terclim by ICS banner
IVES 9 IVES Conference Series 9 OIV 9 OIV 2024 9 Short communications - Economy and law 9 Anticipating consumer preference for low-alcohol wine: a machine learning analysis based on consumption habits and socio-demographics

Anticipating consumer preference for low-alcohol wine: a machine learning analysis based on consumption habits and socio-demographics

Abstract

The global wine consumption landscape is undergoing a transformation, marked by a growing trend towards reduced consumption and a preference for healthier lifestyles. However, comprehensive studies investigating consumer preferences for low-alcohol wine remain scarce. This study aims to anticipate consumer preference for low-alcohol wine based on consumers wine consumption habits and their socio-demographics. The data were collected from 2022 to 2024 through online surveys to consumers in France, Germany, Italy and the United States, the four largest wine-consuming countries in 2021. The survey comprised questions regarding consumers’ socio-demographics, wine consumption habits and preference for alcoholic content of wine. We employed various machine learning algorithms, treating each survey year as a repetition to identify the best modeling approach. After evaluating the performance of these models, Random Forest emerged as the top performer. The analysis revealed that interactions between socio-demographic factors and buying habits factors are significant predictors of low-alcohol wine preferences. The study provides actionable insights for wine producers and marketers, offering a data-driven approach to more effectively target low-alcohol wine markets using accessible customer data such as demographics and sales records.

Anticipando la preferencia del consumidor por el vino bajo en alcohol: un análisis de aprendizaje automático basado en los hábitos de consumo y las  características socio-demográficas 

Objetivos y alcance: el panorama global del consumo de vino está experimentando una transformación, marcada por una tendencia creciente hacia un consumo reducido y una preferencia por estilos de vida más saludables. En línea con este cambio, la regulación de la unión europea (regulación UE 2021/2117) ha redefinido recientemente el vino desalcoholizado o parcialmente desalcoholizado dentro de la categoría del vino. Sin embargo, los estudios completos que investigan las preferencias del consumidor por el vino bajo en alcohol siguen siendo escasos. Este estudio tiene como objetivo anticipar la preferencia del consumidor por el vino bajo en alcohol basándose en los hábitos de consumo de vino de los consumidores y sus características socio-demográficas. Metodología: a través de una revisión exhaustiva de las investigaciónes relevantes publicadas, identificamos las principales características de los hábitos de consumo y exploramos el papel mediador de las características socio-demográficas en la influencia de las preferencias del consumidor. Los datos se recogieron de 2022 a 2024 mediante el envío de encuestas en línea a consumidores en Francia, Alemania, Italia y Estados Unidos, los cuatro países con mayor consumo de vino según la organización internacional de la viña y el vino (OIV). La encuesta incluyó preguntas sobre las características socio-demográficas de los consumidores, los hábitos de consumo de vino y la preferencia por el contenido alcohólico del vino. Empleamos diversos métodos de aprendizaje automático, tratando cada año de encuesta como una repetición para identificar el mejor enfoque de modelado. Después de evaluar el rendimiento de estos métodos, las máquinas de reforzamiento de gradiente (gbm) surgieron como las más destacadas. Resultados e implicaciones: mediante la modelización gbm, descubrimos que las interacciones entre los hábitos de consumo y las características socio-demográficas ejercían una mayor influencia que los hábitos de consumo solos en la preferencia del consumidor por el vino bajo en alcohol. Específicamente, las interacciones entre el país de residencia de los encuestados y la frecuencia de compra de vino, así como entre la ocupación y la frecuencia de compra de vino tinto y blanco, surgieron como factores relativamente más influyentes. En particular, las personas de alemania que raramente compraban ciertos tipos de vino mostraron una fuerte preferencia por las variedades bajas en alcohol, al igual que aquellos que actualmente no estaban empleados y compraban vino tinto o blanco mensualmente. Nuestro estudio subraya la importancia de considerar tanto los hábitos de consumo como los factores socio-demográficos para comprender las preferencias del consumidor por el vino bajo en alcohol. Comprender estas dinámicas puede informar estrategias dirigidas para la producción, comercialización y promoción de variedades de vino bajo en alcohol para satisfacer las cambiantes demandas y preferencias del consumidor.

Anticipando le preferenze dei consumatori per il vino a basso contenuto alcolico: un’analisi machine learning basata sulle abitudini di consumo e sulle caratteristiche socio-demografiche

