Commentary

Media Metrics: Go Forth and Multiply

There's no longer a question of going with old vs. new media, but rather how they can be used together to improve ROI. Investment in digital marketing, search marketing and social media has moved from the "garage band days," if you will, to a significant line item in marketing budgets.

Most marketers have proved the utility and impact of digital media, which has officially taken its place in their tool kits. This migration has fueled the growth of digital marketing, in some cases causing a dramatic budget reallocation away from traditional media for a significant incremental investment.

Amid this sea change, marketers are struggling to understand the synergies between online and offline media. This is a complicated, multilayered challenge that can prove painfully expensive. It can cost companies millions through what is known as "misattribution" - mistakenly crediting the wrong medium for sales. This is a problem with far-reaching consequences and serious strategic implications for future investment decisions.  

Beyond this, there are missed opportunities to capitalize on the interactions between online and offline media that in some cases, through an integrated marketing campaign, can produce a synergistic effect.

An example might put this in perspective. An online retailer we know theorized that the ROI of search advertising was four to five times greater than that of all other media vehicles. This was based on isolated search metrics that failed to identify media synergies or control for marketplace changes. Had the retailer acted on these metrics, he would have missed an opportunity: Nearly one-third of its search traffic was actually being driven by other media. By applying advanced analytics that examined all media, not just search, the retailer could uncover the synergistic relationships among search, TV and print that were responsible for nearly one-third of all search traffic. This knowledge saved the retailer from overusing search to the point of diminishing returns.

As many of our clients continue to pour money into digital media such as display ads, branded search, unbranded search, social media, microsites and widgets, we have learned much about the effectiveness and efficiency of each vehicle. One of the most interesting trends we have encountered is the synergistic relationship between offline and online media. By this we mean that media no longer operates in a vacuum - and that what you do online affects what you do offline, and the reverse. This, in turn, allows marketers to understand what's working (what to keep) and what isn't (what to kill), as well as how they all work together in an integrated campaign.

Marketers are now being asked to do more with less. Taking advantage of synergies and intergrated marketing techniques allows marketers to drive more sales without necessarily buying more media.

In trying to understand the synergies between offline and online media, marketers have encountered two major challenges:

1. Establishing common metrics in order to credit the right media: There is a statistical challenge in dealing with situations where multiple influencers interact with each other. A metaphor for this phenomenon is dieting. Imagine you are following a weight-loss program of both dieting and exercise. As you lose 20 pounds, you can easily conclude that your weight-loss program is working, but you can't say how many pounds lost were attributable to the diet versus exercise. You decide to stop one and end up gaining weight. In this example, the risk is that your experimentation slows down your weight loss; in marketing, this could cost millions.

Take TV and search, as an example. For many clients, we have found that a significant amount of search traffic, especially branded search, is driven by offline media (TV, print, PR, events, etc.). If you can't quantify the percentage of search traffic driven by offline media, then you may overestimate the impact of search and end up making the wrong investment decision. To get the right answers and make the right marketing decisions, make sure your marketing analytics account for these synergistic relationships.

2. Dealing with increasing complexity: Marketing has become complicated by the number and characteristics of digital media tactics and the various ways consumers interact with online and offline media and with each other. Our historic concept of a consumer funnel has been stretched, reorganized and disrupted by consumer entry and exit points. In the past, we used linear processes to describe how consumers became aware of a brand or product and moved through various sequential steps until they became customers. This just isn't the case anymore. Through the use of the Internet and social media, consumers can now enter the process at almost any point, share their views and recommendations with other consumers, and conduct research - jumping in and out of our traditional linear construct until they decide to buy, or not to buy, our products.

This requires us to understand which media are responsible for driving changes in consumer attitudes, changes in pre-purchase consumer behaviors and interruptions in the consumer decision process. Social media adds a wonderful new dimension and complexity to synergistic relationships. At a recent advertising conference, 28 percent of U.S. marketing executives said their focus during the next two years will be on gaining a deeper understanding of social media, and on how the resulting consumer data will influence marketing and product development.

Once we have established an apples-to-apples comparison and understand the complexities, we then need to measure the synergistic effects and identify opportunities to capitalize on these effects. Can one plus one equal three? In this case, the answer is yes. The use of four or more media vehicles to support a campaign is significantly more effective than using less - provided the creative is strong and the campaign is integrated across the different media.

A recent research project involving more than 50 brands across a number of industries quantified the tipping point: Companies that support integrated marketing with four or more media vehicles can expect an average increase of effectiveness of more than 20 percent. One durable-goods manufacturer we examined found that when display ads were included in an integrated campaign, ROI jumped by more than 40 percent.

In the case of digital media, we often find display ads to be more effective when they are part of an integrated campaign.

To get the right answers, analysts need to acquire a deep understanding of the intent of the media, be curious, and apply the right techniques relevant to their sectors. Being able to take advantage of these opportunities allows companies to significantly improve marketing ROI while giving them a competitive advantage. Unfortunately, there is no "one size fits all" tool or technique to address these questions, but in a time of shrinking marketing budgets and higher demands for marketing to produce, it's worth exploring.
1 comment about "Media Metrics: Go Forth and Multiply".
Check to receive email when comments are posted.
  1. John Grono from GAP Research, January 14, 2009 at 8:21 a.m.

    An excellent post Douglas. There is of course yet another dimension (actually multiple dimensions) at play here - what were your competitors doing at the same time? Does 100 GRPs or a million banner ads work the same as when you competitors are also using 100 GRPs or a million banner ads. Do these stimuli work the same when you are at price parity, above price parity or below price parity. The list is virtually endless.

    In my experience of such analyses using non-linear micro-economic time-series modelling, simple is NOT better. They key is that it is rare that all permutations of marketing scenarios rarely occur. While the permutations may be in their thousands the actual occurences tend to be in the dozens. Therefore with sufficient historical data, you can tend to work out how a burst of TV advertising run in conjunction with online display ads and search works when you have a price cut and when you don't have a price cut.

    Also, it is important to note that what works for the brand reported in the published research paper will almost certainly NOT work the same way for your brand. Brands are different and respond to marketing stimuli differently. These research papers are handy for giving directional information but they can not be used to quantify your likely ROI. I'm afraid that marketers have to do the hard yards, have an impecable MIS, and then be prepared to get an experienced statistical modeller to spend months poring through the data (often to find and correct the errors) in order to start to quantify HOW their marketing and sales support spends work.

    While this sounds incredibly complex, in most cases the goal should be to construct a parsimonious model (Occham's Law) with around half a dozen KEY marketing variables that REALLY drive sales (the old Pareto principle at play) that the marketer can manipulate to best effect.

    John Grono
    GAP Research
    Sydney Australia

Next story loading loading..