Proper Forecasting Can Be Life or Death
The importance of proper forecasting and results measuring in business cannot be understated. Countless times I have witnessed businesses and sales teams over-forecast sales or under-forecast losses due to their inability to accurately interpret data. This can be a million-dollar mistake and in some cases, a life or death situation. Let me explain why.
There is great debate over the reported rise in COVID cases in the U.S. today. On one side the message is that COVID cases are increasing every day at a rapid pace. On the other side the message is that the number of cases haven't necessarily grown, the number reported have grown, due to the increase in testing. Both are right and wrong. Both are right in the sense that they are basing their messaging on the facts of the data. Both are wrong because of the framing that they are using in their messaging, which focuses only on the part of the data that supports their message.
When referencing data, especially in business, it is important to provide the results in multiple formats while making sure that your data sources remain consistent. The results should be provided in numbers as well as percentages to verify the validity and clarify the messaging. For example, if the number of tests per day is 10,000 and the number of positive tests is 1,000, that means that 10% or roughly 1 in every 10 tested have shown to have COVID. If the tests jump to 20,000 per day and the number of positive cases jump to 2,000, then yes, you can say that cases increased 50%; however, according to the data, the number of positive tests remain at 10%, which would prove, in this hypothetical example, that COVID cases were stabilizing. Choosing to report the 50% growth absent of the percentage provides a very different narrative. Reporting that COVID is stabilizing without the added information of the additional 10,000 cases is also a fail in transparency as it provides a false narrative that the amount of people with COVID among us has not grown.
So what is the issue? Aside from the slant on the narrative, when I researched the Centers for Disease Control & Prevention's published data, I found that the variables are inconsistent, making the data unreliable. The CDC provides conflicting data on the number of cases vs the number of tests vs the number of deaths reported. The date ranges are inconsistent, the results are lumped in on some days and not on others (lack of consistent reporting of data), and the amount of testing vs the population is inconsistent across states, therefore the data is unreliable for forecasting. Unless we were testing everyone in the country, at the same rate using the same form of measurement, which we are not, it can be virtually impossible to have an accurate reading of the status of COVID based on the data samples. While this is a real example, COVID is not the focus of this blog.
In business, inaccurate data pools can lead you to bankruptcy if you are measuring inconsistently or from a lens you believe to be correct. For example, if you were to fast forward to 2021 and were a grocery store measuring sales year-over-year, you would see a spike in sales during the COVID lock down in 2020 (April-June) vs where you were in 2021. If you set your sales forecast based on a year-over-year measurement, you would likely be setting yourself up for failure as the pandemic caused a spike in grocery stockpiling. To prevent this, always research any anomalies in your data (good or bad) to seek to understand the contributing factors for the variance. This will also allow you to review how your team responded to the variance to see if best practices can be developed or created. If you are uncomfortable with forecasting or analytics, get familiar with the unfamiliar by taking online courses in forecasting, partner with a coach for development or utilize software that assists you in interpreting your data. While it may not be a life and death situation like COVID cases are, it could be a life or death situation for your business.
Biz Lesson: Always seek the “Why Behind the What” when reviewing data and measuring your results.