Evaluating and prioritizing business processes

3:00 min

Medium

In any analytical data modeling project, it's crucial to evaluate and prioritize business processes effectively. Focusing on the most impactful and feasible processes first ensures that your data models can deliver the highest value to the business as soon as possible, while minimizing risks. It allows for a deep focus on the necessary details, and helps accumulate valuable knowledge for future projects.

Impact vs. Risk: The Core of Prioritization

To prioritize effectively, you need to consider two main factors: impact and risk.

Impact: This refers to the potential value a business process can bring to the organization. Does the process help the business make or save money? Does it improve efficiency, enhance customer satisfaction, or contribute significantly to key performance indicators (KPIs)? Processes with high impact are often those that the business already knows are critical; if they aren't yet available to the customers, they likely have already been asking for insights in these areas.
Risk: This refers to the challenges associated with implementing a data model for a particular process. Key considerations include the availability and quality of data, the complexity of the required business logic, and any technical hurdles that might be encountered. Assessing risk typically involves discussions with data source experts and conducting data profiling to evaluate feasibility.

Start with Low-Hanging Fruit

A pragmatic approach is to begin with "low-hanging fruit" - processes that offer high impact with low risk. These are the most critical and feasible projects, providing easy wins that can immediately demonstrate the value of your data modeling efforts. Starting with these processes allows for quick successes and helps build momentum for tackling more complex challenges down the line.

Our example company "Stellar Sales" has used this approach to prioritize the following business processes:

Sales and Revenue Tracking: Monitoring the performance of different product categories across planets was identified as a high-impact, low-risk process. It directly influences revenue generation and helps customers optimize their sales strategies. Data for this process is readily available in their data warehouse and already relatively clean, making it an ideal starting point.
Customer Behavior Analysis: Understanding where users are buying from and what influences their purchasing decisions was another high-impact process. By analyzing customer behavior, Stellar Sales can help customers tailor their offerings to specific markets. As this process is slightly more complex, and some of the data isn't captured or natively available, this process brings along some risk and potentially needs to be further prepared or scoped.
Order Management: Analyzing typical order sizes and fulfillment efficiency impacts both customer satisfaction and operational efficiency. The source data required is well-documented and consistently available, but it's currently partially spread across different data sources (e.g. third-party APIs of shipping carriers); Stellar Sales will therefore first need to plan some further development to consolidate the required data in their data warehouse.

As a quick win, Stellar Sales decides to first focus on perparing for insights around sales and revenue, as those data points provide the most value while being feasible to achieve!

Evaluating and prioritizing business processes is a critical step in any data modeling journey. By focusing on high-impact, low-risk processes first, you can ensure that your initial models deliver substantial value to the business while minimizing the likelihood of setbacks. In the next article, we'll provide some useful tips and tricks that should facilitate the design process of your (first) analytical data models!