��������pk10

It turns out you can have too much of a good thing. According to research conducted by Dr Lisa Barnard at Ithaca College, New York, overly personalised ads can reduce purchase intent. Now, it’s only one study and possibly designed to embarrass for success (a student sample and acne remedy ads), but it makes the point that somewhere in personalisation lies "personal" – and it’s a line that’s easy to cross. As Dr 

Barnard put it: "Even these digital natives were bothered by this. They know they’re being marketed to. And they don’t like it." 

Yet most pundits agree there is little to stop the programmatic juggernaut, and the meteoric rise of data management platforms (DMP was last year’s three-letter acronym of choice) is further proof that there
will be ever more data to work into segmentation of audience and personalisation of ads.

Speaking at this year’s Direct Marketing Association conference on data privacy, the UK’s information commissioner, Christopher Graham, talked of data as "not only the ‘new oil’, but also the new asbestos" – it may redefine the basis of mass commercial communications but an over-reaching industry would also undo the current acceptance of advertising as a positive force in culture. 

The DMA’s own research indicates consumer sentiment moving in favour of more rather than less data
use, but even they still have nearly a quarter of the population as actively resistant "fundamentalists". 

So if it’s not exactly an open door; there’s some circle-squaring to do. How do we stop turning this new oil into the new asbestos?

Most obviously, some data is more personal. According to Millward Brown, my social-media profile and my search and web-browsing histories are, in principle, likely to be more sensitive than data around the things I’ve a declared interest in – whether hobbies, categories or brands.

Second comes control. Show me the easy-to-locate opt-out button and I’m more likely to welcome than resent you using what you’ve found out about me. And this doesn’t need to be in condensed font disclaimer-speak either – Channel 4’s All4 data-usage video, presented by Alan Carr, demystifies with clarity and charm and is prominently flagged when viewers see ads with data applied.

Third, you don’t need to draw my attention to data in order to be effective. Knowing where I live gives you the opportunity to increase my engagement with reference to local stores or a local landmark; showing my street not only feels too close for comfort, but insults my intelligence – so, of course, I will disengage.

Personalisation works like good product placement: give me the opportunity to spot the brand and I’ll draw my own inference; rub it in my face and the viewer contract breaks.

Respect the recipient
In all of this there’s ancient learning: put yourself in the recipient’s place and you will likely judge strategy effectively. Add value; have a reason; respect the individual.

Finally, we may all be looking at personalisation through the wrong end of the telescope: there’s as much danger in the hype of hyper as there is in the personal of personalisation.

Rory Sutherland points out that programmatic is too often seen as "an optimisation problem" – success driven by a sharp-end selecting and working with the most apposite data, an essentially reductive process.

Instead, it should be seen as a "de-crappification problem" – working out which large swathes of audience aren’t going to engage regardless of the finessing of the message. In a world of mass-adoption of data, such an approach – making fewer mistakes than the next guy – might well be the biggest source of competitive advantage. 

by David Fletcher, chief data officer, MEC UK