Collaboration: A New Definition of Analytics
A new approach to analytics is starting to emerge – we call it Collaborative Analytics.
We appreciated the feedback and the questions that asked for a deeper definition of the process underlying the idea.
We define the phases of the process in the following way:
- Integrate: Align disparate data sets into one common body of knowledge using some form of data science. The key is to align insights that are from various qualitative and quantitative sources.
- Validate: Move from historical measures of fit such as R-squared to broader measures of model specification and perform holdouts that not only recreate the past, but reliably forecast to guide decisions that go into the market.
- Evaluate: Conduct analyses that identify how the entirety of the marketing effort performs when any one component changes. Such analyses test all parts of the brand ecosystem simultaneously and interdependently.
- Explore: Perform “what-if” scenario analysis that not only optimizes strategies, but also de-risks them. The “what-if” process helps eliminate ideas that are deemed too risky and then guides the optimization of winning ideas.
- Activate: Deliver signals to various “agents” in the network – programmatic platforms, agencies, sales teams, media buyers, strategists, creative teams, and customers in order to ensure that programs come to life in the network.
Ultimately, collaborative analytics brings together people, processes, data, and tools into one cohesive system that is enabled by technology and by an open, adaptive, and knowledgeable culture.
Rewards and resources flow freely to the highest value-creating activity.
Power and influence are no longer held by those who have asymmetric insights they use to their advantage, but to those who help a brand outpace its competitors through speed and knowledge.