Producing reports has, from time to time, been the bane of my existence, almost as bad as consuming reports produced for me by others. I am often reminded of a design heuristic I learned many years ago: “No one knows what they want until you give them what they ask for.” It has always seemed that no matter what we put in a report, and no matter how well the content meets the needs of a specific audience, other audiences will want something different. The tweaks may be large or small, but they are time consuming. Every tweak risks introducing errors or distortions into the interpretation of the underlying data. Essentially, reports are out of date as soon as they are produced.
This style of retrospective reporting might be OK if we only want to figure out how we did last quarter. However, reporting in this way negates the huge improvement in information accuracy and relevance that came with the implementation of enterprise-resource-planning systems over the past two decades. Amid all the pain of ERP implementation, it’s easy to lose sight of the shift from averages to actuals that ERP systems made possible.
For the first time, any business could look at how it was operating on a very granular, transaction-by-transaction, customer-by-customer, product-by-product basis and dig into variances to understand what was going right and what wasn’t. This new ability was largely the trigger for a host of process-improvement techniques, from just-in-time to six sigma and lean, and it’s the foundation to a lot of what we call business intelligence (BI). However, it hasn’t made much of an impact on reporting. We are all too often still looking backwards at summarized information that isn’t quite what we want, or need, to know.
So what should we be doing? We need reliable, accurate and timely information to run the business efficiently and to meet regulatory reporting requirements. It’s also a given that we can’t look at everything all the time, although someone should be paying attention to each thing that matters. Some summarization is inevitable but so is the need to get to the details when we need to. How do we get to the right balance?
To fix reporting mishaps and misery, try these steps:
1. Ask, “What do we really need to know?” In most cases, a simple red, amber or green (RAG) indication of good health is sufficient, provided that I trust the information sources and the summarization scheme that determines the indicator color. I need to periodically audit some of the greens as well as the ambers and reds. After all, one person’s doing just fine may be another’s impending disaster. Trust but verify doesn’t go away.
2. Replace static reports with dynamic dashboards. Data that is accurate to the minute, hour, day or week — dependent on the characteristics of the underlying processes and our ability to respond to changes — is usually going to be better than a once-a-month snapshot.
3. Connect dashboards to the underlying data sources so that you can drill down to look at the details of the indicators. I’m assuming you have a way to load into your systems the plans and budgets that define your expectations, and that you can measure actual performance against these expectations in a controlled fashion. This may require improving the level of detail in the plan, as well as integrating different sources of data and tools, but it’s essential if the dashboards and indicators are to be useful. If attention isn’t paid to how the drill-down process works, too much digging in the data can have a negative impact on the performance of the transaction systems that are generating the data. There are plenty of ways to avoid this issue, but I’ve seen many situations where it was missed with ugly results.
4. Consider some degree of self-service capability. With the right tools and a little help (training and support) most executives can get what they want from the reporting platform without resorting to specialist analysts and information-technology resources. Turning the CFO into a BI developer isn’t a good use of skills and time, but the available tools are quite adequate for an on-demand, self-service approach once the underlying platform has been deployed and proven. And doing it yourself, at least some of the time, is a good first step to judging just how badly you want the answers.
John Parkinson is an affiliate partner at Waterstone Management Group in Chicago. He has been a global business and technology executive and a strategist for more than 35 years.