The Retailer Spring_09.05_FA

#reinventionretail

Are you geared up to manage your weather sensitive stock this spring? Perhaps not as much as you could be…

RECENT RESEARCH BY BRC AND THE MET OFFICE HAS SHOWN A CLEAR LINK BETWEEN WEATHER AND NON- FOOD SALES, WITH SOME RETAIL CATEGORIES BEING MORE SENSITIVE THAN OTHERS TO WEATHER IN SPRING. VARIOUS POSSIBLE LINKS BETWEEN SALES AND WEATHER DATA WERE INVESTIGATED, WITH WEATHER ACCOUNTING FOR THE STRONGEST RELATIONSHIP IN LIKE FOR LIKE SALES VARIANCES Research highlights The relationship to temperature was strongest in women’s clothing, where a year-on-year 1oC change in temperature for weeks in spring, creates a 2% change in year-on-year sales. In autumn the profile for winter women’s clothing sales is even stronger. A summary of the key relationships for all product lines is in Table 1.

So how can this information be exploited? First of all, it is possible to assess the likelihood of a particular week being warmer or colder than the same week last year. For example, if week 17 was very warm last spring, the odds are against it being even warmer this spring, so the expectation will be that like-for-like sales will be reduced in week 17 this year. By using historical weather records and the year-on-year sales relationship, an anticipated sales profile for the spring period can be created. This type of analysis produces quantitative and plausible sales profiles for the current season by placing last year into its historical context, helping you to maximise on sales opportunities, but more can be done. Seasonal forecasts give additional information that enable certain outcomes to be weighted more highly than others. For instance, spring last year may have been very warm and conducive to high sales, so statistically we would expect sales this year to be reduced. However, if the seasonal forecast is predicting an increased likelihood of above average temperatures, there will be an increased prospect of this spring being as warm, or warmer than last year. This means the probabilities shift in favour of another good year for sales. The Met Office are experts in applying weather information to retail operations and will carry out all these calculations behind the scenes, so that you receive information presented in a manner that you and your systems are able to consume and easily interpret. This could be a data feed of all possible outcomes, a probability distribution or just the most likely outcome with a high and low probability range for each week. This information would be updated weekly, starting three months ahead of the week in question. Looking further ahead When looking years ahead, the sales profile over the spring period will change when using weather intelligence, with total seasonal like-for-like sales remaining approximately the same, year-on-year. Met Office services will help you to get that profiling right and improve the prospects of successful promotions. Developing the weather relationships The weather relationships have been developed using UK-wide sales as a whole. Embedded within these data will be stronger relationships for some product lines and weaker relationships for others. The variability in weather across the UK will also act to degrade the relationships. Therefore, if you possess several years of more localised, product-specific data there is every prospect of creating more concise relationships than the general relationships developed here, giving extra confidence in the guidance that is created. Using this understanding helps make the awkward statistics more palatable and for there to be more measured euphoria from the rosy results. What superficially looks like a good or bad set of results may actually be the opposite when weather is accounted for.

Effect on weekly sales growth for each degree its warmer than last year -1.1% (mid-Aug to mid-September) -2.7% (mid-Aug to mid-September) +2.5% (mid-May to mid-June) -2.4% (September to early October) +2.9% (late March to early April) -2.5% (Mid-August to early October) -2.4% (September to early October) +1.8% (March) -3.6% (September to early October) -1.5% (Mid-June to Mid July)

Period most affected by temperature differences Mid-August to early October Mid-August to early October Late March to mid-June September to early October March – May Mid-August to early October September to early October March September to early October Mid-June to mid-August Late March to early October

Category

All Non-Food

Women’s Clothing

Men’s Clothing

Women’s Footwear

Men’s Footwear

Children’s Footwear

Furniture

Home Textiles -2.3% (Mid July to Mid August) Table 1: Summary of periods of peak temperature influences on sales and size of the effect

30 | spring 2019 | the retailer

Made with FlippingBook - Online catalogs