Sustainability and Quality in Wine Pricing: Insights from the Primitivo E-Commerce Market

Abstract

The wine sector faces urgent challenges related to climate change, resource constraints, rising energy costs, and shifting consumption patterns. In this context, sustainability represents a lever of competitiveness and resilience, enabling producers to differentiate themselves through their connection with the territory and the preservation of artisanal practices. This is particularly evident for native varieties, such as the Apulian Primitivo, whose market repositioning is based on production choices that combine sustainability and craftsmanship, generating wines with distinctive organoleptic characteristics while also contributing to the environmental, economic, and social protection of the production areas. At the same time, the spread of e-commerce offers new opportunities to enhance these characteristics: online descriptions enrich the communication of details compared to the physical label, reducing information asymmetries and making signs of quality and sustainability more visible. Therefore, this study applies a hedonic model to estimate the implicit price of such practices, analyzing 656 Primitivo labels through simultaneous quantile regressions. The results show that sensory characteristics, such as alcohol content and ageing, determine the price throughout distribution, while identity and sustainable attributes have heterogeneous effects: ecological practices and Geographical Indications affect the lower quantiles, while indigenous yeasts and sapling vines are configured as premium signals in the upper segments.

1. Introduction

In recent years, sustainability has emerged as a key strategic lever in agri-food systems, becoming increasingly relevant in business and public policy. Food systems are currently at the crossroads of global challenges such as resource depletion, climate change, and socioeconomic inequality, as noted by El Bilali et al. (2024) [1]. These issues have a significant impact on the production, distribution, and consumption of food. In this context, agri-food markets are intricate, dynamic systems whose operations have a direct impact on social, economic, and environmental balances [2].

They are strategic hubs to address global tensions and to implement the principles of the 2030 Agenda, promoting sustainable practices along the entire supply chain. Addressing these challenges implies a systemic reconfiguration of production and distribution models and requires a transdisciplinary approach to identify which system functions should be phased out, supported or redesigned, while building a shared agenda of research, policies and actions.

Borsellino et al. (2020) [2] underline that the transformation underway represents both a challenge and an opportunity: creating effective, inclusive, and nutrient-conscious marketplaces that can reduce environmental externalities, promote collaborative governance models throughout the supply chain, and value low-impact agricultural approaches (integrated, organic, and precision farming) is essential. However, the literature shows that the transition to sustainable production models encounters numerous barriers, which vary according to firm size, sector, and adopted technologies. In the case of small and medium-sized enterprises, for example, the most recurrent constraints concern the lack of resources, high initial capital costs, and the lack of managerial and technical skills [3] (Álvarez Jaramillo et al., 2019). The systematic review by Campuzano et al. (2023) [4], which focuses precisely on identifying barriers to sustainable innovation in the agricultural sector, highlights the persistence of both internal (financial resources, skills, organizational mindset) and external barriers, including the lack of favourable policies and regulations and the scarcity of skilled labour.

Despite these constraints, innovation remains an essential tool to face the challenges of the transition. According to Cecchini et al. (2018) [5], the “sustainability-oriented innovation” paradigm integrates social, environmental, and economic objectives going beyond the traditional linear research-development logic. Precision technologies, conservative agriculture, process and product eco-innovations, but also organizational solutions and multi-stakeholder collaborative models design paths to reduce chemical inputs, save water, cut emissions and enhance biodiversity. From this perspective, innovation extends beyond the technological aspect; it also includes supply chain management, governance practices, and consumer communication in an effort to influence their perception of value and their willingness to pay for sustainable products. As sustainability in the food sector is increasingly conceptualized as a multidimensional construct, which consumers perceive and hierarchize differently, from “protection of natural resources” to “accessibility to healthy food” to “decent working conditions” [6], it becomes crucial to understand how these meanings influence purchasing behaviours. Consequently, assessing consumers’ propensity to change their eating habits and to recognize the added value of sustainable products becomes a priority not only to guide supply, but also to promote more responsible and equitable eating patterns [2].

As reported by Cecchini et al. (2018) [5], the post-modern consumer is more informed and inclined to consider ethical, social, and environmental values when making purchases, rather than being driven by price or taste. Multiple studies provide evidence of this behavioural shift, showing that sustainability labels (e.g., organic, fair-trade, reduced carbon impact, Geographical Indications, etc.) have become significant drivers of purchasing decisions [7,8]. However, experimental research indicates that consumers are willing to pay a substantial premium for “responsible” products, only when both informational transparency and sensory quality are ensured. Without these conditions, the search for sensory enjoyment is prioritized over sustainable awareness, highlighting the need to combine taste performance with value-driven content [9,10].

