Scene: A bar, midday. It’s empty, except for one despondent customer and a bartender wiping glasses.
Bartender: Hey, buddy. Why the long face?
Marketer: I’m the CMO at that big CPG firm down the street. My CFO is demanding accountability for our marketing dollars, but every time I try to measure the impact of our spending, I run into trouble.
Bartender: What sort of trouble?
Marketer: Where do I start? First, my data turned out to not be as solid as I thought. My marketing team forgot to include a promotion or two, and it turns out that some of my local TV is measured with diaries, which are notoriously inaccurate. Diaries, for crying out loud! This is 2013, and the industry is asking consumers to write what they’re watching in a paper journal?
Bartender: Unless you’re going to finance a whole new local TV measurement system, there isn’t much you can do about that. You can use set-top box data, but since it only accurately measures household TV usage, you are giving up some target accuracy. You should set standards for how marketing activity is measured and set up validation procedures to ensure business rules are being followed.
Marketer: That’s a great idea! Here’s another problem: My digital and traditional marketing teams don’t even speak the same language. It sounds the same — with CPMs and reach and target audiences — but they define those terms in completely different ways. Take CPMs, or cost per thousand: TV buyers talk about CPMs in terms of an age-and-gender, but digital buyers talk about CPM in terms of everyone reached by an ad. And don’t get me started on gross rating points, or GRPs. Sometimes I think my online specialists don’t really understand the math behind GRPs. My traditional media folks don’t seem to understand that online exchanges allow them to buy specialty target audiences. No one cares if they’re women age 18 to 49 or not, as long as they are likely buyers. It makes me think spending is the only common measurement.
Bartender: Sounds to me like you need consistency. Everyone providing data for measurement and analysis should work off the same base. However you determine that — total impressions, target reached or even dollars spent — you have to look beyond what’s reported to what the numbers mean. Your advertising agency can help with that, so ask them for clarification. Like a bartender building a base of loyal patrons, the payoff is down the road.
Marketer: I like the idea of involving specialists to help understand media data. Even I know that reach is typically modeled, and GRPs are relative. Assuming I have accurate, consistently reported data, what do I do about my marketing mix model? I invested hundreds of thousands of dollars in it, but it doesn’t answer all my questions. It’s predicting my sales based on advertising and promotion, but since my product is a more considered purchase than most CPGs, I care about awareness as well. A model won’t measure that, will it?
Bartender: Nah, you have to use a brand tracker for that. By the way, you should probably test your creative. It’s the wild card in a lot of models produced, and it can have a huge impact on both your sales and awareness.
Marketer: Yeah, I should do more of that. I just wish my model could take absolutely everything into account, so I could really believe in it.
Bartender: That starts with you, of course. Getting your data correct and synchronized is the first step. Then you need to consider anything else that could be impacting your results. Things like the weather, the economy — stuff you don’t control but definitely impacts your sales. Even with all that, a model cannot perfectly predict every consumer’s behavior, nor can it allow for unexpected societal events that could have an impact. We’re talking probability here, not certainty. You want another drink?
Marketer: Sure, why not? You sure are a smart bartender.
Bartender: I keep my ears open.
Marketer: Just between the two of us, here’s my biggest struggle: My model shows the print advertising I championed isn’t working. I’m afraid I’m going to lose my job! It made so much sense to use local newspapers, and they worked great in my last position. Now everyone is going to know I screwed up.
Bartender: You can’t take it that way. You need to approach this as an opportunity. Every time you learn something new, it’s a chance to improve. If all you want is applause, you don’t need to spend money on a marketing mix model.
Marketer: That’s true.
Bartender: Knowing something is better than knowing nothing. Even if you don’t know everything about what works in your marketing plan, it’s better to measure, analyze and evaluate than to wait for the perfect solution. As many of my patrons say, “You can’t let perfect be the enemy of good.”
Marketer: What about the future? We’re getting into mobile advertising, apps and addressable TV. How do I determine the effectiveness of those?
Bartender: Test and learn, my friend. Test and learn. Try them out on a small scale and see if they do what you want them to. Just be sure all the stakeholders are aligned on the metrics you’re going to use to evaluate those new options. A model can’t do that for you. You have to move onto the proverbial dance floor and take a chance.
Marketer: I don’t know if it’s our conversation or the two drinks, but I’m feeling a lot better. I’m ready to move forward and measure marketing with confidence.
Bartender: Atta boy! That will be $12.50.
Marketer: I just wish I could measure the return on investment for that.
Bartender: Might be out of your league, pal.