The last region, which was observed in our research is Ural. The weather conditions in this area do not changes often, on the other hand the changes is very rapid. The main reason is the continental climate of this region, so it was expected that model for the Ural will have the high level of the R square parameter.

The models equations with fixed effects Year, SKU, City:

Sales Volume in Ural region (daily data) =fixed effects + 0,068*PricePerDecalitre + 0,557*Weather.T_mean - 0,244*Weather.Po_maх - 0,275*Weather.U_mean- 1,171*Weather.Ff_min -39,746*Day_Off	(14)
Sales Volume in Ural region (weekly data) =fixed effects +0,450*PricePerDecalitre + 3,591*Weather.T_mean - 1,910*Weather.Po_maх - 39,148*Weather.Ff_min+133,411*Day_Off	(15)

And the result of the model prove this hypothesis, the R squared value for the weekly is 0,592, the highest level from all the regions, which were discussed earlier. Furthermore, the model includes only five parameters. Average temperature leads to the rise of volume of sales, the increase the average temperature on 1 degree сelsius provide the growth of volume of sales on approximately 3,6 dekaliter. Atmosphere pressure and speed wind have negative impact on the volume of sales, their coefficients are - 1,910, -39,148.

Table 10 Regression results for Ural region

For daily and weekly data connection with price, average temperature and volume of sales are positive. Humidity and atmosphere pressure in millimeters of mercury, wind speed have negative influence on volume of sales. It is external effect, so, the direction is clear. Non-working days shows the same connection as it was described earlier.

The Ural region characterizes by sharply continental climate, so the high wing speed is not usual for this region, so the rise of wind speed leads to the decrease of sales volumes.

Finally, in table 12 was gathered all results for the regression models for daily data for all regions to estimate the similarities and differences. the R-squared value ranges from 0,383 to 0,408.

The similarities between the regions, which are noticed by the regression model, that in all regions the volume of sales relates to the average air temperature and the influence of this parameter is positive. The weather parameter of wind speed in all region except Siberia and FE leads to the decrease of volume of sales. The humidity parameter in all regions, where this weather factor is significant has minus coefficients. Atmosphere pressure in millimeters of mercury in Moscow Center and North West region has positive influence for volume of beer sales, on the other hand, in all other region this parameter is significant as well, but with opposite way of relation.

Table 11 Regression results for Ural region with factor variable

The R-squared value for the models with weekly data ranges from 0,525 to 0,594. Consequently, approximately 60% of sales volume could be explain by these models. The results of the models on table 10 shows us the main differences between the models in different regions.

Table 12 Regression results for regions in daily data

==================================================================================

Dependent variable:

-------------------------------------------------------------------------------------------------------------------------------

Sales_volume Daily data

Ural Moscow-Center North-West South Siberia &FE

-------------------------------------------------------------------------------------------------------------------------------

PricePerDecalitre 0.068*** -0.031*** 0.015** -0.025***

(0.004) (0.004) (0.007) (0.005)

discount -0.139***

(0.006)

Weather.T_mean 0.557*** 1.039*** 1.051*** 1.444*** 0.450***

(0.018) (0.019) (0.033) (0.028) (0.021)

Weather.Po_max -0.244*** 0.059** 0.128*** -0.560***

(0.026) (0.028) (0.040) (0.042)

Weather.U_mean -0.275*** -0.222***

(0.014) (0.015)

Weather.U_max -0.436***

(0.039)

Weather.Ff_min -1.171*** -1.781*** -2.007*** 0.601***

(0.161) (0.166) (0.302) (0.141)

Weather.Ff_mean -0.312***

(0.113)

Day_Off -39.746*** -41.170*** -42.567*** -27.953*** -42.638***

(0.410) (0.450) (0.690) (0.411) (0.494)

--------------------------------------------------------------------------------------------------------------------------------

Observations 636,990 608,591 263,144 620,703 449,041

R2 0.395 0.383 0.386 0.391 0.408

Adjusted R2 0.394 0.383 0.385 0.390 0.408

Residual Std. Error 133.381 (df = 636591) 140.609 (df = 608168) 149.106 (df = 262784) 136.035 (df = 620344) 131.993 (df = 448720)

==================================================================================

