Let’s say you have an investment advisor, and when you ask him how your investments are doing, he says, ‘Well I got the stock at a very good price.’ So you say, ‘Great, I’m glad you got the stock at a good price, but how is it performing now?’
And he says, ‘We don’t really know.’
Well, you might consider getting a new advisor.
Now if you’re a marketer and that advisor is your media company, you’ve got the general situation as far as media measurement in Canada right now, says Bruce Claassen, president of Toronto’s Genesis Media.
Because your media company can tell you it purchased a certain number of GRPs or magazine pages, and it can tell you what it paid for them – but it can’t tell you what impact those ads had on brand awareness and sales.
That’s assuming your media company still only has access to the old media metrics, which are really just about measuring audience – or the number of people exposed to your advertising message – and which are just not good enough any more, says Claassen. That’s why the Canadian Media Directors’ Council is tackling the measurement issue head-on come April 2 in Toronto, at this year’s annual conference.
‘I think this industry has to realize that the measures that have been used for the last few decades have been principally measurements of eyeballs, or ears,’ says Claassen. ‘But there is a fairly weak link between that and a client’s business objectives. The relationship between measurement of audience and the actual effect – which is to influence, educate and engage consumers in transactions – is very distant.’
Fine, but is it reasonable to expect a media company to measure the direct effect of media choices on sales? Could media companies really one day look back over the last year, isolate the decisions to run this much weight in TV one week, that much weight in newspaper another, and link that back to how many tubes of toothpaste flew off the shelves?
Not only is it reasonable, not only is it possible, but Rob Young, co-founder of HYPN in Toronto, is already doing it – and for such blue chip companies as Unilever and Cara Foods to boot.
‘It’s just so cool,’ he says, describing the process. ‘You can find out all sorts of things as they relate to consumer behaviour.’
Young takes an imaginary tinned tuna marketer to illustrate how it works.
Every time a tin of tuna is purchased at the grocery store, the cashier scans in its bar code, collecting such information as the brand name, SKU and price. That information is collected in a central database, where ACNielsen packages it with additional data, such as display ratings, shelf space, media activity and even the total advertising activity for the tuna category.
Young then purchases that data from Nielsen, conveniently sent on a CD, pops it in his Dell computer and puts on his data mining helmet. He has all the information he needs at this point, but the real work is just beginning.
Because as every marketer knows, monthly sales are dependent on a huge number of variables. There’s the price, the shelf space, the distribution, the amount of advertising, the quality of the creative, and the media channels chosen to reach the target. And oh yeah, if you’re selling ice cream, you have to take the weather into account, and if you’re selling duct tape or plastic sheeting, you should probably factor in the latest bulletins from the U.S. Department of Homeland Security.
But whatever those factors are, there’s only a finite number that have a measurable effect on sales, and new software tools that allow regular folks to perform multi-variable regression analysis can isolate their individual effects. It’s called econometrics, and it’s still in its infancy in Canada, but it can help marketers determine the impact individual factors like media channels have on sales.
Along with Genesis and HYPN, Toronto-based M2 Universal has been getting into the econometrics game, and as with Claassen and Young, M2 president Hugh Dow sees it as key to the emergence of the media company as marketing partner.
‘It opens up a whole new dimension to the business relationship,’ he says. ‘I think that’s why we and a number of other media companies have really taken up the challenge of providing this service.’
Like Young, Dow says that he can already think of several cases where the results of such analyses have allowed the media company to not only advise the client on media choices, but pricing and other sales factors as well.
Dow says that with the analytical software in place, the only real limit on who this service can work for is the client’s history of collecting the right data.
The packaged goods companies are in the lead right now, thanks to having centralized sales data years ago, but Dow says financial services and automotive are also starting to take part.
‘The stumbling block is that many advertisers don’t have historical data, at least in the form that can be used for these models,’ he says. ‘I think at the moment, as an industry, we’re at a stage where we recognize the importance of this tool, but we’re just putting the measurement techniques in place. We’re really at a turning point.’
But before econometrics (or neural networks, a related method for getting at the same results), takes hold in Canada, more marketers are going to have to be convinced of its worth.
Paul Meehan, director of marketing at Vancouver-based Mike’s Hard Lemonade, for one, isn’t entirely convinced.
‘There’s a danger in measuring things way too fast after you execute a campaign,’ he says. ‘The bulk of what you’re trying to accomplish is to change the way that people think about your brand. Ultimately, you need to know emerging trends, rather than just sales blips.’
All the same, Meehan says, as a marketer, ‘you can never have too much information,’ so he sees econometrics as a useful tool – but just one of many. His fear is that lazy marketers could rely too much on the information, causing them to second-guess gut decisions and put a damper on innovation.
Dow confirms that this is an issue, at least when it comes to innovation. ‘The key to remember is that all of these models are based on historical cause and effect. What they can’t take into consideration is doing something totally new. You have to balance the need for historically-based information with the need to secure a competitive edge by doing something totally different.’
But Mike Welling, VP of brand development, foods at Unilever Canada, seems pretty happy with his company’s investment in econometrics so far. Over the last few years, Unilever has decided to take a more scientific approach to media planning (see ‘Grenades from Unilever,’ p. 8), and such an approach is already affecting the way his company plans media. For instance, Unilever has already begun shifting money from television to other vehicles thanks to findings from its Communication Channel Planning initiative.
‘Obviously, it helps to define what is providing value for you,’ he says, ‘and then based upon how you perceive that value, you then approach each one of those individual media vehicles prepared to pay only for what it does for you.’
In the end, it seems, using econometric modeling to analyze your media choices isn’t much different from reading up on cars before heading out to the dealership. ‘Like any purchase,’ says Welling, ‘the more information you have before buying your media, the more informed choice you can actually make.’
For more information on econometrics, see the session ‘Making media pay’ at the CMDC’s ‘Media’s New Metrics’ conference, taking place April 2 at Toronto’s Metro Convention Centre. Go to www.cmdc.ca for details.
