CRM Analytics Predictions Survey
This content is part of Tableau Blueprint—a maturity framework allowing you to zoom in and improve how your organization uses data to drive impact. To begin your journey, take our assessment(Link opens in a new window).
Before creating stories/models and deploying predictions using Einstein Discovery, each Line-of-Business Sponsor should survey their departments and teams to help them prioritize use cases and prediction needs. Each business team who will be using Einstein Discovery should complete the Predictions Survey or work with team members to facilitate documentation of the information. The purpose of the survey is to identify the business use cases that require optimization (descriptive and prescriptive predictions) and the sources of data that are needed. Additionally, this survey will help you determine if data scientist need to be involved, plan and execute prediction deployments, and assign model monitoring responsibilities.
- What is the team's business function?
- Who are the targeted users?
- Do you need to collaborate with an existing data scientist team?
- Does your team have access to the data that needs to be analyzed?
- Who on the team will need full license access to CRM Analytics (to bring the data in, create stories/models, and deploy predictions)?
Selection and Management
- What business use cases will be used for predictions? Have the use cases been been qualified as good or bad fit?
- What and where are the key sources of data for stories/models?
- How does your team source data (Salesforce, databases or warehouses, file exports, third-party, etc.?
- Do you have or require approval for models from a data scientist team?
- Have you incorporated data modeling methods such as CRISP-DM to qualify the business use cases?
- How does the data look for phase 1 (quick and dirty approach) and can you use csv files to populate the dataset?
- How does the data look for production deployment and are you able to leverage data prep and build the orchestration needed?
- How is data secured?
- What business scientist skills and capabilities exist within the team?
- Who will be identified and trained as the Einstein Discovery Champion (e.g., business scientist) within the team?
- Do you have the necessary business skills and insights to prioritize use cases?
- Do you have the necessary data engineering skills for any derived fields and feature engineering needs?
- Have you already vetted and approved Salesforce cloud products and data centers with regards to using CRM Analytics?
- How are you going to monitor the accuracy of predictions (e.g., Model Manager, custom dashboard)?
- What is the accepted model accuracy threshold and who is going to be notified and alerted if the value is out of bounds?
- How often will you need to refresh model data?
- Who will be responsible for measuring business impact, performance, and adoption rates of the model once it is deployed?
- How will the business monitor and measure the impact on processes and results from deployed predictions?