This ambivalence of demand is reflected in the challenges of supply: if consumers reward sustainability only when it is accompanied by transparency and quality, companies are called upon to innovate without neglecting efficiency and sensory performance. The tensions between innovation drives, implementation limits and consumer expectations find an emblematic example in the wine supply, which, to regenerate ecosystems, reduce emissions, and ensure social equity, must address five key challenges along the entire value chain: the impact of climate change and adaptation strategies; reducing emissions and creating carbon sinks; the responsible use of inputs in the vineyard; innovation in packaging; and the social and economic sustainability of producing communities [11] (Wagner et al., 2023). As Golicic (2022) [12] shows, many companies follow an evolutionary path, starting with environmental aspects, then moving on to economic balance, and finally integrating the social dimension. This gradual process reflects the structural difficulties of the transition, as “green” approaches can lead to higher labour requirements, temporary drops in yield, and increased unit costs, generating tension between reduced inputs, production targets, and profit margins.

At the same time, several studies have shown that the transition from conventional practices to lower-impact solutions can, under specific soil and climate conditions, strengthen economic margins by combining yield stability with improved product quality. Research conducted in Italian wine contexts has shown that the adoption of integrated approaches makes it possible to combine environmental and economic sustainability, albeit with some trade-offs related to yields [13]. Similarly, low-input regimes increase operational complexity and short-term risks but, under specific yield profiles and pedoclimatic conditions, allow for the joint achievement of environmental and economic performance [14,15].

In the vineyard, the answers to the challenges of sustainability, aimed at reducing the use of chemicals, water and energy, are realized through three main guidelines: the adoption of traditional low-yield training systems, which have demonstrated significant physiological and microclimatic benefits in conditions of increasing climatic stress, helping to limit water loss and protect the bunches during heat waves [16]; the use of resilient native grape varieties, which offer a concrete solution for the reduction in pesticide use and for the promotion of more sustainable viticulture, in line with the objectives of the European Green Deal [17]; and the progressive extension of vineyard areas conducted with sustainable methods, in particular organic and biodynamic, which is recorded in both traditional and emerging producing countries [18].

In the winery, practices that focus on sustainability and quality are being explored. This includes using local yeasts, spontaneous fermentations, limited sulphites, ageing in low-impact materials, and reducing fining and filtration.

The use of indigenous yeasts has proved to be an effective strategy to strengthen the authenticity of the wine, enhancing the sensory characteristics linked to the microflora of the territory [19]. Spontaneous fermentation, which is based precisely on these yeasts naturally present on the grapes and in the vineyards, makes it possible to obtain unique wines, with a strong regional identity, while requiring careful management to ensure their quality and stability.

The use of sulphites is also the subject of attention: although they are commonly used as antioxidants and antimicrobials, their use raises growing health concerns. For this reason, alternative solutions are being experimented with, such as natural extracts rich in phenolic compounds, which make it possible to significantly reduce the sulphur dioxide content in wines, offering concrete answers to the growing demand for more natural and healthy products [20].

In parallel, research is exploring natural fining agents that are more sustainable than traditional ones: among these, pomace has been shown to be effective in removing tannins and anthocyanins, with equal or superior performance compared to commercial agents [21].

Finally, the refining materials are also evolving. Several wood species, in addition to oak, are being studied for their unique chemical compositions and their effects on wine quality [22]. In addition, ceramic and cementitious materials are emerging as alternatives to wood and stainless steel. These alternative ageing materials offer winemakers new tools to improve wine quality, reduce costs, and improve the sustainability of the industry.

These choices, along with lightweight packaging and short-chain logistics, lead to product features that appear on wineries’ labels and stories. They play an increasing role in consumer buying decisions and act as strategic tools for strengthening market competitiveness.

1.1. Recent Dynamics of the Wine Market

The wine industry is one of the agricultural sectors with the highest added and symbolic value, where quality, sustainability, and territorial identity are all closely interconnected. However, the industry is undergoing a deeper restructuring process on a global scale. According to ISMEA (2025) [23], global wine production in 2024 reached its lowest level in the last ten years, standing at around 221 million hectolitres. Europe maintains its leadership, but its weight has dropped to 64% of the total, while international trade has stabilized at around 36 million hectolitres and 36 billion euros. Italy keeps playing a major role in this scenario: in 2024, it reclaimed the top spot as the world’s largest producer, reaching 44 million hectolitres—a 15% rise compared to 2023. It also confirmed its position as the top exporter by volume (21.7 million hectolitres) and second by value (8.1 billion euros) [23]. With 529 recognized PDO and PGI designations, the industry is valued at over €14 billion, equal to 10% of the national agri-food turnover. Even on the national market, wine remains a cornerstone of Italian food culture, with consumption stable at around 22 million hectolitres, corresponding to approximately 37.8 litres per capita.

