Media analytics has become cool and it’s a good thing because data is the only way to prove marketing’s value to executives who control budgets. You can’t talk “impressions” to a CFO; you need to talk dollars—X dollars in gets you Y dollars out.

So media analytics are critical for convincing the C-suite that marketing is an investment, not just an expense. But analytics are hard. Everyone wishes for a single, simple platform that answers every question. But this is impractical, and pining after it deters marketers from tackling the complexity to be able to optimize their media investments more often with more confidence.

What makes media analytics so complex?

  • The tension between the need for macro and micro data
    • The data sets needed to answer essential macro questions about timelines, annual budgets and scope are different than the ones relating to micro questions about the performance of specific tactics and how to make media perform better
  • The inherent messiness of media tests
    • There’s always a difference between what you want to test vs. what you can test
    • Furthermore, many variables are out of the control of the marketer
  • Consulting firms that want to sell their special “hammer” even if the problem isn’t a nail
  • Hiring, promoting and keeping qualified data analysts in a hot new field
  • The variability of language
    • Words like attribution mean different things to different people in this young field

So where does that leave us? While analytics is still defining itself, it seems the most productive approach is one that uses small, focused tests, along with purpose-built measurement tools. This lets marketers adjust media spend throughout the year to iterate more often. Finding lots of small effects and stacking them up leads to the greatest impact in performance. Seeking a silver bullet gets you fool’s gold.

The first step in finding those small, meaningful actions is pulling together the right analytic resources. They aren’t likely to come from a single source.

1. Build an internal team. It’s not easy, but an objective internal team is essential. These are smart people who can not only analyze, but dive into the weeds and evaluate available tools, and the merits of a particular hammer. To build a team:

  • Find a dynamic leader—a director or senior manager who understands all the technical details but can also talk to people throughout the company about the value analytics brings; you lose the argument if analytics is just the geek team in the corner.
  • Get buy-in for the investment across the organization—the leader should be able to explain how each constituent and stakeholder will benefit.
  • Build a team of at least three—there need to be junior people to handle projects and tasks, and levels within the team to offer a career path.

2. Hire a specialist firm—when needed. Experts are useful, and there are many good ones out there. But they need internal folks who can direct them. Experts are also expensive. So to make sure that money is well-spent, be able to check at least one of these three boxes before hiring.

  • They have unique expertise that you don’t, and can’t imagine being able to hire.
  • The project is limited in scope—limited scope means it doesn’t make sense to bring on new employees in Q1 that you’ll have to let go in Q2; it also prevents an open-ended commitment that could lead down a never-ending rabbit hole.
  • The specialist offers scale you don’t have.

3. Tap into the expertise of your existing vendors. They already know your business. You probably are already sharing data with them. They may even be working in your office. Because of that business relationship (vs. consulting firms) existing vendors can often offer a more cost-effective solution.

Focus on closing the loop

Together, these resources can turn a company into a nimble learning organization. That’s what’s needed for analytics to help your business. Think of a generic media cycle. It goes from plan, to create, to execute, to evaluate. You only learn things when you close that loop, when you evaluate: this worked, that didn’t. The faster you complete the cycle, the more frequently you learn new things and the quicker you can take action to improve the performance of your marketing investments.

There’s no silver bullet for media analytics. Rather, there are lots of little opportunities to do things better. The suite of resources outlined above helps achieve that. For more on this topic, view the Adweek webinar, “Why Marketers Don’t Need a Grand Analytics Solution.”