Obiettivi e ambito: il panorama globale del consumo di vino sta subendo una trasformazione, contraddistinta da una tendenza crescente verso un consumo ridotto e una preferenza per stili di vita più sani. In linea con questo cambiamento, il regolamento dell’unione europea (regolamento UE 2021/2117) ha recentemente ridefinito il vino dealcolizzato o parzialmente dealcolizzato all’interno della categoria del vino. Tuttavia, studi completi che indaghino sulle preferenze dei consumatori per il vino a basso contenuto alcolico rimangono scarsi. Questo studio mira ad anticipare le preferenze dei consumatori per il vino a basso contenuto alcolico basandosi sulle abitudini di consumo del vino dei consumatori e sulle loro caratteristiche socio-demografiche. Metodologia: attraverso un’ analisi bibliografica estesa, abbiamo identificato le principali caratteristiche delle abitudini di consumo e esplorato il ruolo delle caratteristiche socio-demografiche nell’influenzare le preferenze dei consumatori. I dati sono stati raccolti dal 2022 al 2024 inviando sondaggi online ai consumatori in Francia, Germania, Italia e Stati Uniti, i quattro paesi con il maggior consumo di vino secondo l’organizzazione internazionale della vite e del vino (OIV). Il sondaggio comprendeva domande riguardanti le caratteristiche socio-demografiche dei consumatori, le abitudini di consumo del vino e la preferenza per il contenuto alcolico del vino. Abbiamo utilizzato vari metodi di machine learning, trattando ogni anno del sondaggio come una ripetizione per identificare l’approccio di modellizzazione migliore. Dopo aver valutato le performance di questi metodi, le gradient boosting machines (gbm) sono emerse come le più performanti. Risultati e implicazioni: abbiamo individuato che le interazioni tra le abitudini di consumo e le caratteristiche socio-demografiche esercitavano una maggiore influenza rispetto alle sole abitudini di consumo sulla preferenza dei consumatori per il vino a basso contenuto alcolico. In particolare, le interazioni tra il paese di residenza dei rispondenti e la frequenza degli acquisti di vino, così come tra l’occupazione e la frequenza degli acquisti di vino rosso e bianco, sono emerse come fattori relativamente più influenti. In particolare, gli individui provenienti dalla germania che acquistavano raramente certi tipi di vino manifestavano una forte preferenza per le varietà a basso contenuto alcolico, così come coloro che non lavoravano e acquistavano mensilmente vino rosso o bianco. Il nostro studio sottolinea l’importanza di considerare sia le abitudini di consumo che i fattori socio-demografici per comprendere le preferenze dei consumatori per il vino a basso contenuto alcolico. Comprendere queste dinamiche può informare strategie mirate per la produzione, il marketing e la promozione delle varietà di vino a basso contenuto alcolico per soddisfare le mutevoli esigenze e preferenze dei consumatori.

DOI:

Publication date: November 18, 2024

Issue: OIV 2024

Type: Article

Authors

Mingze Rui1, Federica Rosa2, Andrea Viberti2, Filippo Brun1, Simone Blanc1 and Stefano Massaglia1,3

1 Department of Agricultural, Forest, and Food Sciences, University of Turin, 10095 Grugliasco, Italy
2 RV Studio s.r.l, 12051 Alba, Italy
3 Centro Interdipartimentale Viticoltura e Vino (CONViVi), University of Turin, 12051 Alba, Italy

Contact the author*

Tags

Full papers OIV 2024 | IVES Conference Series | OIV | OIV 2024

Citation

Related articles…

Plastid genomics of Vitis vinifera L. for understanding the molecular basis of  grapevine (Vitis vinifera L.) domestication

The precise molecular mechanisms underlying the domestication of grapevine (Vitis vinifera L.) Are still not fully understood. In the recent years, next-generation sequencing (NGS) of plastid genomes has emerged as a powerful and increasingly effective tool for plant phylogenetics and evolution. To uncover the biological profile of the grapevine domestication process comprehensively, an investigation should encompass both the cultivated varieties (V. vinifera subsp. Vinifera) and their wild ancestors V. vinifera subsp. Sylvestris) across all potential sites of their distribution and domestication.

Smoke tainted wine – what now?

The frequency of bushfires close to wine regions around the world has increased in the last two decades. The economic losses incurred when grapes and wines are discarded due to ‘smoke taint’ are substantial (i.e., hundreds of millions of dollars). Efforts to mitigate and ameliorate smoke taint are therefore crucial. Chardonnay, rosé and cabernet sauvignon wines made from grapes exposed to smoke during the 2020 wildfires in eastern Australia were subjected to various amelioration techniques: the addition of activated carbons, molecularly imprinted polymers (mips), and a proprietary resin (either directly, or following membrane filtration); spinning cone column (scc) distillation; and finally, transformation into vinegar.

Genetic traceability of the varietal origin of wines: a robust application for must and wines during alcoholic fermentation

Industry and regulatory agencies have developed regulations to ensure authenticity and compliance with wine composition limits. However, this can be truncated by the absence of simple and robust analytical methodologies, uninfluenced by the environment, different oenological techniques and cultural practices. Genetic fingerprinting is the most powerful tool for unequivocal varietal identification; it is not affected by the environment or agronomic practices; however, its usefulness in musts and wines has been controversial and there is currently no routine certification of varietal origin based on DNA analysis.

Effect of moderate wine consumption in animal models

In 1979, the so-called “french paradox” was proposed, that is, a correlation between wine consumption, a diet rich in saturated fats, and a low mortality from coronary heart disease. On the other hand, it has also been described that alcohol consumption has negative effects on aging and increases the risk of liver cirrhosis and cancer. However, both hypotheses are based on population studies that may present distortions due to multiple factors (geographic, diet, smoking, socioeconomic level, etc.).

Contribution of grape seeds to evolution of acetaldehyde, pigments and tannins reactive towards salivary proteins of red wine over time

This study investigated the impact of the gsk/gse ratio on the evolution of acetaldehyde and of major phenolic compounds of aglianico wine in wine like solution and real wine. Four model solutions and the correspondant control wines were prepared. The natural weight ratio between grape skins and seeds was determined on the real grapes, and a control wine was obtained from those.