The best marketing programs have intentional measurement strategies planned in advance. So as part of planning any program, you need to answer these three questions:
• What will you measure?
• When will you measure?
• How will you measure?
In almost every case, you will need to take specific steps to make your marketing programs measurable. This often includes setting up test and control groups or varying your spending levels across markets to measure the relative impact. Without variance in your marketing, you may not be able to use modeling to tease apart the incremental impact of your marketing programs and improve your marketing precision and mix.
See Section 5 for more on measuring ROI using test and control groups. Data Collection A key part of planning for measurement is simply tracking the appropriate attributes for all programs (and their variants). This can include target audience, message, channel, offer, investment level, and any other relevant attributes. Most companies do not begin this process early enough in their lifecycle, and they pay for it later.
Even if you don’t use the data right away, it will become invaluable down the road when you attempt any of the more sophisticated approaches towards measuring program effectiveness. These attributes can be stored in anything from your automation system to a simple spreadsheet hosted on a share drive – what matters the most is that you start to build the history as early as possible. “It is more important to periodically capture potentially high-impact insights than to frequently measure less important outcomes simply for reporting purposes.”