Why multiple media in a single campaign? And at what frequency for each channel?

That’s the question Roy Patel’s team tackled at Metrix Lab when they came to chat with us at the IAB Europe Leadership Council last week. Let’s face it it’s not an easy one. The theory of combining media channels together may seem simple, but quantifying the effects is a massive challenge.

The theory goes that if you see a message in one channel such as newspapers, and it’s echoed in a second channel such as the web, then the combined effect is much greater. A first exposure in an additional channel often delivers a bigger impact than a second exposure in the same channel. When I teach marketing I usually call this the 2+2=5 effect. Add to this the fact that different channels talk to our senses differently, and you start to appreciate the complexity. Then factor in the differences in the nature of the impact – for example how radio needs a high frequency to build recognition, but cinema needs only a few views to do the same. On top of this there are the differences in the price of each media channel, the targeting, the wastage levels and the calls to action: no, being a strategic media planner is not an easy task.

And the web just complicates it all further by creating a marcoms channel that can address pretty much any part of the advertising process from awareness, through brand activation to delivering the sale. You’d need some cutting edge econometric modelling to read this landscape and distil the rules for an effective campaign.

Enter Metrix Lab: their landmark studies in strategic media mix modelling in Europe are showing planners not only which media channels to use, but the exact frequency of exposures you need to get the best results for your campaign.

How does it work? Metrix Lab look at which media channels people are viewing and from that learn about their behaviour and the type of advertising they’re exposed to. This forms the basis of an analysis of advertising effectiveness. By marrying up media channel consumption, advertising exposures, and the viewers’ attitudes to these brands, they’re able to start unbundling cause and effect.

“For me it’s all about OTS: when have people actually confronted media” explains Patel. “The real challenge is that it’s tough to really test the effectiveness advertising across different media. This type of econometric modelling needs big samples and a powerful methodology, but get it right and it changes the way a business behaves.”

Theory: follow the steps...
• The weight of each media channel in a campaign needs to be change
• New research techniques are unlocking the answer
• When they’re run they typically reveal classic media are over used, with such a high frequency of views per person that the final few exposures do almost nothing for the advertiser
• If budgets are moved into the less used media – typically including the web – then the uplift in the campaign’s results are significant
• To advertise smartly online advertisers don’t need to create new budgets, just move the budgets they have around and they can enjoy better results

Case study: Spain
“In Spain the challenge for the industry was really clear: how do you get the big spenders to change their use of the media mix, upweighting online”, says Patel. “On the McDonalds campaign we worked with MSN and OMD Digital to gain the insights, taking large samples of more than three thousand people to look at the impact of the different channels.”

The research explored the branding effect of online: awareness, image and purchase intent. The objective most under the microscope was purchase intent because McDonalds were looking for the link to sales.

Results:
In terms of the campaign, there was an integrated look and feel in the creative work across all media, but with very different messages carried in each of the media channels. The web looked like press, but different calls to action.

In terms of results, the campaign proved that online built the brand. The uplifts in brand image were a massive 19% across the overall campaign, but the role of the internet was even greater. With some of the campaigns Metrix Lab researched, the frequency of TV campaign topped 22 views per person, suggesting that there could be significant wear-out. Crunch the numbers and it revealed those last few views were doing almost nothing to boost brand metrics. In fact, after the first 20% of the TV budget had been spent, there was hardly any incremental reach on the campaign: more spend wasn’t touching more people, it was just hitting the same people again and again.

With web advertising running at much lower levels (not much more than an average of 6 views per person), raising the frequency of the campaign would still deliver a significant uplift. And for younger audiences who were less exposed to classic media, it would also reach new people. That was a finding echoed from US research for McDonald’s where it had also become clear that a large segment of 18-34 year olds just couldn’t be reached cost-effectively through television any more.

In Spain, Metrix Lab’s analysis proved that the last 10% of the budget spent in TV delivered no significant change in purchase intent. Conclusion? Ditch the final part of the television budget and switch it into online. The findings for the media planners were that the optimal media mix for this campaign should have had the web on around 9% of total advertising spend rather than the 2% actually invested. It’s a dramatic change, but it echoes the cross-media optimisation research IAB US started back in 2001. The wider implications are even greater. This campaign was a brand building campaign for a product that couldn’t be bought or even researched online. In sectors like finance, motoring and travel, where the web now forms the main place for purchase research, the mix should clearly be even higher.

Takeouts from the session?
• Pioneering study
• Campaign frequency in TV was more than 22 vs just over 6 on the web
• Media wastage was high in television
• There was a 51% increase in message association
• New research proved that media mix should have been 9% in online, not 2% for McDonalds
• Further studies from Metrix Lab have placed the web even higher, at 14% for the best results
• And online with offline is a big boost: same budget, better results.
• McDonald’s has now changed its marketing mix as a result of the work

I'm sure Roy's team will be pleased to share with you more of the details. if you'd like to talk about presenting research at the European IAB meetings then just mail me: danny@digitalstrategyconsulting.com ...or add comments on this blog page.