Oil and data are much alike, but perhaps the most important commonality is this: In raw form, each is virtually unusable. If you want to fuel your car with oil, you need to refine it first. In the oil refining process, “impurities” are removed, and the remaining elements are selectively reconfigured into more useful products.
Similarly, raw data must be cleansed and transformed into actionable insights to fuel consumer packaged goods (CPG) advertising campaigns. Uncleansed data can muck up your campaign engine. Data collected without viable privacy practices creates risk. Low-quality raw data sets can turn out to be deep potholes, putting campaign results at risk.
A future-proofed marketing strategy consists of consumer purchase insights where brands have made mid- to long-term investments. Advertisers should feel confident the data they are using to build their advertising strategy will be around for the long haul.
Sufficient in scale. Today’s CPG marketers place a high value on insights sufficiently large enough to scale the entire United States, representative of the current marketplace and actionable.
But at some point, it becomes just as important that the source data is diverse and balanced.
CPG marketers can take a prudent approach to data management, balancing out their data portfolio to include all types of retailers, grocery stores, receipt capture, drug stores, convenience stores, and ecommerce sources. It’s important to continually evaluate your data sources for gaps -- for instance, too few sources of grocery store data -- and take steps to close them.
Representative of the market. To deliver representative CPG insights with precision and accuracy, data scientists start with consumer purchasing data representative of:
With this nationally representative foundation, data scientists have the raw ingredients to derive a view that provides an accurate, ongoing view of household shopping behaviors. Shopping patterns are continually changing, and data resellers that don’t adjust their sources to account for this will have outdated purchase models.
Actionable and applicable. A data set can be representative and sufficient, but without granularity, it lacks application. Data with application is like refined oil, ready for consumption.
Using newer technologies like machine learning, data scientists learn purchasing behavior at the household level and make predictions about consumer purchase activity. These models allow marketers to observe or derive, in a privacy-safe way, the current and future consumer purchase behavior of households across the U.S. at every retail channel.
This level of precision and granularity allows CPG marketers to segment audiences, optimize in the bid stream, and measure success. If done right, they’ve a roadmap for future engagements.
A path forward
As advertising becomes more data-driven, marketers will feel pressured to make large investments in consumer data. By prioritizing refined insights over raw data, they can fuel more powerful advertising strategies.