About a year ago I traded in the dark suit and oxford shirt uniform of my investment career for the casual, laid-back, jeans and sneakers vibe of digital marketing analytics.
Hello, career 180.
When people hear about my massive career shift I almost always get the question, “is your new job completely different then what you were doing before?” The answer surprises almost everyone. It’s really exactly the same, except for I play a lot more ping pong, I get to wear jeans, and I have increased access to Cheez-its (my favorite cheese flavored cracker …). Analytics, is analytics, no matter what you’re using for data. While I went from analyzing companies’ financial statements, to analyzing how marketing effects consumers’ propensity to purchase, the same skills are required.
When I was working in the investment industry, one of my roles was looking at individual companies and providing recommendations on if we should include the company in our portfolio or not. This involved a mix of different ways to analyze data, but they were all considered, “bottom up” investing. Meaning, this data was specifically focused on the company at hand. How much debt did the company have on the balance sheet? What was their free cash flow? Were they offering any new products or services that could potentially increase their revenue? Etc.
The reality is, however, that all of these metrics were pretty useless unless they were compared to what was happening in the broader market. For example, what was the Fed up to with regards to Quantitative Easing? How was China’s economy doing? What were consumers buying? While starting with these questions is considered more of a “top down” approach, I learned that even if you are starting with “bottom up” questions, those questions must be asked in the context of “top down” information to make successful investment decisions. For example, if in the latter half of 2014, you bought an oil company with an amazing income statement and balance sheet, but didn’t understand your investment within the broader economic context of rapidly decreasing oil prices, you probably would have lost a fair amount of money.
As a self-proclaimed nerd, imagine my pleasant surprise when I realized the process of analyzing granular data in the context of more macro industry data didn’t just apply to investing, but applied to digital marketing analytics as well.
**insert snort/laugh of excitement**
At Ovative/group, we measure and analyze our clients’ media in order to drive more value – primarily financial – to their businesses as a whole. In order to accomplish this task, we ask a lot of questions and use a wide variety of tools in various ways. These tools consider more granular level marketing tactics, the company’s media strategy as a whole, as well as external economic factors. Similar to what I learned as a financial analyst, while no one tool completely solves a company’s ability to measure their media, using these tools in tandem can provide significant value in making strategic business decisions.
In digital marketing analytics, the “bottom up” analysis is accomplished with understanding individual user interactions with every media touch point (in the fancy digital analytics space, we call this “multi-touch attribution”). While there are many different types of multi-touch attribution methodologies, and many different providers, the end goal is the same: to determine the most accurate value of individual marketing tactics.
At Ovative, we use different multi-touch attribution methodologies to answer different types of questions to drive media optimization decision-making. We answer questions such as “what keywords should I spend more on during certain promotional events?” or “what display tactic is most useful for driving new customers?” While these are great and necessary questions, they don’t necessarily answer questions surrounding strategic budget shifts, or how different marketing channels relate to one another. For example, while you can optimize your display channel to drive new customers, a multi-touch attribution solution isn’t going to tell you if display is the right channel to make this optimization. There might be other channels that could more efficiently drive this strategy, but with that, we need another tool.
At Ovative, the tool we use to answer “top down” questions that revolve around a broader marketing channel strategy and budget, is Media Mix Modeling. At a high level, Media Mix Modeling conducts statistical analysis on a company’s data to determine the impact marketing has on revenue or conversions. Words like “multivariate regression” and “econometrics” get thrown around, so we’ll spare you the deep and potentially terrifying details. However, it is important to note that there are many different providers of Media Mix Models and all models are going to be different depending on the provider’s process as well as an individual company’s data. In the end, regardless of provider or methodology, what Media Mix Modeling provides us is a way to analyze and estimate the impact of marketing channels on a company’s revenue at a high level. This is the tool that’s going to tell you Non-Brand Paid Search is actually better at driving new customers than your Display channel, but it’s not going to tell you what keywords are doing the best job. One of the many examples of why using these tools together is so crucial. What’s also essential is a robust testing program, but we’ll leave that to my next blog post.
As marketers, we’re also investors. We’re taking a limited budget and investing it into different media tactics to drive value for our clients. A wise investor would never take a position without robust “bottom up” and “top down” analysis and wise marketers shouldn’t either. Having a partner who understands the ins and outs of these tools and how they work together is an irreplaceable part of a successful digital marketing strategy. That’s what we do at Ovative/group. Let us know how we can help.