Note: *p<0.1; **p<0.05; ***p<0.01

In the table 13 we could witness the results of regional regression on the basis on weekly data. The results of the models in table 13 shows us the main differences between the models in different regions. In more detail, in each model was included the price factor. In Ural and Moscow Center region price per dekaliter in rubles found out to be significant and positive effect on volume of sales. In North West and Siberia region the rise of discount in rubles per liter leads to the decrease of volume of beer sales. Air temperature significant and positive effect variable for all the regions. The rise of minimum wind speed in Ural and South regions leads to the decrease of volume of sales. Controversially, in the North West region the increase of maximum of wind speed connected with growth of beer sales. The R-squared value for the models with weekly data ranges from 0,525 to 0,594 . Consequently, approximately 60% of sales volume could be explain by these models. At this moment company tries to implement forecast process based on big data analysis. It is not an easy task in case of hard to gathered data from different sources, moreover, maintaining master data and setting the importing the reliable external data. So, nowadays the value of accuracy of forecast , which is made by people by using different sources and analytics close to 50%, it is the best numbers of weekly forecast accuracy.

Table 13 Regression results for regions in weekly data

==================================================================================Dependent variable:

-------------------------------------------------------------------------------------------------------------------------------

Sales_volume Weekly data

Ural Moscow-Center North-West South Siberia &FE

-------------------------------------------------------------------------------------------------------------------------------

PricePerDecalitre 0.450*** 0.146*** 0.134***

(0.028) (0.033) (0.035)

discount -1.014*** -0.448***

(0.087) (0.047)

Weather.T_mean 3.591*** 6.804*** 9.648*** 4.748***

(0.128) (0.162) (0.290) (0.190)

Weather.T_max 7.103***

(0.258)

Weather.Po_max -1.910*** 15.431*** 11.790*** 5.098*** 7.292***

(0.231) (0.524) (0.810) (0.372) (0.433)

Weather.U_mean -0.543***

(0.154)

Weather.U_max 8.927***

(0.580)

Weather.Ff_min -39.148*** -19.713***

(1.508) (1.691)

Weather.Ff_max 6.716***

(1.683)

Day_Off 133.411*** 131.049*** 183.475*** 124.573*** 100.629***

(1.783) (2.016) (3.354) (1.999) (2.600)

-------------------------------------------------------------------------------------------------------------------------------

Observations 138,023 130,415 53,245 129,865 97,136

R2 0.594 0.525 0.536 0.551 0.558

Adjusted R2 0.592 0.523 0.533 0.549 0.557

Residual Std. Error 472.596 (df = 137625) 539.145 (df = 129992) 563.121 (df = 52884) 530.319 (df = 129506) 491.433 (df = 96815)

==================================================================================Note: *p<0.1; **p<0.05; ***p<0.01

Looking at the R-squared value, which shows that the model could describe a variable on approximately 60% we could conclude that usage of our model for forecasting process could add a value and accuracy for it. Therefore, these regional regression models on the basis of weekly dataset is considered as final for the research.

Overall, the results of the study confirm the hypotheses, which were stated for the research. The first hypothesis are proven by the multiple regression model with fixed effects that weather conditions parameters effect of volume of beer product sales in Russia. The second hypothesis is partly confirmed. Weather variables as air temperature, humidity, atmosphere pressure and wind speed are significant for the research, but not for all 5 regional model. The third hypothesis is fully confirmed, as all five regional models shoes that air temperature is significant and the increase of air temperature leads to the rise of volume of beer sales. The fourth hypothesis about negative effect of humidity level on volume of sales was proven only for the South region. In North West the rate of humidity has positive influence on sales. Furthermore, the most significant value is the maximum of the humidity rate in a week. The results of other regional regression models denying the significance of humidity rate on the beer consumption. Fifths hypothesis was approved by the four of five regional models, all the regions except of Ural illustrates positive connection with the atmosphere pressure and volume of beer sales. In Ural increase of atmosphere pressure leads to the decrease of volume of beer sales. The fourth hypothesis about negative effect of wind speed was proved in Ural and South region. Otherwise in North West region the increase of wind speed leads to the rise of volume of beer sales. Finally, based on the analysis of the model results, we can conclude that model for different regions shows different lists of significant variable and different relation between then and independent variable.