Apulian Viticulture: The Case of Primitivo

Apulia is the leading Italian wine-producing region by volume, with over 90,000 hectares of vineyards and more than 1.4 million hectolitres of wine produced in 2024 (ISTAT, 2024). The region includes 29 DOC, 4 DOCG, and 6 IGT that protect native grape varieties. Among these, Primitivo stands out and has gained growing attention in both national and international markets.

Originally from Dalmatia, the Primitivo grape arrived in Apulia more than two thousand years ago. It found its ideal growing conditions in the Manduria area and, by the 1990s, became the ninth most planted grape variety in Italy [24]. For many years, Primitivo was mainly used for blending. However, lower yields and better winemaking methods helped to reposition it as a red wine with strong character, appreciated again by the wider public [24]. Its commercial revival also coincides with efforts to strengthen local identity. “Primitivo di Manduria” received DOC status in 1974 and was later awarded DOCG recognition in its “Dolce Naturale” version in 2011. The production area includes 18 towns in the provinces of Taranto and Brindisi, covering more than 5000 hectares. Many of the vineyards still use the traditional bush-trained vines method, which fits well with the Ionian-Salento climate [25].

According to Circana data processed for Vinitaly [26], Primitivo saw a 12.3% increase in value in 2024 compared to the year before, reaching €40.6 million in sales in large-scale retail, with about 7.5 million litres sold and an average price of €6.20 per bottle. Over the 2019–2024 period, it recorded the second-highest annual growth rate (+10%), just behind Vermentino (+11%). Primitivo is also among the top five red wines by average price per litre, ahead of Barbera, Nero d’Avola, and Montepulciano.

This combination of strong regional roots and growing interest in native, resilient grape varieties make Primitivo a good case for exploring whether the market recognizes a premium for quality and sustainability attributes shown on labels. To study this, the literature often uses hedonic pricing models, which are common in wine research to estimate how product features affect prices.

1.2. Bibliographic Review: The Hedonic Price Applied to Wine

Martínez-Navarro et al. (2024) [27] describe how the literature on wine marketing has evolved through three distinct periods: a first phase of emergence (1990–2009), followed by a phase of growth (2010–2016), and more recently, a phase of rapid expansion (2017–2022). Among the common topics across all these stages, the relationship between quality and price has remained central, only overtaken in impact by themes concerning communication and promotion. Early studies mainly aimed to identify the elements that influence wine prices and to understand how pricing affects consumer choices. In recent phases, the discussion on quality has gone beyond purely economic factors. Researchers have begun to investigate which attributes affect how consumers and experts perceive quality and how these perceptions translate into a willingness to pay a premium price.

In this context, the hedonic approach has been used a lot to investigate how different product features affect wine prices. A recent systematic review of hedonic pricing models applied to wine [28] showed that attribute selection, market definition, data sources, and econometric methods have a significant impact on the explanatory capacity of these models. In fact, selecting the variables to be included in the hedonic function is one of the main empirical challenges [29]. Núñez et al. (2024) [28] propose a detailed hierarchical-deductive classification, which builds on a common distinction in the literature [30,31] between: (I) objective attributes, (II) quality and reputation, and (III) those related to production, consumption, and the economic environment. Each of these categories includes multiple layers. Among objective features, a further distinction is made between observable intrinsic attributes, such as wine type and colour, and attributes not directly observable but shown on the label, including vintage, grape variety, ageing, alcohol content, and environmental certifications. Based on an international sample of more than nine thousand wines sold online in France, Italy, Germany, and Australia, Gonçalves et al. (2021) [32] confirm that vintage and alcohol content are key price drivers across countries and market segments. They also find differentiated effects for colour, blend, bottle closure type, and awards, revealing notable heterogeneity in price determinants across national markets.

