These measurement models are shaped by the goals that advertising programs are crafted to address, the technology that can be brought to bear on measurement, and even the marketer’s view of the role new media may play – awareness, response, engagement and more.
In many ways, the models themselves are based on surrogates for success – key performance indicators that can directionally inform the bottom line and the advertising’s contribution to it. Whatever exists today may not be perfect, but hopefully represents what an advertiser needs to be comfortable with in order to make investments in advertising, and build senior management’s confidence that the programs contribute to the bottom line.
But what happens when those ROI models shift – potentially radically?
Many things can contribute to a change in how advertisers view success of ad programs. A new CMO or head of emerging media could arrive on the scene. Or senior management could decide that the surrogates being used to generate the model are not accurate enough predictors of success. New technology might emerge in the measurement space that can shed new light into dark corners.
Whatever the reasons, marketers and their agencies should be proactive about anticipating changes, so that the marketer can make a transition to a new model as quickly and as painlessly as possible. Here are some tips:
- Always be asking “What would we recommend if we had the opportunity to improve our ROI modeling?” No model is perfect. There may be things about it that you already don’t like. Thinking proactively about what you would do differently to more accurately gauge success is never a waste of time.
- Adopt redundancy into your model. For instance, multiple studies that are run in conjunction with a campaign can increase confidence in its ROI, even if the measurement model shifts somewhat. For example, an awareness study conducted by a CPG advertiser could be supplemented with an exposed-control sales lift study. If one study shows a significant lift in unaided awareness, and that correlates strongly with off-the-shelf sales, that can increase confidence that media investments were in the right place. If the correlation isn’t that strong, it might expose a weakness in the model, but you will at least be armed with additional data that could help find a more accurate predictor of success.
- Hang on to your data. Analytics data in its rawest form, whether from a website, mobile app or campaign can be useful when your ROI model needs an overhaul. Consider that if you’re shifting metrics or looking for other success predictors, historical data can be a great ally that can prevent having to start over at square one. You might even be able to evaluate historical programs with the new model.
- Get smart with data visualization. If you’ve followed all of the tips above, it may be that changes to an ROI model might be less chaotic than anticipated if you’re good with visualization. Sometimes, just a few clicks can set up new metrics and dashboards. If the data points are already there and the pipes are there to access them, looking at success a different way might just entail a few moments’ worth of making adjustments in Tableau or Domo.
The most solid of ROI modeling foundations can be shaken, sometimes in an instant. Be prepared to see changing attitudes toward measurement as an opportunity, and don’t be left scrambling when new requirements demand change.