Conclusion

This study set out to determine the relation between volume of beer sales and the weather conditions. The empirical findings in this study provide a new understanding of the weather conditions as a variables, which affect the consumption of beer in Russia and in particular regions, cities. According to considered in this research regression models in general approximately 60% of beer sales could be described by price, weather condition and non-working days in the region of sales.

These findings suggest that the list of significant variables consist of 5 weather parameters: air temperature (degrees Celsius), atmospheric pressure at station and sea level (millimeters of mercury), relative humidity (%), wind speed at an altitude of 10-12 meters above the earth's surface, averaged over a 10-minute period immediate, 2 price factors as price per dekaliter and discount and control variable as non-working days.

Key trends and changes in the behavior of Russian consumers are connected with becoming consumers insightful and influential. Experts believe that most organizations need to invest much more in improving the consumer experience. The evolution of retail is driven by the need to adapt to a changing environment and respond quickly to these challenges. Organizations that do not do this will fall behind on the way to winning customers. The result findings of this survey could make positive effect to the FMCG and retail industry, because it shows on what factors companies should concentrate their attention in making forecasting models to make it more accurate. The more accurately company could predict future demand, the less losses they will generate at all stages of the supply chain: purchasing, manufacturing, warehouses, and transportation. Looking at the other end of the supply chain, an underestimated forecast level leads to low shelf availability (OSA), and as a result, lost sales revenue. Furthermore, companies could use our coefficient and model to make a predictive model for other their needs.

A key to success for the FMCG company is to make sure that it's products always are on the shelves in the market when the consumer is ready to buy it. If company produce too little and the shelf is empty, the consumer may choose the competitor's brand and may never come back, if company produces too much, the beer become spoiled. Furthermore, FMCG companies might even incur penalties from its customers if companies do not deliver the right products at the right time in the right quantities.

Therefore, companies needs to forecast as precisely as possible how much the consumers are going to buy in future, and plan it's production and procurement of raw materials accordingly.

Finally, a number of important limitations need to be considered. First, observed data sample could be bigger, it could include information about sales not only from one Russian company, but from the whole industry. Second, the model could be more complicated, using other methodology tools. Third, the survey could be added with some kind of qualitative analysis, for instance, the questionnaire of consumers.

It is recommended that further research be undertaken considering such important variables such as particular sales, bonuses, use other methodologies to understand more about relations of weather conditions on sales.

References

1. Adebanjo, D., & Mann, R. (2000). Identifying problems in forecasting consumer demand in the fast moving consumer goods sector. Benchmarking: An International Journal, 7(3), 223-230.

2. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior.

3. Aleksina, S.B. (2018). Scenarios for the development of distribution strategy in the FMcG market under increasing pressure from federal trade networks. Russian Journal of Entrepreneurship.

4. Ares, G., & Gбmbaro, A. (2007). Influence of gender, age and motives underlying food choice on perceived healthiness and willingness to try functional foods. Appetite, 49(1), 148-158.

5. Babongo, F., Appelqvist, P., Chavez-Demoulin, Valйrie, Hameri, Ari-Pekka, Niemi, Tapio. (2018) Using weather data to improve demand forecasting for seasonal products. 53-76.

6. Badorf,F., Hoberg, K. (2020) The impact of daily weather on retail sales: An empirical study in brick-and-mortar stores. Journal of Retailing and Consumer Services.

7. Bahng, Y., &Kincade, D. H. (2012). The relationship between temperature and sales: Sales data analysis of a retailer of branded women's business wear. International Journal of Retail & Distribution Management, 40(6), 410-426.

8. Bahng, Y., &Kincade, D. H. (2012). The relationship between temperature and sales: Sales data analysis of a retailer of branded women's business wear. International Journal of Retail & Distribution Management, 40(6), 410-426.

9. Behe, B. K., Getter, K. L., & Yue, C. (2012). Should you blame the weather? The influence of weather parameters, month, and day of the week on spring herbaceous plant sales in the US midwest. HortScience, 47(1), 71-73.

10. Bertrand, J. L., Brusset, X., & Fortin, M. (2015). Assessing and hedging the cost of unseasonal weather: Case of the apparel sector. European Journal of Operational Research, 244(1), 261-276.