The second major category (quality and reputation) is composed mostly of subjective attributes. It includes both wine quality indicators, observable (e.g., brand, quantity produced) or non-observable (expert evaluations, awards, aggregate judgments), and company reputation indicators, which are also distinguished between observable (size or age of the winery) and non-observable (overall rating), up to forms of collective reputation, such as membership in associations or designation of origin. Several studies confirm the relevance of such indicators. Benfratello et al. (2009) [30], Cacchiarelli et al. (2014) [33], Di Vita et al. (2015) [34] and Caracciolo & Furno (2020) [35] show that geographic designations like DOCG, DOC, and IGP are solid price drivers, although their effects differ depending on the market and product segment. Caracciolo & Furno (2020) [35], for example, note that the same attribute may hold different meanings for different consumer groups, while Di Vita et al. (2015) [34] show that certifications tend to generate higher premiums in upper price tiers and are less relevant in the lower end of the market.

As regards non-observable quality indicators, several contributions [31,33] show that expert scores, awards, and recognition contribute positively to price formation, particularly in the mid-to-high segments. However, the role of reputation is broader and more layered: Benfratello et al. (2009) [30] highlight that, in contexts of imperfect information, the reputation consolidated over time can be a more reliable signal than isolated recognitions. Moreover, Boatto et al. (2011) [31] show that the signalling effectiveness of these attributes depends on the distribution channel, being more marked in contexts frequented by less experienced consumers.

Finally, the third group of attributes refers to production, consumption, and the economic environment. It includes factors such as farming and winemaking practices, packaging, conditions of consumption (time, place, pairing), and consumer habits like brand loyalty or deeper motivations linked to place and socio-economic context. In this area, environmental and organic certifications have been frequently investigated, but with controversial results. While environmental certification alone tends to generate a price premium, visible environmental labelling can sometimes negatively affect the product, especially in the presence of low consumer environmental literacy [36]. Abraben et al. (2017) [37] note that the premium for organic is concentrated in the mid-range segments, while in high-end wines it can also translate into a discount.

Despite the emerging focus on sustainability and process quality, as noted by Núñez et al. (2024) [28], the characteristics associated with the production process are still little explored in hedonic models. Often, these practices reduce production yields and require greater labour intensity. Still, they can contribute to the sensory distinctiveness and identity of the product play. They can play a crucial role in the enhancement of the oenological heritage, in the conservation of the rural landscape, and in the promotion of the environmental, economic, and social sustainability of the production areas. The influence of these practices was investigated with respect of the cost or sensory profile of the wine [38,39], but no one, so far, has evaluated the implicit value attributed to the communication of the adoption of cultivation and winemaking practices that strengthen the link between wine and territory. This overlooked dimension offers a promising avenue for further investigation, which the present study aims to address.

1.3. Research Objective

In recent years, attention to sustainable production practices has grown significantly, but the literature still shows important gaps. Few studies systematically analyze how these practices translate into price changes, and even fewer focus on native varieties undergoing repositioning. Contributing to this gap is the difficulty of finding such detailed information through wine labels in traditional distribution channels.

However, the expansion of e-commerce offers an opportunity to fill these gaps, as product sheets and online descriptions make not only traditional quality signals (appellation, vintage, alcohol content) accessible and comparable, but also elements related to sustainability and craftsmanship. As demonstrated by Lynch and Ariely (2000) [40], the reduction in search costs and the greater transparency of information in digital markets systematically influence consumer choices: direct access to credence attributes favours more quality-oriented decisions and reduces price sensitivity.

The choice of Apulian Primitivo as the reference market offers a singular and significant context to examine how attributes related to sustainability and tradition translate into economic recognition. It is, in fact, an autochthonous grape variety, once relegated to blending and now at the centre of a process of qualitative enhancement. Moreover, the structure of its supply chain, composed largely of small and medium-sized enterprises, makes strategies based on such practices particularly relevant, which can represent a more accessible and distinctive way to differentiate oneself on the market compared to solutions that require substantial investments and economies of scale. For these practices to effectively contribute to the sustainable transition of the wine sector, however, they must also be valued by consumers. This remains a fundamental step, since the transition from process sustainability to perceived sustainability is anything but linear: not all attributes are equally recognized, nor do they generate the same impact in terms of willingness to pay.

Considering this evidence, the present study aims to answer the following research questions:

RQ1. To what extent do artisanal and sustainable practices (e.g., low-yield farming systems, use of indigenous yeasts, environmental sustainability mentions, and Geographical Indications) affect the price of Primitivo wine in the e-commerce market?

RQ2. What is the relative weight of these practices in the different market segments, from the mass market to the premium market?

