Article

Using front-end data analysis for media planning

Anyone in media planning has an appreciation for data. But the amount and sophistication of data we’re working with today is so different than it was 10 years ago. The accessibility to complex data makes working in media really exciting. That said, data is typically used to either target media messages or to measure the effectiveness of media in delivering impressions, clicks, conversions, etc., either during or after a campaign.

At Callahan, we’ve developed an innovative front-end data analysis, which takes the way data can be used in media planning to a whole new level. I feel like we are finally able to do what we have always wanted to do, which is to have a bigger and better impact on our clients businesses.

What I mean by front-end data analytics

Simply put, front-end data analysis is about digging deep into business, sales and marketing data before media planning begins. It’s about understanding business metrics in a way that sheds light on opportunities for growth by understanding the biggest drivers of business. Those insights are so valuable because we are able to design media plans specifically based on the insight of what drives business and what doesn’t. It’s liberating and empowering from a media planning standpoint.

The benefit to advertisers, with this front-end data approach, is a focus on business results. What’s interesting is that you rarely find a client who will articulate it that way. You also never find a client who will disagree with it. Even though they may say it, I’ve never had a client hire me to buy them more efficient media impressions or a better CPM. What they really want is to sell more widgets. Typical media analytics focus on measuring media efficiency. In contrast, front-end data analysis focuses on establishing a path to the desired business outcomes, and then measuring impacted business results on the back-end.

So, tactically, as we move into media planning, the more we can shape media plans based on the insight around what does and doesn’t drive business, the more effective we can be because we are focused on the effectiveness of selling widgets rather than the efficiency of buying media. Who cares about a cheap CPM if nobody buys the widgets? The CPM is after the fact, and if you are focused solely on the lowest CPM, you may miss the opportunity to drive more business.

The way we have traditionally reported on media is a bit lazy. It’s convenient for a media planner to say I delivered what I purchased from a media standpoint and I’ve made good on my post-buy, but the ability to demonstrate the value to the bottom line is lacking, frankly. In this day and age, media planners ought to hold themselves more accountable, to a higher standard, and clients ought to expect more from their agencies.

How data lets national act locally with media

An example of how you can use business data to affect media planning is to think about the fact that even for national chains, all brick-and-mortar store sales are local. Most marketers and media planners know that no two stores are created equal, and factors like local demographics, weather, etc., make business opportunities different from store to store. National brands still have to think about the local sale. I worked with a restaurant chain that sold hamburgers. They never once sold a national hamburger. Hamburgers are sold one transaction at a time, to one customer at a time, at one distinct time, in one distinct location. Getting down to that level of detail, when you dive that deep into the numbers, you discover a variability of performance across a system.

With today’s digital media technology, we finally have the ability to act on this variability. Ten years ago, we could start to get an idea of this variability from a data analysis standpoint, but we couldn’t act on it very effectively. Advancements in digital technology (search, programmatic display, programmatic and addressable TV, social media, etc.) now provide media tactics that we can use with great levels of sophistication to take advantage of this variability.

Let’s stick with the restaurant example. We can look at a chain that has thousands of locations across the country and start to understand how the different locations perform. Then, we can develop a segmentation of these stores based on sales-related opportunities or based on their vulnerability to weather, to demographic skews in their particular area, or other business factors that we uncover by digging deep into the data. Then, we can activate media strategies using new technologies to leverage those differences. Spend more in this part of the country, less in those other parts. Do more Spanish language radio in this part of the country, do less in others. Follow rain storms in these markets, don’t in others. All levels of variability can be acted on and decisions can be made based on the impact to business outcomes, not just media efficiencies.

So why aren’t more brands prioritizing business outcomes in media planning?

It may seem obvious that granular data is available, and it can be extremely valuable in achieving business outcomes. Even so, most national and brands don’t take advantage of this approach. For example, they may still run the same media plan everywhere. Why?

First, they may not have access to the level of data analysis that will really shed light on desired outcomes and KPIs. I worked with a client recently who anecdotally would talk about the impact of weather. They knew it impacted business, but they didn’t know how. So, lack of data and analysis can sometimes be a limiting factor.

But that’s not always the case. Some marketers do have access to the data and still don’t do something with it. For some, it may simply be that changing the status quo is hard. For a lot of organizations, there’s a way of doing things, and with change comes vulnerability. An individual within an organization may be hesitant to stick his neck out to say, “We’re going to try something different.”

Perpetual Testing Environment (PTE) can mitigate the fear of change

We’ve found that there’s a way to get over the fear of change, and it’s tied directly to opportunities that arise from the use of front-end data. It’s the idea of creating a PTE — perpetual testing environment. That’s a good approach for clients who are reluctant to try something new. They can try it on without having to buy the whole outfit. Test and learn: Prove success on a small level before making bigger changes. Build internal credibility and positive momentum.

One of the watch-outs in PTE is that because it works in Denver doesn’t mean that it’s going to work in Cincinnati. Just like there are varying factors in every market, there’s going to be variability of what works where. With PTE, you continue to test and learn as you carefully expand on successes and adjust as needed. The beauty of basing decisions on business data (especially when you have access to a real-time dashboard like the Callahan Intelligence Platform) is that you’re never flying blind with your PTE.

How to get started

If you’re thinking about taking the front-end data approach outlined in this post, what’s next? Well, first you need to have the pieces in place to access the data. You must be able to identify, gather, store, analyze and visualize the data in a way that’s truly helpful. (That’s no small task, but Callahan can help.) Then, get comfortable with analytics and what it tells you to do. Some of the ways you’ve determined success in the past may be a little bit different in the new world. Be willing to evaluate plans before they’ve happened. As you’re deciding about media plan details, be willing to see these with a new, fresh perspective, informed by potential business outcomes rather than the old view using analytics to only measure media efficiency. I promise this new perspective will be rewarding!