Research on Climate Risk Management in the Drug Store Industry

Introduction

The Japan Meteorological Agency (JMA) is conducting a project to promote climate risk management (CRM) in various sectors. CRM involves reducing climate risk caused by heat waves, cold spells and other extreme events using extended-range weather forecast data such as Early Warning Information on Extreme Weather (two-week forecasts) and One-month Forecasts.
In the research reported here, which is part of this project, JMA investigated the relationship between drug store sales and temperature along with other weather factors. The use of two-week forecasts for added benefit in sales promotion was also investigated.

The Japan Association of Chain Drug Stores (JACDS) and related companies kindly contributed to the study by providing several years' worth of drug store sales data. Factors in the relationship between weather and drug store sales have frequently been discussed, and stakeholders have developed ways of modifying their sales management using two-week forecasts.
A significant correlation was found between sales volumes and weather factors. It was concluded that two-week forecasts can help sales managers plan matters such as which items should be prioritized and how to manage the supply of products.

Results

There is a significant correlation between sales and temperature.

There are threshold temperatures at which sales increase remarkably (Table 1, Figure 1).
Sales tend to fluctuate in phase with weekly average temperature variations.

Table 1 Threshold temperatures for various products (Tokyo)

Product	Threshold temperature	Approx. potential sales increase
Sunscreen	Above 10℃	×4.7 at 5℃ above threshold temperature
Cockroach insecticide	Above 11℃	×2.7 at 5℃ above threshold temperature
Fly / mosquito insecticide	Above 18℃	×3.2 at 5℃ above threshold temperature
Insect sting treatments	Above 18℃	×2.6 at 5℃ above threshold temperature
Sports drinks	Above 25℃	×1.6 at 5℃ above threshold temperature
Cold medicine	Below 25℃	×1.5 at 5℃ below threshold temperature
Hand cream	Below 25℃	×2.9 at 5℃ below threshold temperature
Lip balm	Below 25℃	×1.6 at 5℃ below threshold temperature

Figure 1 Time-series representation of oral rehydration solution sales,
average temperatures and heat illness cases for Tokyo

Oral rehydration solution is a typical summer product.
Sales start to increase in June as temperatures rise.
Sales increase sharply when the temperature reaches 23℃.
The threshold temperature for oral rehydration solution sales is 23℃.

Analysis by drug store experts:

The outcomes of the research indicate that the start of temperature rise and the peak vary significantly by year, and that sales of certain products exhibit a correlation with temperature.
Certain types of products appear to exhibit a correlation with temperature, such as those emphasized as initial measures in light of usage relative to the product volume and those whose sales growth is continuous at or above certain temperatures. Full awareness of such patterns for individual items therefore appears important.
The results of this research scientifically clarify the relationship between temperature and product sales. Based on product lifecycles, the outcomes could potentially be utilized for setup in the introductory phase and in particular for inventory control in the decline phase.

Adaptation to Climate Risk

Stakeholders can use forecasts to support the implementation of climate risk countermeasures. For example, if a two-week forecast indicates a high possibility of temperatures exceeding 25℃(as seen in Figure 2), drug store managers should frequently replenish stocks of sports drinks to ensure sufficient supply. Floor managers may also wish to prioritize the display of sports drinks in the selling area.

Based on the results obtained and on the outcomes of discussions with drug store experts, possible sales promotion plans include:

Modification of supply volume from warehouses to stores
Strengthening of visual merchandising (VMD) methods such as point-of-purchase (POP) placement
Development of sales talk for customers
Arrangement of sales floor plans
Efforts by store pharmacists to raise awareness of the climate's impact on people

Figure 2 JMA two-week forecast
Initial date/location: 28th June 2013, Tokyo

Cumulative probability: forecast (blue), normal (black)
Probability density: forecast (green), normal (gray)

Expert Feedback and Future Strategy

As average temperature data are hard to understand intuitively in many industries, forecasts of maximum or minimum temperatures would be better.
More accurate forecasting over a period of three weeks or more would be better because it takes drug stores at least this long to print promotional materials.
Guidelines on the use of JMA's probabilistic forecasts should be provided.