To answer these questions, the study applies a hedonic pricing model to the online marketplace. More specifically, the adoption of quantile regression in an e-commerce context represents a significant methodological innovation, as it allows not only the modelling of average market effects, as in previous hedonic studies, but also the analysis of distributional tails, while exploiting the richness of digital marketplace information, where standardized descriptions reduce information asymmetries. This approach delivers a threefold contribution to the literature: theoretical, by extending wine hedonic models to incorporate previously unexplored sustainability dimensions; methodological, through the application of quantile regression in this field; and practical, because quantifying the economic contribution of individual sustainability and quality practices to the formation of the final price, can provide strategic guidance to producers on investments, denomination policies, and digital communication, as well as policy recommendations for valorising native varieties and supporting the sustainable transition of the entire supply chain.

2. Materials and Methods

As reported by Petrontino et al. (2023) [41], taking up previous studies [42,43], the hedonic pricing model (HPM) considers a product as a set of characteristics/attributes, whereby consumers buy the product that includes this set, maximizing their utility. Similarly, manufacturers maximize their profits by determining the price of a product based on its attributes. Thus, differentiated goods can be considered as a set of various quality attributes that distinguish them from other similar goods, so that the market equilibrium price can be considered a function of the implied prices of each attribute of the good.

However, this theoretical framework is based on robust assumptions: (i) perfect identification of the relevant market; (ii) independence of attributes or related implicit marginal prices; (iii) perfectly competitive markets in long-term equilibrium; (iv) perfect and symmetrical information between consumers and producers [29].

In the wine-growing context, Miller et al. (2007) [44] highlight how these conditions are difficult to achieve. Furthermore, as pointed out by Unwin (1999) [29], Rosen’s framework [42], does not provide operational indications on the functional specification of the price function nor on the selection of the relevant attributes, exposing the analysis to econometric criticalities such as multicollinearity, endogeneity and sample representativeness problems.

Despite these limitations, HPM has established itself as a reference methodology in wine sector studies, by virtue of its ability to estimate the economic value of quality and sustainability attributes incorporated in wines [45]. The literature reports a wide variety of methodological approaches used for the estimation of hedonic price functions. Ordinary least squares (OLS) regression is the most widely used technique, used in 93% of studies and 85% of estimates [28], thanks to the simplicity of interpretation of the coefficients and the availability of correctives for heteroskedasticity and autocorrelation (51% of studies).

The functional form used is another important methodological issue. According to Núñez et al. (2024) [28], 76% of the studies adopt a semi-logarithmic specification, allowing the coefficients to be interpreted as semi-elasticity. Less frequent is the use of the pure logarithmic form (7%) or mixed log-linear models (7%). More complex functional shapes, such as Box–Cox transformations, are residual, mainly used to improve the fit of the model in the presence of strongly asymmetric distributions.

However, the increasing focus on market segmentation and consumer heterogeneity has raised questions about the adequacy of conventional conditional averaging models. As Amédée-Manesme et al. (2020) [46] point out, traditional hedonic models assume homogeneous market behaviour, assuming that the relationship between price and attributes remains constant throughout the price distribution. This assumption risks hiding differentiated dynamics between market segments, generating partial or distorted estimates. In response to these critical issues, quantile regression (QR) is an alternative approach capable of capturing changes in price premiums associated with attributes along the different percentiles of the distribution [34].

Unlike OLS regression, which estimates the average effect of attributes on price, QR allows these effects to be analyzed at different points in the conditional price distribution, explicitly modelling the tails and intermediate segments of the market. The QR is based on minimizing weighted absolute deviations and estimating conditional quantile functions for each percentile q ∈ (0,1), by means of linear programming algorithms [47]. The objective function is represented by:

$$Q\left(\beta q\right)={\sum }_{i:y\ge x_i^{\prime }\beta q}q\left|y_i-x_i^{\prime }\beta q\right|+{\sum }_{i:y<x_i^{\prime }\beta q}\left(1-q\right)\left|y_i-x_i^{\prime }\beta q\right|$$

(1)

where βq represents the estimated coefficients for each quantile [34]. As a semi-parametric estimate, QR is robust compared to heteroskedasticity and outliers, offering more flexibility in analyzing segmented markets [48]. Although the application of QR to the wine sector is still relatively limited, it has proven effective in analyzing reputation awards and market segmentations, as evidenced in previous studies [34,49].

Consistent with this evidence, the present study adopts a quantile regression model to estimate the hedonic price function, in order to quantify the marginal contribution of sustainability and quality attributes in the pricing of Primitivo wine sold online.

2.1. Data Collection and Dataset Construction

The dataset was built through systematic web scraping conducted between September and November 2024 using Web Scraper, a browser-based tool for extracting structured data from e-commerce platforms.

