💫The end-to-end approach to Decision Intelligence is a key to success.
If you are a data and analytics leader #CDAO #CDO #CAO, do you remember the last mile of analytics problems? Let me explain…
The last mile of analytics was about the belief that if data and analytics teams build recommendations, insights, dashboards etc. it will change the organization's behavior and create better business outcomes.
Guess what? IT DID N🚫T!
There are two core issues 👇holding their solutions back:
1. They were not based on explicitly defined business decisions – therefore, many often team solved the wrong problem (type 3 error)
2. Sometimes, the team got lucky to win a lottery and solve the right problem, BUT it was very isolated and not integrated into systems and processes and the day-to-day workflows.
💡#DecisionIntelligence is an evolution of decision management in conjunction with other technologies. In this sense, if leaders make the same mistake inherited from data and analytics that caused the “last mile of analytics,” they will still face the problem of the “last mile of analytics” no matter what.
How to avoid this?
🔹 Model business decisions explicitly and manage them systematically. Modeling business decision is not a rocket science, but it is the most critical part of the successful approach. Do yourself a favor and leverage the open standard that is widely accepted and used in many industries called “Decision Model and Notation – DMN” for this part and not a proprietary notation.
🔹 Make sure you cover the full decision cycle, meaning your decision models should be able to interact with outside i.e. the environment they are put in. They should be able to access data from operational databases, files, online services, 3rd parties, etc. They should be able to massage data for on-the-fly for processing and pass it to the decision model for execution.
🔹 The solution must be able to reflect back the decision outcomes (e.g., results, insights, recommendations, etc.) into organizations' systems, processes, and workflows. Leaving the decision outcomes in external systems or dashboards isolated and disconnected from actual business operations will not deliver any change.
💣That's what the end-to-end is: a fully integrated approach involving data, process, decision execution, and integrating results back to organizations. The caveat is that everything MUST be around and for explicitly defined business decisions not #data #machinealearning #optimization #workflow #llms and etc.
Posted here.
Published August 14th, 2024 at 07:30 am
