1. Choose the right trends. It might sound obvious, but companies often don’t monitor the most relevant data. A basic materials company selling to the automotive sector, for example, drove its sales projections by the number of cars sold by region, which seems logical. However, its orders were actually tied to the number of new automotive projects (e.g. prototypes, factory modifications). The company simply got the correlation wrong. One way to figure it out is to “analyze your analysis.” Trend analyses are often based on false or old assumptions that haven’t kept pace with changes in the market place. Each data set you’re examining should tie directly to your sales goals. If you can’t make that direct connection, you’re looking in the wrong places. One critical way to identify the trends that matter to you is to ask yourself, how could this affect my customers? It’s a basic question but one that’s often overlooked when looking at long-term trends.
2. Keep it manageable. Even when you’ve narrowed down the trends to the ones that really matter to you, it’s still easy to drown in complex models that become both time consuming to run and hard to make sense of. You need information to help make decisions — investments, resource allocations, hiring — not awe-inducing spreadsheet models to impress particle physicists.
3. Tie insights to operations. Too many companies fail to take effective action or manage performance based on the insights they uncover. If you’re still rewarding the frontline based on short-term results, then any long-term effort will fail.
Harvard Business Review