The collection involved exclusively the main pure players in Italian wine e-commerce: Tannico, Vino.com, Callmewine, Xtrawine, and Bernabei. The decision to focus on these specialized platforms responds to a twofold need. On the one hand, these are direct sales channels that play an important role in the Italian market. According to the most recent Mediobanca report [50], pure players represent a market share close to that of producers’ direct sites, with an overall sales growth of +8% in 2024 compared to the previous year. Tannico confirmed its position as market leader (€29 million in revenues), followed by Bernabei (€28.5 million), Vino.com (€27 million), Callmewine (€13.6 million), and Xtrawine (€9.1 million). On the other hand, the choice of the online channel allows access to detailed and standardized product information, reducing informational asymmetries and allowing a homogeneous detection of the technical and commercial characteristics of the wines.

The focus of the survey concerned the labels marketed as Apulian Primitivo. At the end of the collection, comments without essential information (e.g., price) were excluded from the sample, as well as sweet wines (known as “Passito”) or other special versions, as they were associated with specific production practices that could significantly affect the typological reference.

The final dataset includes 656 distinct labels.

Table 1. Description of variables used in our hedonic price model analysis.

Variable	Category	Description	Code
Score Winery	Dichotomous	3.4 ≤ score ≤ 3, 7 = 1; otherwise = 0	Score_A
		3.8 ≤ score ≤ 4, 0 = 1; otherwise = 0	Score_B
		4.1 ≤ score ≤ 4, 3 = 1; otherwise = 0	Score_C
Winery Site	Dichotomous	Apulia = 1; otherwise = 0	CA_PU
Price_75	Continuous	Final price of wine (euro/ 75 cL)	Price per 75 cl
Years	Dichotomous	0 ≤ years ≤ 3 = 1; otherwise = 0	Y0_3
		4 ≤ years ≤ 6 = 1; otherwise = 0	Y4_6
		more than six years = 1; otherwise = 0	Ymore_6
Denomination	Dichotomous	IG = 1; otherwise = 0	IG
		DO = 1; otherwise = 0	DO
Grape Variety	Dichotomous	Primitivo 100% = 1; otherwise = 0	MONO_V
Alcohol	Dichotomous	Alcohol ≤ 14% = 1; otherwise = 0	ALC%14
		14% < alcohol ≤ 15.5% = 1; otherwise = 0	ALC%14–15.5
		Alcohol > 15.5% = 1; otherwise = 0	ALC%15.5
Type	Dichotomous	red = 1; otherwise = 0	RED
		rosè = 1; otherwise = 0	ROSE
Overripening	Dichotomous	overripening = 1; otherwise = 0	TARD
Planting	Dichotomous	apulian bush-trained vine = 1; otherwise = 0	ALB
		other planting system = 1; otherwise = 0	Imp_oth
		not present = 1; otherwise = 0	Imp_no
Refinement	Dichotomous	wood = 1; otherwise = 0	AF_LE
		steel = 1; otherwise = 0	AF_AC
		other material = 1; otherwise = 0	AF_oth
		not present = 1; otherwise = 0	AF_no
Packaging	Dichotomous	product sold in a wooden case or box = 1; otherwise = 0	Pack_ast_cass
Philosophies	Dichotomous	indigenous yeasts = 1; otherwise = 0	IND
		old vines and vines over 25 years old = 1; otherwise = 0	OLDV
Eco_Practices	Dichotomous	Presence of at least one environmental practice (organic, biodynamic, synergic) =1; otherwise = 0	Eco_Practices
Awards	Dichotomous	Single reward > 95 = 1; otherwise = 0	AW_U95
		Multiple rewards, mean >95, SD < 2.5 = 1; otherwise = 0	AW_M95_LSD
		Single reward 90 < score < 95 = 1; otherwise = 0	AW_U90_95
		Multiple rewards, mean 90–95, SD < 2.5 = 1; otherwise = 0	AW_M90_95_LSD
		Multiple rewards, mean 90–95, SD > 2.5 = 1; otherwise = 0	AW_M90_95_HSD
		Single reward < 90 = 1; otherwise = 0	AW_U90
		Multiple rewards, mean < 90, SD < 2.5 = 1; otherwise = 0	AW_M90_LSD
		Multiple rewards, mean < 90, SD > 2.5 = 1; otherwise = 0	AW_M90_HSD

shows the 15 variables identified, divided into four main macro-categories: company information, production characteristics, attributes of product, and awards and recognitions.

