The CEO of a media company recently asked me about my thoughts regarding the near-term and long-term future of the economy due to the COVID-19 crisis. The conversation turned to what type of recovery would we see and how long will it take. What do executives need to know and how can they prepare for the rebound. Here are a few thoughts from a data science perspective.

We aren’t out of the woods yet.

At 5EA, we are modeling the virus spread and population effects as best as possible with the data we have available. We are providing insights to our customers about the virus growth curve, with a focus on NY and projecting when the peak will come. The virus as Dr. Fauci, Director of the National Institute of Allergy and Infectious Diseases, states, “has its own timetable”, and therefore the public’s participation of government recommendations and guidelines is critical to slowing the spread and ultimately the start of the recovery.

Starting the Engine

When we reach the peak of the virus, which will occur in different parts across the country, the economy will be like car accelerating. We should see regional recoveries which would then lead to a more national and then global recovery. The Great Recession of 2008, saw negative GDP from the beginning of the recession in Sept 2008 until the 3rd Quarter of 2009. A key difference between 2008 Recession and our current economy, is it’s strong structural foundation. No other recession has ever been characterized by a voluntary shutdown of an economy. Therefore, according to our preliminary research, we feel that a recovery should be quick and the acceleration will be rapid, targeting somewhere around 4-6 months after parts of the economy starts between May 1st and May 25th.

Be Smart, Use Your Data

Your company’s recovery will be dependent on actions you take now. Start by examining where you would have been had the shutdown. Leverage your data prior and post shutdown to examine changes in your customers, or their behavior. Using this data you can begin to make projections about how your economic conditions will change when the recovery starts and is complete.

The more data you can collect and analyze the better you will be prepared for the recovery. We at 5EA strongly believe that companies that can leverage data science will be in a far better position to explore new opportunities and manage the intermittent shocks that will occur in the coming months.

What are some specific actions you can take

a) Analyze the change in your customers – you may have seen an influx of customers based on the business you are in. If this is the case consider yourself lucky, however, these customers may not stay with you. They may have used your service due to convenience and they may exit when this is over. Conversely, if you have lost customers, many of those customers will come back, but you must stay in touch with them. Now is a good opportunity for brand and product awareness projects.

b) Model your revenue – Many companies, will see a strong dip in revenue. Begin using the growth curves of the virus as a factor in your equations. The virus itself doesn’t have a direct economic effect, but its an indicator of government responses. By modeling the virus in different areas of the country or world, you can get a better idea of how your revenue curves will react to government policy

c) Target – Allow data science models and classifications to help target areas and identify products that will be in demand. Your previous models, if you have them, will assist greatly in this effort. Connect these previous models with new short term models from the above recommendations and you could generate some very specific insights.

d) Research – With the downturn, now might be the time to revisit some data projects you haven’t had the time to do. Every company has a significant amount of data and as such should leverage the time wisely to prepare for the rebound.

In the coming days, we will be providing more information about virus growth, and economic projections. If you have some specific questions or are in need of assistance and would like to talk to one of our data scientists reach out to us at info@5eanalytics.com and we would be happy to assist.

Let’s keep up the good work! As we move through this crisis together, we are affecting the curve.  All of our efforts from increased testing to social distancing and business shutdowns, each designed to slow the growth of the virus and prevent more people from being infected have begun to take effect. 

5EA has been downloading and examining the number of confirmed cases each day. Our initial results are showing signs that we are making very good progress toward the first goal which is to slow the growth of the curve.  We predict that by Tuesday 3/25 the impact on the growth rate should be significant and consistent enough to see a consistent downward trend in the growth rate, assuming no new clusters appear and government actions and compliance remain in place.    

The growth rate of confirmed cases in NY spiked on Thursday with a rate of approximately 1.87, and steadily moved down to 1.37 Saturday. The numbers reported Sunday at 4:15PM from NY showed the growth rate at 1.29, a significant downward reduction in new cases, a likely indication that the efforts are having an effect in NY.

