless to say, as Davison points out, this kind of practice is not permitted in telephone communications.

Imagine a world where covert communications surveillance was allowed, and indeed was commonplace. If one merchant could do it, without repercussion from the consumer, no doubt all would do it; a level playing field would result. Davison's analogy may not be totally damning then.

However, such a world would not be right (and will never exist). It's not the merchant's rights we have to worry about, it's the consumer's. As the All Parliamentary Communications Group concludes, Phorm's proposed system is an opt-out one, and one that cannot be allowed. Worse still for Phorm, they have already conducted secret trials of their technology, where even opting out was not possible. They have been condemned for this, and rightly so.

With cookie-based behavioural targeting systems, the consumer is in complete control of their participation. If they want to disassociate themselves from any targeting profile that has been established, they merely clear their cookies. Moreover, together the IAB and Google provide facilities for the consumer to actively prevent a lot of cookie-based behavioural targeting from operating on their browser.

The interception of ISP data in ad targeting systems would allow the consumer considerably less freedom. As the All Parliamentary Communications Group report points out, Phorm will not be operating in the UK in the foreseeable future.

Take a look at the report and form your own opinion. It addresses a number of hot topics, in addition to behavioural targeting. Thanks again to @zootcadillac for sending me the link.

Phorm and the Open Internet Exchange

I met a week or two ago with a couple of guys from internet technology company Phorm. They presented what struck me as an exciting proposition: a network of business relationships, of which phorm would be the administrators, named the Open Internet Exchange, or OIX.

Phorm's OIX will involve publishers, advertisers and, crucially, internet service providers in a way that will allow advertisers to target internet users whose behaviour on the internet identifies them as having relevant consumption interests. This is great: behavioural targeting identifies consumers who have actually started down a particular road to consumption (purchase, subscription, or whatever); rather than just identifying those who are only likely to start down this road, because, for example, of their demographic.

With behavioural targeting, once the user is identified as a target, they can be shown relevant advertising while they're viewing any internet page, not just pages whose content is relevant to the marketing proposition. The advantages here are that publishers can monetize all their content, even that which isn't easily categorised and wouldn't otherwise be valuable; and advertisers can avoid the clutter of competitive ads that forms on the pages whose content pertains to the advertising.

But behavioural targeting is something that advertising networks do already. So what makes the OIX different? Exisiting methods of behavioural targeting work like this:

1. a user suddenly appears, browsing one of the sites in an advertising network


2. he or she exhibits while on that site a degree of behaviour that identifies him or her as a potential target: he or she conducts a relevant search, clicks on a relevant ad, or looks at a relevant page


3. when he or she next appears back on one of the sites in the advertising network, he or she gets shown a relevant ad

The limitation here is that the behaviour of the user is observed only while the user is in the advertising network. Some eighty or ninety per cent of the internet population might be reached at least once each month by traditional advertising networks, but only a fraction of a user's total internet behaviour is actually observed by any one network. This limits the volume of relevant users available, and the accuracy of their match to the advertising.

Phorm's OIX, which involves internet service providers, will make use of the ISPs' data on users' entire browsing histories: it will 'see' all the pages they visit, and all of the searches they make. This will allow adveritisers to define and successfully target a practically infinite number of behavioural advertising 'channels' - combinations of keywords and URLs that define specific behaviours.

Sounds good, doesn't it? - like a more highly evolved approach, that makes use of a massive resource that has been under our noses all along. But, while it employs one resource of the ISPs, for it to work Phorm's OIX must sustain another: the user. While the user won't have to download any software onto his or her computer, and no personally-identifiable data will ever be recorded, there will be cries of 'Big Brother is watching us!' Moreover, users will be generating a revenue stream for all those involved in the OIX. So what's in it for them?

Well firstly, we all see advertising anyway. What Phorm is proposing with the Open Internet Exchange is a way for us all to see more relevant advertising - advertising that might even be helpful!

Secondly, Phorm is in a position to provide the user with whatever services that can be derived by the monitoring of his or her internet behaviour. The obvious ones here are internet security services, like warnings against phishing websites.

Given how important it will be for them to preserve the ISPs' good relationships with users - to keep the user sweet - I wonder if Phorm will be able to come up with anything else ground breaking, by way of services for the user. Indeed, they may have to come up with something of this nature before the OIX will take off. Whether Phorm can provide something where ISPs have long had the opportunity to do so already will be interesting to see.

Keeping ahead of the advertising

It was reported on Tuesday that the entertainment giant Viacom, which owns MTV and Nickelodeon, will be launching a billion-dollar lawsuit against YouTube and its owner Google. YouTube, Viacom argues, does not take sufficient responsibility for removing copyrighted Viacom content from its video sharing site.

The most common response to this news that I have read about has said that Viacom is making a mistake: that they are struggling against a market whose natural flow is in fact in their best interests. As Jeff Jarvis observes, Viacom is 'trying to spread stupid', by criticising the actions of its own fans, who are publicising Viacom content.

