One of the draws of a job in advertising used to be its simplicity. Decide what you need to say; craft an engaging way to say it; find relevant places to stick the ads; strike a hard-but-fair bargain with the media owner; lunch; receive praise when sales go through the roof.
The industry has developed "science bits" to add rigour to the simplicity.
Groups help to frame the message; clients can "pre-test" copy to check whether ads will engage; surveys help identify the optimum placement; ad-tracking, sales monitors, call-centres and web-hits justify repeat business.
Interactive TV disrupts this benign straightforwardness. It adds layers of subtlety to a communications model built around a binary stimulus- response mechanism. This is mirrored by the evaluation processes we use, which tend to count responses.
Viewers who press red opt in to a branded communication that allows them to have a different experience. Direct response advertisers aside - for whom counts usually suffice - it's in the depth of engagement with a relevant few that interactive TV's potency lies. Interactive TV is moving beyond brochure requests - it is here that evaluation models most need to change.
Not that either Barb or Millward Brown should worry - there is still a manifest need to quantify the scale and profile of input delivery and consumer out-take. It's the bit in the middle that needs filling out. While interactive TV is growing rapidly, it needs a focus that almost replicates the process we use with linear spots.
So, first we need to know what to say. As interactive TV develops, tactics will change. But there is an increasing body of evidence from syndicated studies, TV sales houses and interactive TV practitioners such as Press Red and ZiP TV about the strategies that drive engagement.
While this helps with a general understanding, it might not be enough for an individual campaign. Say "pre-testing" in most agencies and the response is unenthusiastic. Testing linear copy gives you reductive linear results: you make one component more prominent at the expense of another.
Moreover, you expose respondents to an ad and ask them what they think of it. Ads are a trivial component of most people's lives, so it is difficult to get beyond a forced, unnatural response.
Interactive TV pre-testing allows a real, rather than forced, response.
Respondents in iBurbia's lounge-like Chiswick facilities who are engaged will engage, those confused by an interactive TV option won't fill it in. As importantly, the non-linear material allows you to evaluate the effectiveness of the call to action, the clarity and purpose of the content, the ease of navigation and the participation through to response. You can also ask them what they think about it.
Monitoring exposure is more fraught - not least because it's not so much engagement as exposure that we're after. Barb can only tell us whether someone is interacting, not how. Nor is Barb designed to cope with the relatively low numbers of inter-actors at any one time. Similarly, ad- tracking studies are not cost-effective at collecting sub-samples of people who've pressed red.
Various solutions exist, with some compelling case studies (see box), and larger solutions are in the pipeline.
For advertisers whose interactivity extends to an online request (free sample,brochure, competition etc), both Sky and ZiP use an opt-in sample of interactors as a base. This is benchmarked against people who saw the ad but didn't interact, as well as against people who didn't see the ad.
Sky is working hard to produce a panel of some 20,000 Sky homes - 5,000 of which will be on one of the Taylor Nelson Sofres sales panels (Superpanel, Fashion and Impulse). Due to launch in the third quarter of 2005, this will link interactive TV exposure to sales effects and also, in time, give a deeper understanding of how people are navigating an interactive TV site.
Perhaps the most telling statistic so far is the 78 per cent of advertisers who come back for more. We might not yet know exactly how it's all working, but we have worked out that the potential is real, and that just because something isn't simple, that doesn't mean that it can't be very good.
- David Fletcher is the head of MediaLab at Mediaedge:cia UK
THE POWER OF INTERACTIVE: CASE STUDIES THAT PROVED THE MEDIUM'S WORTH
Volvo's "mystery of Delaro" TV campaign was accompanied by an eight- minute spoof documentary - showcasing the interior and performance of the S40 - made available via web and interactive TV dedicated advertiser location.
Some 435,000 people pressed red, representing 11 per cent of Sky Digital viewers aware of the ad. They went on to spend an average six minutes in the DAL.
Continental Research's telephone survey showed that spontaneous brand recall was 12 times higher among interactors than among those who had viewed the ad but not interacted.
Kit Kat's "you are not a salmon" ad incorporated an interactive TV game that some 300,000 (Barb) individuals engaged with, the majority being 16-34 (67 per cent) and ABC1 (77 per cent) (Barb). Two per cent of these went online to register their score, having spent an average 13 minutes interacting.
Continental Research's monitoring reported 96 per cent of interactors rating the campaign as "good" or "excellent". In every case, interactors reported higher levels than non-interactors in terms of spontaneous recall and perceptions of quality, innovation, "for me" and classic.
To promote its Sonicare electric toothbrush, Philips made a two-minute infomercial available on a mini-DAL with viewers incentivised to interact with a combined free radio and prize draw offer.
Half-a-million people interacted with the campaign. Manifestly from the right 16-34 ABC1 women target audience (Ipsos), they spent two minutes each, and 35,000 registered to enter the competition. Voucher redemption among these was ten times higher than previously experienced in direct marketing campaigns.
Record sales followed, doubling brand share year on year.