Nicole Moreo, Senior Vice President, Analytics, Ketchum Inc

One of the biggest frustrations with measurement and analytics in our industry is that it tends to be linear or forced to play one role. While catching up on the reams of New Year’s lists, I saw an article that seemed to clearly define the roles that analytics will have to play as we move into 2020 and beyond. The article in CIO, written by Thor Olavsrud outlined a bimodal approach:

• Mode One: analytics for how you run your business
• Mode Two: analytics that support the ability to experiment, uncover hidden insights, and then incorporate those insights into your production analytics

The article pointed out that historically even when organizations use both modes, they usually kept them separate. Mr. Olavsrud put into clear terms what has been frustrating me, and most of the industry for a while now. We are still forcing measurement to play a basic reporting role- not allowing it to play the full analytical and evaluation role it should. We are leaving data and analytics out of some of the most critical stages of communications work, and it ends up hurting us in the long run.

Effective programs and strategies need to be supported from the beginning to support planning, then during to support optimizations, and then after to learn and report. Done properly, PR measurement and evaluation should make up the bimodal approach that Olavsrud outlined- the integrated approach.

To hopefully help, I have essentially turned some of my frustrations into 5 points that I think we all need to discuss and work towards:

1. Point One: Bring in analytics, research and measurement early
   1. Analytics needs to be brought in before planning starts, at the first table read of a RFP, for brainstorms, and as the strategy is being developed. Too often research and analytics is brought in after the fact. This means that strategies are left incomplete, insights are not actually insights, and a unique opportunity to look at things from a different angle is lost.
2. Point Two: Stop only budgeting for reporting
   1. Analytics needs to be integrated into the process. This does not simply mean that a tool should be purchased. This means that budget should be built in to have analytics members join meetings regularly, research should be a part of all planning, optimization check-ins need to happen frequently, and data needs to be used. This will only happen if there is budget. Otherwise it is usually forgotten and cut.
3. Point Three: Set-up is key
   1. With all the changes in the data API space, historical data is going to be harder and harder to access. Too often measurement is left until the end. That is when it simply becomes another PowerPoint deck, another report that gets a check mark but is not used. By creating a data plan and KPI framework at the beginning, you can ensure that you will have access to the data needed, you can bake the proper insights and optimizations and reporting will resonate internally.
4. Point Four: It is not one person’s “job”
   1. As Jenna Clark from Google said in a recent interview “If you have common sense, know how to use a computer, understand objectives and KPIs, you can do measurement. It’s about understanding what your objectives are, starting with those objectives and figuring out how to track against them.” This means that analytics should be an integrated team member. Whether that means including an analytics professional, training the team, or incorporating a consultant- analytics needs to be involved- and the entire team should know the basics. This makes sure that research is actually used, goals and KPIs are followed, optimizations are recommended, and reporting aligns with the work.
5. Point Five: Never settle for guessing or rinse and repeating
   1. I tend to hear that people do not use research or analytics when planning, building a strategy or initiating a campaign because it takes too long and is too expensive. This is not true. There is a wide range of ways to research, test and ideate before you settle with repeating what you always do, copying and pasting a framework from a previous campaign, or guessing that something will work.

These points are not going to solve everything, but hopefully they can start a few discussions, change a few minds, and support a few measurement transformations.