Kirk Klasson

The Search for El Dorado

Mad men, production functions and social media marketing

Years ago, a colleague of mine, Barry Clapp, explained to me his theory of advertising. I’m sure I don’t remember the subtleties but it went something like this…. Demand for any consumer good or service over any prospective population of buyers is essentially random, consequently the value proposition for those goods and services must be constantly communicated in order to intercept the demand when it occurs.

Now the predictive marketing types are shaking their heads and muttering about big data and recency and a bunch of other stuff. That notwithstanding, the fact remains that demand results from numerous factors that don’t always align in a convenient manner and the messaging that’s meant to promote that demand can frequently fall on deaf ears. Put another way, the efficacy of advertising is never 100%. So the allocation of resources to promote demand is tricky under any circumstances. Today that’s become even more complex by the fragmentation of communication channels and the rise of social media.

So, in today’s cacophonous world, how does one go about allocating messaging resources across a multitude of channels, including social media, and to what effect?

Recently eMarketer declared Facebook the hands down winner of the social media marketing race as measured in social media ad spend declaring that it should top $4.0b in 2011. Also weighing in was comScore with “the Power of Like” as a testament to the networking value of Facebook’s members when leveraging friends and fan-based relationships. But what exactly does that mean with respect to the efficacy of social media over other types of media whether that be conventional or internet based?

PwC that measures these types of things has been watching media ad spend for a while and has predicted that internet ad spending will likely be the fastest growing advertising media for the foreseeable future, growing at an annual rate of around 8.0%. And since ad spending itself isn’t growing, in fact it seriously declined in 2009 to the tune of about 14.4%, in the face of a 2.8% decline in consumer spending, and only came back by 5.4% in 2010, the internet must be taking some serious share from the 8 other conventional channels (TV, radio, newspapers, etc.) that PwC measures. Given recent economic conditions over all ad spending isn’t likely to grow significantly since there is no intrinsic elasticity attributable to messaging in the face of a double dip recession, 9% unemployment, and consumers who are busy deleveraging after the last big buying binge.

Now internet ad spend is itself a pretty noisy space including categories as broad as utilities (search, shopping, proximity), information services (news, opinion, blogs), entertainment (games, gossip, music, video), not just social media. In fact, by some estimates, there are at least 20 different categories of internet channel and each category can have numerous players. By 2013 PwC estimates that internet channels will garner about $36b in ad spend in North America alone. But in the face of all these options it would be good to know which internet channel can be expected to produce what results.

According to MarketingSherpa that recently conducted a survey of social media marketers, the results that advertisers are seeing vary widely. Most agree that when it comes to increasing awareness that social media is by and large and effective messaging platform.

However, when measured solely on the dimension of revenue generation , the results aren’t nearly as positive. In fact, if the primary criterion for social media ad spending is an expected increase in revenue, this might make a marketing exec think twice. What it does suggest is that social media might not be the exclusive channel for the realization of revenue goals but could be an effective compliment in a larger marketing program.

This notion of social media as a marketing compliment was recently substantiated by research conducted by Telnor for iPhone sales as well as the results that Wrigley is seeing in the use of gaming as recently reported by the WSJ.

Telnor, a Swedish telecom company, measured the propensity of a consumer of iPhones to influence the buying decisions of other consumers through the use of social media and determined the influence was statistically significant. Not a big surprise given the cult like following of Apple consumers. What might be more interesting would be to see the role that Apple retail stores played in the same correlation of sales.

Next, Wrigley, the gum maker, has been driving sales of its “5” gum through an elaborate use of different types of social media but mostly based on the roll out of a game called “The Human Preservation Project”. While mentioned on Facebook and other media the hook is that clues distributed with the gum give players of the game an added advantage. Seems to be working as sales of the gum have hit $500m in five years.

So one challenge facing marketers anxious to exploit social media would be to determine how resources allocated to social media best compliment other resources and channels both on line and off to drive results.

The idea of combining compliments to yield a specific result has been around for a long time in the form of a production function, a concept that was originally developed to optimize manufacturing yields through the refinement of inputs but has recently been applied to the allocation of marketing resources to achieve revenue results. Basically, a production function is a recipe. You combine this much flour, sugar, and salt, bake at a certain temperature for a certain amount of time and you will get a pleasing result.

The graphic at the top of this post is a marketing production function that was developed by researchers at Ohio State University, Notre Dame and the National Taipei University and published in a paper entitled “A Direct Approach to Evaluating Technical and Allocative Efficency in Marketing” to describe the optimum allocation of marketing resources for a services firm operating in 21 distinct geographies. While such techniques are not necessarily common to most business their adoption is likely to grow given the fact that marketers must now decide how to allocate resources over 8 conventional advertising channels each with numerous demographic and geographic options and an internet that has at least 20 discrete categories of channel each with their own numerous options. But if that weren’t complex enough there are as many variables associated with a given prospect as there are with a given channel or bundle of channels, not the least of which are things like age, education, disposable income, etc. A prime example of why this type of analysis is increasingly important was reflected in a recent report that indicated that AT&T was one of the most aggressive advertisers on social media. Is there somebody out there who isn’t aware of AT&T? If, as some experts contend, social media works well at building awareness, not necessarily sales, for lesser known brands, then one might wonder whether AT&T’s social media advertising was money well spent.

So how would one go about determining the advertising efficacy of social media?

The truth is you probably can’t. The variables have as much if not more to do with the pool of prospects as it does with the media employed to reach that pool. While consultants will tell you all about the networking value of your customers, the revenue potential of site membership, the yield rate of discrete insertions at the end of the day you have to deal with real dollars in the door. In the scheme of things social media will likely be at least as good as advertising in other media which suggests how one might go about establishing a benchmark, a floor really, of expectations.

In 2009, according to readily available stats, US consumers spent approximately $9,990b across all categories of expenditure predicated on approximately $161.1b in ad spend across all media channels including the internet. This is not a strictly apples to apples comparison since the ad spending numbers probably include some B2B stuff. So the simple yield is approximately $62 of consumer spending for every ad dollar spent. This figure is probably low as the ad spend is total but consumer spending is only a portion of measured economic activity. Extrapolating that with the US advertising expected to be spent with Facebook by eMarketer ($2.19b) and you get expected US sales of $135.7b. Not an unreasonably big number and probably one that you might be able to use to baseline your expectations. Perhaps another way to establish a baseline would be to find a publicly traded company, if one exists, that exclusively uses social media and see what their marketing to sales ratio performance is over time.

Obviously, your mileage may vary but when it comes to marketing spending you might want to develop a plan for the allocation and measurement of resources to compliments, including social and conventional media, on line and off, before you strike out in search of El Dorado.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Insights on Technology and Strategy