11. Blom, J. E. (2009). Hedging revenues with weather derivatives. Magistarski rad. Norwegian School of Economics, Norway.

12. Blom, J. E. (2009). Hedging revenues with weather derivatives. Magistarski rad. Norwegian School of Economics, Norway.

13. Blondй, B., &Ryckbosch, W. (2015). In `splendid isolation'. A comparative perspective on the historiographies of the `material renaissance'and the `consumer revolution'. History of Retailing and Consumption, 1(2), 105-124.

14. Bratina, D., &Faganel, A. (2008). Forecasting the primary demand for a beer brand using time series analysis. Organizacija, 41(3), 116-124.

15. Carvalho, F., Moors, P., Wagemans, J. and Spence, C. (2017) The Influence of Color on the Consumer's Experience of Beer. Front. Psychol. 8:2205. doi: 10.3389/fpsyg.2017.02205

16. Cohen, L. (2004). A consumers' republic: The politics of mass consumption in postwar America. Journal of Consumer Research, 31(1), 236-239.

17. Changhui, C., Eyunghee, K., and Kim, Changki. (2011) A Way of Managing Weather Risks Considering Apparel Consumer Behaviors.

18. Colen, L., &Swinnen, J. (2016). Economic growth, globalisation and beer consumption. Journal of Agricultural Economics, 67(1), 186-207.

19. Conlin, M., O'Donoghue, T., & Vogelsang, T. J. (2007). Projection bias in catalog orders. american economic Review, 97(4), 1217-1249.

20. De Goeij, M. C., Suhrcke, M., Toffolutti, V., van de Mheen, D., Schoenmakers, T. M., & Kunst, A. E. (2015). How economic crises affect alcohol consumption and alcohol-related health problems: a realist systematic review. Social Science & Medicine, 131, 131-146.

21. De Graaf, C., Kramer, F. M., Meiselman, H. L., Lesher, L. L., Baker-Fulco, C., Hirsch, E. S., & Warber, J. (2005). Food acceptability in field studies with US army men and women: Relationship with food intake and food choice after repeated exposures. Appetite, 44(1), 23-31. https://doi.org/10.1016/j.appet.2004.08.008.

22. Denissen, J. J., Butalid, L., Penke, L., & Van Aken, M. A. (2008). The effects of weather on daily mood: A multilevel approach. Emotion, 8(5), 662.

23. Digon, E., & Bock, H. B. (1966). Suicides and climatology. Archives of Environmental Health: An International Journal, 12(3), 279-286.

24. Divakar, S., Ratchford, B. T., & Shankar, V. (2005). Practice Prize Article--CHAN4CAST: A Multichannel, Multiregion Sales Forecasting Model and Decision Support System for Consumer Packaged Goods. Marketing Science, 24(3), 334-350.

25. Drбpal, J., & Pina-Sбnchez, J. (2019). Does the weather influence sentencing? Empirical evidence from Czech data. International Journal of Law, Crime and Justice, 56, 1-12.

26. Dubй, J. P. (2004). Multiple discreteness and product differentiation: Demand for carbonated soft drinks. Marketing Science, 23(1), 66-81.

27. Ferreira, Johnson, K., Hong, B., Lee, A., Simchi-Levi, D.(2016) Analytics for an online retailer: demand forecasting and price optimization. 69-88.

28. Gaure, S. (2013). lfe: Linear group fixed effects. The R Journal, 5(2), 104-117.

29. Gustavsen, G. W. (2005). Public policies and the demand for carbonated soft drinks: a censored quantile regression approach(No. 724-2016-49244).

30. Hartwick, E. (1998). Geographies of consumption: a commodity-chain approach. Environment and planning D: society and space, 16(4), 423-437.

31. Haugtvedt, C. P., Petty, R. E., & Cacioppo, J. T. (1992). Need for cognition and advertising: Understanding the role of personality variables in consumer behavior. Journal of Consumer Psychology, 1(3), 239-260.

32. Hinnosaar, M.. (2016) Time inconsistency and alcohol sales restrictions. European Economic Review. 108-131

33. Howarth, E., & Hoffman, M. S. (1984). A multidimensional approach to the relationship between mood and weather. British Journal of Psychology, 75(1), 15-23.