The clever action, so goes Jarvis' argument, would be for Viacom to do as CBS and, in fact, the BBC are doing: They are condoning the posting of copyrighted content on YouTube, not only by leaving it on the site, but by replacing it wherever it exists with high-quality copy. CBS's and the BBC's idea here is to use YouTube for the marketing tool that it provides, concentrating YouTubers' attention around single high-quality copies of their content.

As one comment on Jarvis' blog puts it, however, the argument is not so one-sided: YouTube is illegally hosting the best of Viacom content, deriving its own potential revenue from it, and developing but containing its own audience - because they don't necessarily have to go anywhere else to get the Viacom content they want.

Basically, Viacom aren't stupid, they do have a point, and perhaps in their case they should be taking the action that they are.

The CBS/BBC model is a good one, but the merit of it is found not so much in the 'clever' strategy itself, as in what it intends for the quality and nature of their content. The secret to making use of YouTube as they plan, will be for CBS and the BBC continually to offer, on their own channels, fresh, appealing product, which competes with what should be regarded only as its own advertising on YouTube.

It will be a shame if Viacom wins a lot of ground against YouTube in the outcome of this lawsuit: the CBS/BBC model is one that will only drive better content.

... but with great power comes great responsibility

It was reported last month that five families at a court in Los Angeles jointly filed a negligence and fraud suit against Rupert Murdoch's News Corporation. The company owns the social networking site MySpace, which enabled paedophiles to groom and sexually assault the children of those families.

I first read about this case shortly after writing my last blog entry, and I think it provokes a discussion that follows from it: about the responsibility that new media providers must be willing to accept if they are to protect the health of new media. Empowering media consumers as broadcasters is a dangerous business. By encouraging us all to become personal broadcasters, new media providers are giving each of us the ability to devastate our own privacy in an instant. It's not just Google or News Corporation that we share our information with, so that they can show us adverts; imagine the consequences of accidentally sharing a personal document with the whole internet, instead of a small group of friends. And it's not just issues of privacy that accompany broadcasting. Media providers are also empowering individuals to commit libel and infringe ownership rights.

One immediate response to these concerns would be to say that it's down to individuals and the government to be responsible for these issues of identity and interaction; and it's true, we do have personal responsibility, and of course the government does too. But I think that responsibility for these concerns should lie most of all with media providers, because assuming this responsibility is in their own best interests. Let me explain why:

As I mentioned at the end of my last entry, moral integrity and trustworthiness are likely to become increasingly valuable as a consequence of changes in the branding environment. This somewhat superficial reason isn't the only one though for why media providers will need to be moral, trustworthy, and responsible: they will need to be these things in order for new media, for consumer content creation to work at all.

As they collect and help us to share more information, media providers are demanding more trust of consumers than has ever been asked of them before. In addtion to the information that we actively upload, media providers are in a position to collect information relating to every single discrete action we take in using the internet. When a transfer of information occurs, it's already being recorded:

• where geographically in the world the consumer is; and more specifically, what internet service provider he or she is using. This is to become increasingly important when, as with Virgin's new 'quadruple-play' TV, landline, mobile phone, and internet offering, the channels of our media consumption ('old' and new) begin to merge.
• When exactly the information transfer was made;
• exactly what information was transferred;
• and how, i.e. with exactly what media device the transfer was made.

That last category of recorded information is very sweet icing on the cake for the media industry. By using 'cookies', the data gathered about each of our information transfers is made uniquely identifiable, and so can be made continuous with the data on all our other internet activity.

We're looking in new media at the ready availability of a massive amount of information on consumption; and this information is without the bounds we've seen traditionally in media research methodologies, like those of sample size, human error, research budget, as well as others. Where online media consumption can be linked to purchase - and that's where consumption itself is not already purchase - there's a veritable cornicopia of data relating the media we consume, to how we spend our money.

My point here though is mainly to encourage trust, and to urge media providers to warrant it, rather than to warn consumers against it. There's incredible potential for insight in all this information, and I'm sure that that can be for the good of everyone involved; but, and here's the very significant rub: media providers must be responsible for and have the trust of consumers if they want us to invest our information.

Thinking now for a moment of the work I did at the marketing research agency Millward Brown, I'm not sure what all this new-found power in new media is going to mean for the study of consumer attitudes. In a world where on-demand, interactive new media is the only media, it might not be necessary for us to bother trying to research consumer attitudes at all, when we would have a near-perfect picture of what media people are exposed to over time, and the effects of that exposure on media consumption and purchase. In such a world, in the terms of consumer insight, attitudes become consumption itself.

I would probably guess also that advertising pretesting, as we know it - out of 'field' - will become a redundant practice. What would the point be of having a separate out-of-field process for pretesting a piece of advertising, when it's actual performance in field, amongst a perfectly controlled audience, can be monitored, perfectly?