The company information includes the score attributed to the winery (based on user evaluations) and the geographical location of the production site, distinguishing between companies located in Apulian region and outside the region. The information relating to production characteristics includes the planting system (bush-trained vines, other forms or undeclared), the method of refinement (wood, steel, other materials or undeclared), the adoption of certified or declared environmental practices (organic, biodynamic or synergistic agriculture), the use of indigenous yeasts, the use of old vines and the possible application of over-ripening techniques of the grapes. The product information concerns the type of wine (red or rosé), the varietal purity (Primitivo 100% or blend), the alcohol content (classified into three categories), the presence of a Geographical Indications (DOCG/DOC or IGT), the year of production and the retail price for a standard 75 cl bottle.

Finally, about the awards and recognitions, the scores assigned by sector guides (e.g., Gambero Rosso, Bibenda, Vitae AIS, VITIS Vitae, Wine Spectator, Mundus Vini) and by experts of international importance (e.g., Parker, Maroni, Suckling), each operating with different evaluation scales and criteria, were considered. In order to harmonize the data, all scores were traced back to a common scale on a basis of 100. For each label, the number of awards received (single or multiple) was recorded. In the presence of multiple scores attributed to the same label, mean and standard deviation were calculated as indicators of consistency between the evaluations from different sources.

2.2. Specifying the Quantile Regression Model

In line with the methodological approach illustrated, the present study applies a quantile regression model aimed at analyzing the contribution of sustainability, quality, and reputation attributes to the determination of the price of Primitivo Pugliese wine in the different market segments. The analysis, in fact, was conducted through the use of the STATA 15 software, on the 10th, 25th, 50th, 75th, and 90th percentiles, in order to capture any differences in the price premiums associated with the attributes along the distribution. The dependent variable is represented by the natural logarithm of the price per 75 cl (ln_Price75). Functional specification by the Box–Cox test suggested a value of λ ≈ −0.71. However, the corresponding transformation implies an excessive compression of the variability at the extremes of the distribution (

Table 2. Comparison of price distributions (€/75 cl).

Statistic	Price_75	ln_Price75	Boxcox_Price75
Mean	19.1	2.75	1.19
Std. deviation	17.09	0.57	0.07
Minimum	4.9	1.59	0.95
1st quartile (Q25)	10.3	2.33	1.14
Median (Q50)	13.63	2.61	1.18
3rd quartile (Q75)	22.59	3.12	1.25
Maximum	187.88	5.24	1.37
Variance	292.18	0.32	0.01
Skewness	4.43	1.13	0.08
Kurtosis	30.87	4.54	2.58

), with the risk of distortion precisely in the upper tail, which is central to our analysis. As Cacchiarelli et al. (2014) [33] point out, in the wine market, the extremes of the price distribution should not be considered mere outliers, but relevant observations that reflect market segmentation and deserve to be analyzed. For this reason, instead of adopting transformations that compress these values, it was considered more appropriate to apply quantile regressions, which make it possible to capture the effects of explanatory variables across the entire conditional distribution of prices, while avoiding parametric assumptions on the distribution of errors. In line with established practice in hedonic wine studies, the logarithmic transformation of the price was therefore adopted, as it ensures econometric robustness, guarantees comparability with the reference literature [34,35,46,51], and does not compromise the consistency of the estimators. On the other hand, all the independent variables, corresponding to the attributes shown in Table 1, have been encoded as dummy variables, assigning a value of 1 in the presence of the attribute and 0 in the case of absence. As reported in the correlation matrix in Supplementary Material (Table S1), no significant correlations were detected among the variables, ensuring that the estimates are not affected by collinearity issues.

The coefficients estimated in the model, therefore, must be interpreted as approximate percentage changes in the price per 75 cl, associated with the presence of the attribute with respect to the reference category. Therefore, for each percentile of the conditional price distribution, the coefficient of a dummy variable represents the percentage increase (or decrease) of the price associated with the presence of the attribute considered, all other characteristics being equal.

3. Results

3.1. Descriptive Statistics

As shown in

Table 3. Descriptive statistics.