34. Jackson, P., Del Aguila, R. P., Clarke, I., Hallsworth, A., De Kervenoael, R., &Kirkup, M. (2006). Retail restructuring and consumer choice 2. Understanding consumer choice at the household level. Environment and Planning A, 38(1), 47-67.

35. Jackson, T. (2005). Motivating sustainable consumption. Sustainable Development Research Network, 29(1), 30-40.

36. Karjaluoto, H., Karvonen, J., Kesti, M., Koivumдki, T., Manninen, M., Pakola, J., ... &Salo, J. (2005). Factors affecting consumer choice of mobile phones: Two studies from Finland. Journal of Euromarketing, 14(3), 59-82.

37. Keller, M. C., Fredrickson, B. L., Ybarra, O., Cфtй, S., Johnson, K., Mikels, J., ... & Wager, T. (2005). A warm heart and a clear head: The contingent effects of weather on mood and cognition. Psychological science, 16(9), 724-731.

38. Koksalan, M., Erkip, N., & Moskowitz, H. (1999). Explaining beer demand: A residual modeling regression approach using statistical process control. International journal of production economics, 58(3), 265-276.

39. Kossova, T., Kossova, E., &Sheluntcova, M. (2017). Alcohol consumption and individual time preferences of Russians. International Review of Economics, 64(1), 47-85.

40. Kossova, T., Kossova, E., &Sheluntcova, M. (2017). Investigating the volume and structure of alcohol consumption in Russian regions. Journal of Economic Studies, 44(2), 266-281.

41. Krasilnikova, E.A. and Nikishin, A.F. (2018) Regional retail of the Russian Federation: trends and prospects. Russian Journal of Entrepreneurship.

42. Kubiљta, M. (2018). Analysis of Weather Effect on Sales in the Czech FMCG Market.

43. Lipsey, R. G., Courant, P. N., & Ragan, C. T. (1999). Macroeconomics, 12e.

44. Maheswaran, D., &Shavitt, S. (2000). Issues and new directions in global consumer psychology. Journal of consumer psychology, 9(2), 59-66.

45. Mani, V., Kesavan, S, Jayashankar, M., Swaminathan. (2015) Estimating the impact of understaffing on sales and profitability in retail stores. 201-218.

46. Miller, D. (1995). Consumption as the vanguard of history in Miller D ed Acknowledging consumption.

47. Moschis, G. P. (1985). The role of family communication in consumer socialization of children and adolescents. Journal of consumer research, 11(4), 898-913.

48. Murray, K. B., Di Muro, F., Finn, A., &Leszczyc, P. P. (2010). The effect of weather on consumer spending. Journal of Retailing and Consumer Services, 17(6), 512-520.

49. Murray, K. B., Di Muro, F., Finn, A., &Leszczyc, P. P. (2010). The effect of weather on consumer spending. Journal of Retailing and Consumer Services, 17(6), 512-520.

50. Nijman, M., James, S., Dehrmann, F., Smart, K., Ford, R., Hort, J.(2019) The effect of consumption context on consumer hedonics, emotional response and beer choice. Food Quality and Preference. 59-71.

51. Nguyen, T.T, Bui, Q.T., Hans, S.S., Yingkui, Y. (2018) The role of packaging format, alcohol level and brand in consumer's choice of beer: A best-worst scaling multi-profile approach. Food Quality and Preference. 92-100.

52. Parsons, A. G. (2001). The association between daily weather and daily shopping patterns. Australasian Marketing Journal (AMJ), 9(2), 78-84.

53. Peresie, J. L. (2004). Female judges matter: Gender and collegial decisionmaking in the federal appellate courts. Yale LJ, 114, 1759.

54. Perlman, F. J. (2010). Drinking in transition: trends in alcohol consumption in Russia 1994-2004. BMC public health, 10(1), 691.

55. Peter, J. P., Olson, J. C., &Grunert, K. G. (1999). Consumer behaviour and marketing strategy (pp. 329-48). London: McGraw-Hill.

56. Peter, J. P., Olson, J. C., &Grunert, K. G. (1999). Consumer behaviour and marketing strategy (pp. 329-48). London: McGraw-Hill.

