Data science is a vital tool in the response to the human challenges and uncertainties posed by the COVID-19 situation. Data holds the key to understanding disrupted supply chains, employee health patterns, changing consumer behavior and emerging credit risks.
Although advanced and predictive data analytics offers a path out of COVID-19 confusion, the volume and variety of new data being generated – and the speed at which this data is being shared – risk overwhelming companies if they don’t have a smart data strategy in place.
Over the past few months, KPMG’s global data and AI analytics platforms (known as KPMG Signals Repository) has been helping clients navigate the fast-moving currents of the COVID-19 situation. As companies respond and prepare for recovery, here are five data trends that should inform business planning, decision making and risk mitigation at this time:
1: Harnessing pop-up data sources
The data signals KPMG has traditionally used fall into three categories:
- static (e.g. the geographical distance between a home and a hospital)
- slow moving (e.g. the ratio of health care professionals to people in an area), and
- fast moving (e.g. first-time unemployment filings).
Absent any specific information about COVID-19 cases, these signals are proving useful in highlighting locales that had or have a predisposition for a broader and longer sustained impact and, by extension, more lasting macroeconomic effects. An example might be the reliance of people in the area on mass transit or the density of people in the area with past diagnoses of diabetes or upper respiratory diseases.
In today’s rapidly changing environment, though, we see a fourth category becoming increasingly important: pop-up data. In our example, this might be the latest number of confirmed COVID-19 cases reported in a county or Google Mobility data (which highlights where mobile phones are congregating).
Combined with traditional signals, pop-up signals are enabling us to glean insights into the likelihood, timing, and size of hyper-local economic rebounds. Harnessing pop-up data requires agility but equips leaders with better insights for in-time decision making.
2: Minimizing economic risk through data
Much of the financial services sector is currently in a holding pattern. Some mortgage repayments have been paused and interest on overdrafts and credit card payments frozen to help consumers whose incomes have been hit.
These changes, however, are just short-term fixes. Lenders will soon need to identify those consumers at greatest risk of mortgage, loan and credit default so they can plan proactively by offering support and advice while protecting their business. Many sectors of the insurance industry similarly are anticipating an influx of new claims in the coming months or lapses in premium payments.
Using data analytics with an appropriate big data fabric, these industries have the opportunity explore and ameliorate risk at a hyper-local level by leveraging all 4 varieties of signals in order to help business leaders predict which enclaves have the greatest risk of defaulting on mortgages or submitting an insurance claim.
3: Using signals to anticipate future demand patterns
Every city and town throughout the world has been affected differently by the virus and, as such, it’s reasonable to expect that each will recover differently. This places major logistical strains on organizations that have previously forecasted demand-based consumer behavior patterns.
Think about a business operating across multiple locations that has been forced to furlough staff. Its recovery relies on operationalizing the right roles in the right locations at the right time to address new customer demand patterns.
Assembling and cross-analyzing data signals such as the daily update of new jobs posted by the National Labor Exchange or the uptick in ride-sharing activity in one area are just two of the thousands of signals that can indicate the pace at which localized economies are bouncing back.
4: Winning and retaining new types of consumers
Historically retail consumers have been predictable in their habits. They have their favorite stores, preferred shopping times and ways to shop.
COVID-19 has complicated those predictable patterns as consumers have had to adapt to a lockdown lifestyle. This means retailers will also need to adapt to new consumer trends. Intelligent forecasting – augmenting traditional inventory analysis with hyper-local and geospatial fast-moving data – puts retailers in a better position to stem the loss of loyal consumers, while also retaining new customers attracted during the lockdown.
Take ’click and collect‘ for example. As more US consumers shift to online shopping retailers can use purchasing patterns, pick up times and geolocation data not just to improve click and collect infrastructure, but also to improve in-store experience. They can plan in-store product picking and inventory replacement to avoid disadvantaging consumers shopping in-store.
The lessons learned from these consumer shifts will help organizations quickly prepare for the era of true contactless commerce that is fast approaching.