However I can see it being useful to be able to apply to our ongoing study of how consumption is impacted upon in new media, that which we understand already about how in traditional media attitudes can be changed. And I don't mean to suggest either that the quantitative analysis of consumption data is ever going to replace the creative, brain-storming processes of qualitative research, or anticipating patterns in consumer behaviour before they form.

And once again, as with the future of the advertising industry, we run into a brick wall when it comes to the number of people actually 'on' the internet. All the traditional ways of examining consumer attitudes remain of paramount importance in the undeveloped world. Whatever man-power that might be freed up by efficiencies and redundancies in the new media world would probably be wisely invested in developing digital media on the ground, in the places where it hardly yet exists.

Our Innocence Lost

You may have seen an article in Tuesday's Evening Standard about the rise to fame of Mia Rose, whose homemade music videos have made her the most-viewed person on the video-sharing web site YouTube.

It got me thinking about the effects of increasing interactivity in our media environment. We can't ignore the fact that interactive modes of media are becoming increasingly popular, so it seems worth wondering what the effects of that interactivity are on the way media works.

We can be sure that interactive media engages consumers more actively than traditional modes of media: a consumer reading and posting a comment on a blog site will certainly be engaging with the blog more actively than if he were being presented with the same information in a television programme. This leads consequently to the potential for interactive media to be comparatively much more communicative to its audience than traditional media. Interactive media is also guaranteed to be more relevant to its audience, as consumers only interact with media that's relevant to them. Interactive media then, is more engaging, communicative, and relevant, which must make it more effective.

The reasons for this additional effectiveness of the media make it a good thing in itself for consumers, but increased effectiveness is also a good thing for advertisers. Media is, of course, the vehicle for advertising; more-effective media means more-effective advertising.

Through effectiveness though is not the only way that both consumers and advertisers benefit from interactivity. Interactive channels provide consumers with the ability to create their own content. Interactivity then probably also makes media on average cheaper, both to consume and to advertise in. Now that those we have traditionally thought of as media consumers are generating content themselves - for free, at home or wherever they are - there's a lot more of it available.

The rise of interactive modes of media - and I'm thinking mainly on the internet - has made it free and easy for anyone to communicate to a global audience. This is likely to make it more important for authors to be able to identify themselves effectively (when they want to); which is the same as it becoming more important for branding to be employed effectively. Branding is going to become more significant than it already is - if that's possible.

The article about Mia Rose in the Evening Standard actually focused on accusations that Mia's YouTube campaign has been conducted secretly by a PR team. (See my last post, 'Big Bully', for more on artifice and engineering in the media industry). You might have been forgiven for imagining that an increase in the significance of consumer-created content would cause the nature of interactive media in aggregate to tend further towards being 'honest' compared to traditional media: after all, the traditional consumer loves and demands honesty, and has no hidden agenda. The accusations levelled at Mia Rose however suggest that this is not the case.

The truth is that whoever creates its content, a degree of what we might describe as deception is inherent in all media, whatever its mode, and that in fact, there is even more opportunity in free and easy, interactive, consumer-created media for this deception. The Mia Rose case illustrates the effect that the removal of the line between consumer and content creator can cause: the creator can easily pose as a humble consumer - that's the corollary to the fact that the interactive media consumer can easily also be a content creator. It's central too to the nature of most interactive media that a single person can have an infinite number of different digital identities (real or otherwise). No truth in consumer-created content is necessarily the whole truth, if even the truth at all. Every single media consumer now has the potential to be multiple content creators, and to be on a level with all other content creators. With that potential have come the agenda that are universal to all content creators: not just the desire for audience, but also that which often drives it: the desire for advertising revenue. Services such as Google's AdSense for publishers have made it very easy for any creator of online content to generate advertising revenue from it. For the interactive media consumer, innocence is a fond memory; the interactive media consumer has come of age.

And there's certainly no going back; this is not a passing fad. There is nothing like the economics of an advertising industry in which the number of advertisers potentially matches the number of consumers to sustain the rate of growth of content production. This growth will in turn sustain the advertising industry. So long as the necessary infrastructure - like online storage space and interactive media services - remains in place, the only limit here that I can think of is the number of people in the world with access to the internet.

An interesting side-effect of this evolution in the media I think will be an increase in the importance of the function of branding to lend credibility to a content creator; while at the same time, branding's capacity for this function will deteriorate, as branding becomes cheaper and easier. Those brand values - i.e. brand image attributes - that are associated with moral integrity, trustworthiness or (somewhat paradoxically) innocence will become even more valuable than at present. Once again, Google, who continue to do a lot to drive the evolution of media, springs to mind, with their slogan: 'Don't be evil'.

Mia Rose on YouTube
Blogger
MySpace
Google AdSense for publishers