Data seems to be everywhere. Big Data is taking us to new frontiers in analysis and predictability. With today's more sophisticated analysis tools, even data in less traditional formats can be a valuable business resource. But, the data can be so much more valuable if it is used in ways that truly reflect an understanding of the dynamics of the business. This makes the data far more meaningful.

Some proponents of using Big Data advocate merely identifying correlations, even if the reasons for those correlations are not fully understood. And, at times that approach may be useful. But, Big Data should not be applied in a vacuum. It is important to think through what the data means, and to understand what it does and does not reveal about the business. So much more value can come from bringing in an understanding of the business that may not be reflected in the data. That's why an understanding of the dynamics of the business should be an added component when using Big Data, especially early on, when little may be known about how it might work or how well it will work.

Understanding the business is especially vital upfront, such as when defining how the analysis should be approached or when predictive models are being developed. Afterwards, an understanding of the business can also aid in interpreting the results of Big Data analytics. It can be a guide for explaining findings, for adjusting them as might be appropriate, and for fine tuning recommendations to better reflect what is known about the realities of the business. And, on an ongoing basis, care must be taken to be sure that what is done with the data continues to reflect the dynamics of the business when conditions change or when something unusual occurs.

There are various ways to do this. In some cases, issues that reflect business dynamics can be incorporated into analysis or predictive models by creating and including additional variables that represent what is known about the business. In other situations, perhaps commentary tied to business issues can be added later, when results of the analysis are reported. Regardless of how it is accomplished, however, bringing in an understanding of the dynamics of the business can greatly enhance the value of Big Data.

An example that offers a taste of what this might entail is from a recent Wall Street Journal article about data obtained from images of places such as parking lots. Before moving on, however, I'd like to point out that my intent here is not necessarily to provide the specifics of what works well when analyzing parking lots. Nor is my intent to explain what may or may not be included in anything offered by suppliers of services or data based on images of parking lots. As someone who spent my early career in quantitative analysis, then later developed a broader business background, my intent here is merely to provide more general insights that might guide overall approaches for effective use of data.

That said, the Wall Street Journal article is "Counting Cars to Predict Earnings" by Bradley Hope, which appeared November 20, 2014. It describes how Big Data offers opportunities to use images of parking lots to predict how well companies with retail locations are performing, although the article does say that the predictions are based upon other factors besides the number of cars in the lots. The article tells about a business that obtains and analyzes this parking lot data, and sells it to the investment industry, which is interested in predicting the revenue and the profitability of companies. Although the article points out that parking lot data can be of value to investors interested in predicting the profitability of a business, I would see parking lot data as also useful to companies for other purposes, such as competitive analysis or evaluating potential new locations.

Nonetheless, since how full the parking lot is may be an indicator of how well a business is doing, parking lot images provide valuable information. As valuable as parking lot images can be, however, augmenting them with an understanding of the business can create even greater value. And, this is important because the implications of how full the parking lot is may not always be clear cut.

For example, parking lots may be packed during a struggling company's going out of business sale, when images of those crowded locations may not indicate profitability at all. The reverse is also true since some companies may decide to focus upon their most profitable customers. These companies may increase their profitability, while the number of customers they serve, and even their revenue, might decline. Their parking lots might be a bit emptier while their profitability charges upward. Thus, although full parking lots might generally indicate profitability, that may not always be the case.

The key point here is that when using data - whether it is parking lot images or whatever - thinking through what the data means and how different meanings might impact the data's usefulness is essential. The nuances in meaning might matter more in some situations than in others. And, in many cases, human input and judgment will be required to identify business issues that should be incorporated into the data analytics. In some cases, the business dynamics to consider may seem closely tied to the data itself, while other times broader strategic issues will be involved. The business dynamics may not always be apparent to data scientists that primarily have strengths statistics, algorithms, computer coding, and related technical areas rather than a strong background in broader business issues.

In conclusion, newer analyses made possible by Big Data can be valuable. But, it is important to think through what they mean. By doing so, analysis methods and reporting of findings can be fine tuned to better reflect the dynamics of the business.