Before starting an analytics or AI project, analytics leaders, business sponsors, and data-product managers should agree on the business decision they want to improve and how someone will act on the results. They should confirm that the available data and analytical approach support that goal, decide how to deliver insights to the intended users, and define a measurable outcome. This gives the team a clear purpose before it chooses tools or begins development.

At FIMA 2026, Juan Liu, Director of Analytics at VanEck, emphasized the importance of defining the decision before choosing the technology: "Start with the decision and everything else follows." In an email marketing example, Liu's team narrowed a broad request to use AI into a specific choice: which subject line tone a campaign manager should select. That detail helped the team identify the relevant data, deliver recommendations to users, and measure results against a baseline.

Another FIMA 2026 conversation showed why collecting data is not enough. In one use case, a financial services prospecting initiative needed to connect external information, CRM activity, and web engagement to produce useful recommendations. Salespeople also needed enough context to act on those recommendations within their existing workflow.

Project leaders should record the five answers in a short intake document and review them with the business sponsor. Name the decision maker, describe the action they will take, and check whether the available data supports that action. Agree on how the team will deliver the insight and measure the result before development begins. If an answer remains vague, resolve it with the business rather than asking developers to make the assumption.

Join financial services data leaders at FIMA 2027 and learn how peers define analytics and AI projects that support better business decisions.