57. Ribeiro, A. P., Harmsen, R., Carreуn, J. R., & Worrell, E. (2018). What influences consumption? Consumers and beyond: purposes, contexts, agents and history. Journal of Cleaner Production.

58. Ribeiro, A. P., Harmsen, R., Carreуn, J. R., & Worrell, E. (2018). What influences consumption? Consumers and beyond: Purposes, contexts, agents and history. Journal of cleaner production.

59. Rind, B., &Strohmetz, D. (2001). Effect of beliefs about future weather conditions on restaurant tipping. Journal of Applied Social Psychology, 31(10), 2160-2164.

60. Rubtsova N.V. &Vygovskaya P. D. Limitation of advertising of alcohol products as factors of influence of the conduct of consumers. Azimuth of research:economic and management
Volume: 8 Number: 1 (26) Year: 2019 Pages: 282-284

61. Sбenz-Navajas, M. P., Campo, E., Sutan, A., Ballester, J., & Valentin, D. (2013). Perception of wine quality according to extrinsic cues: The case of Burgundy wine consumers. Food Quality and Preference, 27(1), 44-53.

62. Sanders, J. L., &Brizzolara, M. S. (1982). Relationships between weather and mood. Journal of General Psychology.

63. Saunders, J., Wong, V., Armstrong, G., & Kotler, P. (2008). Principles of marketing.

64. Sester, C., Deroy, O., Sutan, A., Galia, F., Desmarchelier, J.-F., Valentin, D., & Dacremont, C. (2013). “Having a drink in a bar”: An immersive approach to explore the effects of context on drink choice. Food Quality and Preference, 28(1), 23-31. https://doi.org/ 10.1016/j.foodqual.2012.07.006.

65. Schnaiberg, A. (1980). Environment: from surplus to scarcity. In Environment: from surplus to scarcity. Oxford University Press.

66. Shinkina, N.A. (2013). Technology and features of launching a new brand on the FMCG market . Marketing communications. 2013.No 05 (77). 316-328.

67. Skorobogatov A.S. The effect of closing hour restriction on alcohol use and abuse in Russia. Journal of Institutional Studies 2016, Value 8 (number 3),
pages 72-90

68. Steinker, S., Hoberg, K., Thonemann, U. (2017) The Value of Weather Information for E-Commerce Operations. Production and Operations Management. 1854-1874

69. Љtulec, I., Petljak, K., &Naletina, D. (2019). Weather impact on retail sales: How can weather derivatives help with adverse weather deviations?. Journal of Retailing and Consumer Services, 49, 1-10.

70. Tekin, E. (2004). Employment, wages, and alcohol consumption in Russia. Southern Economic Journal, 71(2), 397-418.

71. Tian, J., Zhang, Y., & Zhang, C. (2018). Predicting consumer variety-seeking through weather data analytics. Electronic Commerce Research and Applications, 28, 194-207.

72. Tomkins, S., Saburova, L., Kiryanov, N., Andreev, E., McKee, M., Shkolnikov, V., & Leon, D. A. (2007). Prevalence and socio-economic distribution of hazardous patterns of alcohol drinking: study of alcohol consumption in men aged 25-54 years in Izhevsk, Russia. Addiction, 102(4), 544-553.

73. Trentmann, F. (2016). Empire of things: How we became a world of consumers, from the fifteenth century to the twenty-first. Penguin UK.

74. Williams, Brent D., Waller, Matthew A., Ahire, Sanjay, Ferrier, Gary D. (2014) Predicting retailer orders with POS and order data: the inventory balance effect. 593-600.

75. Wooldridge, J. M., & Joyner, E. (2013). Multiple regression analysis: Estimation. Introductory Econometrics: A Modern Approach. 5th ed. Mason, OH: South-Western, 68.Yormirzoev, M., &Gabrielyan, G. (2016). Factors Associated with Alcoholic Beverages Consumption in Russia: A Discrete Choice Model. Journal of Agricultural & Food Industrial Organization, 14(1), 133-139.

Appendix

Sales volumes per city

The first eight rows in resulted data base

Seasonality of sales in volume, dekaliter

Seasonality of temperature

Identifying outliers

Relation air temperature and sales volume in dekaliter

Atmosphere pressure and sales volume in dekaliter

Размещено на Allbest.ru