Overall for the United States, we saw the number in the same period move to 1.37 as of Saturday; however, with the NY numbers moving downward, the state with the largest number of cases, we expect the downward trend to continue. Especially as other areas of the country practice social distancing, and lockdown businesses. Localities may still see an increase since compliance with local mandates is critical, and these flare-ups must be monitored closely.  

It’s too early to tell when we will hit the peak, but examining the growth of the virus, we believe the peak will occur around or before April 20th. This downward trend in growth should continue as social distancing and reduction of interpersonal contact measures are enforced, although there is no way to definitively know. A number of factors could adversely affect these numbers including new outbreak clusters somewhere in the country, communities not following local social distancing mandates, or a significant spike in testing or reporting of testing. 

We want everyone to know that to continue this decline, everyone must work to flatten the curve. Let’s continue to watch and monitor what’s happening and do our part.

All of our data was based on publicly available information provided by the NYS department of health, CDC and Johns Hopkins.

Revenue Forecasting presents a major challenge for companies, especially in finance and marketing. Erroneous forecasts can be disastrous for a companies bottom line, especially when money is spent in anticipation of higher revenues. While the challenges are numerous, there are some actions a department can take to ensure their forecasts are accurate and continue to be accurate.

1. Bad Data

This won’t come as a surprise to many. Every company has bad data, but does this mean you can’t get an accurate forecast, no. First, the data should be analyzed for the causes of errors. Once the causes are known, the errors can either be removed or accounted for, in a model, greatly enhancing the accuracy.

Not all your data, therefore, is bad. The trick is to identify the data that is good, which will act as a core baseline. This baseline can then be used to identify variations and fluctuations to determine the cause. Factors, such as spikes in class center volume, vacations by salespeople, or temperature all can impact results.

Our data scientists, continually find ways to mitigate the issues of bad data through statistical imputation and simulations. These methods can greatly enhance an understanding of the effect of the bad data, as well as provide methods to remediate the issues.

2. Changing Environment

“Change is the only constant in life”, was a famous quote by Heraclitus. Many revenue forecasting models preach “teach it and leave it”, as though it can continue to provide accurate forecasting without ever modifying it. However, this isn’t really the case. Economic environments change, organizations change strategies, and customers change preferences. In some cases, models can pick up the changes, but more complex models require retraining of the model. While many companies say this can occur automatically, it cannot.

Take for example a scenario in which a variable was never included in the original model. If this variable becomes important at some point in the future, the model could never compute its impact, yielding an incorrect result, such as would occur with a new product category or removal of an entire product category.

We think of this as driving a car, always making adjustments. While generally a smooth ride, there are times in which scenarios appear, where traditional training doesn’t work and we need to make a split-second decision. Algorithms cant generally do this, without massive computing power, such as what is needed for self-automated driving.

Change is the only constant in life – Heraclitus

Revenue forecast models need regular and periodic review for accuracy and training.

3. Bad Interpretation

Ultimately, the models produce a number for the forecast, models that are too general have increased errors overall, while specific models have fewer errors for training sets, and increased errors for data it hasn’t seen. Analysts’ interpretations are critical when this occurs since the end-user of the forecasting model isn’t fully aware of what events are causing fluctuations.

One way forecasters combat this is to create an ensemble model, which aggregates a number of models together. The challenge with this approach is that many there may not be an insight into which part of the ensemble is more weighted, thus the ensemble simply tries to smooth out these fluctuations, causing a loss of, most likely valuable information.

Creating multiple models allows for a more granular interpretation of the forecast as well as greater insight into the error. This examination of error is critical to understanding the events that are impacting the revenue. Simultaneously, it creates a better-disciplined process by which analysts, forecasters, and end-users are diligent in understanding as much of the variation as possible.

Creating forecasting models in a vacuum should be a serious concern for the organization. This black-box approach, without truly understanding the effect of multiple data points and methods can lead to a forecast accurate in some instances, but which can lead to failures down the road. We believe these black-box approaches create more problems, and that a solid repeatable reviewable process is the most reliable approach and yields more opportunities.