Variable	n	%	Mean	Std. Dev.	Min	Max
Price per 75 cl (€)	–	–	19.10	17.09	4.90	187.88
Score_A	20	3.0	0.03	0.17	0	1
Score_B	617	94.1	0.94	0.24	0	1
Score_C	19	2.9	0.03	0.17	0	1
CA_PU	639	97.4	0.97	0.16	0	1
Y0_3	323	49.2	0.49	0.50	0	1
Y4_6	271	41.3	0.41	0.49	0	1
Ymore_6	62	9.5	0.09	0.29	0	1
IG	367	55.9	0.56	0.50	0	1
DO	289	44.1	0.44	0.50	0	1
MONO_V	638	97.3	0.97	0.16	0	1
ALC%14	287	43.8	0.44	0.50	0	1
ALC%14–15.5	303	46.2	0.46	0.50	0	1
ALC%15.5	66	10.06	0.10	0.30	0	1
RED	639	97.4	0.97	0.16	0	1
ROSE	17	2.6	0.03	0.16	0	1
Eco_Practices	81	12.3	0.12	0.33	0	1
TARD	33	5.0	0.05	0.22	0	1
Imp_no	357	54.4	0.54	0.50	0	1
Imp_oth	104	15.9	0.16	0.37	0	1
Imp_ALB	195	29.7	0.30	0.46	0	1
AF_LE	308	46.9	0.47	0.50	0	1
AF_AC	172	26.2	0.26	0.44	0	1
AF_oth	112	17.1	0.17	0.38	0	1
AF_no	64	9.8	0.10	0.30	0	1
IND	53	8.1	0.08	0.27	0	1
OLDV	117	17.8	0.18	0.38	0	1
Pack_no	624	95.1	0.95	0.22	0	1
Pack_ast_cass	32	4.9	0.05	0.22	0	1
No_award	287	43.8	0.44	0.50	0	1
AW_U95	51	7.8	0.08	0.27	0	1
AW_M95_LSD	42	6.4	0.06	0.24	0	1
AW_U90_95	45	6.9	0.07	0.25	0	1
AW_M90_95_LSD	21	3.2	0.03	0.18	0	1
AW_M90_95_HSD	70	10.7	0.11	0.31	0	1
AW_U90	34	5.2	0.05	0.22	0	1
AW_M90_LSD	45	6.9	0.07	0.25	0	1
AW_M90_HSD	61	9.3	0.09	0.29	0	1

, the dataset analyses 656 Primitivo wine labels available online. The average price is €19.10/75 cl, but with considerable variability: it ranges from a minimum of €4.90 to a maximum of €187.88, highlighting a highly heterogeneous commercial offer. Almost all of the wines examined are made up of reds (97.4%) and produced from a single grape variety (97.3%), confirming the monovarietal vocation of this market segment.

The producing companies are mainly located in the Apulian region (97.4%); however, a residual share of Primitivo wines, while declaring itself Apulian, comes from wineries based in other regions. As for online reputation, 94.1% of companies have evaluation scores between 3.8 and 4 out of 5.

Considering the variable ageing time, it emerges that Primitivo with more than six years of ageing represents a commercial niche (9.5%), while labels with ages between zero and three years and between four and six years are almost equally distributed (49.2% and 41.3%, respectively). The alcohol content is also distributed in a similar way: 43.8% of the wines have an alcohol content of less than 14%, while 46.2% are between 14% and 15.5%; only a marginal share (10.6%) exceeds this threshold.

From the point of view of certifications, about 56% of wines fall into the I.G. category, while 44% have a D.O. certification. Ageing in wood is the most common practice (47%), while steel (26%) and other ageing techniques (17%) are less common. However, information on refinement is not always available. In general, the presence of elements about quality or sustainability practices is rather limited: only 12% of labels report environmental practices, and even more limited is the widespread use of indigenous yeasts (8%), the use of grapes from vineyards over 25 years old (18%), or late harvesting (5%).

Training techniques also show a certain variability: about 30% of wines declare the use of the Apulian bush-trained system, while in most cases (54%) the planting system is not specified. Regarding quality awards, 44% of wines are without awards, while the remaining labels report heterogeneous and fragmented awards. Finally, packaging in wooden cases or crates affects just 5% of wines, suggesting that this choice remains limited to a very small market share.

3.2. Quantile Regression Results

This section presents the results of the estimated simultaneous quantile regressions on five quantiles of the price function (Q10, Q25, Q50, Q75, Q90) for 656 labels of Primitivo wine, with the OLS regression used for comparative purposes. The analysis, as can be seen in

Table 4. Results of Quantile regression and OLS.

Variable	Q10		Q25		Q50		Q75		Q90		OLS
	Coeff.	a Std. Err.	Coeff.	a Std. Err.	Coeff.	a Std. Err.	Coeff.	a Std. Err.	Coeff.	a Std. Err.	Coeff.	b Std. Err.
Score_A	−0.1530	0.1137	−0.1218	0.0994	−0.0165	0.1193	0.0056	0.1634	−0.0876	0.2014	−0.0742	0.0865
Score_C	0.1499	0.10248	0.0908	0.1558	0.2709 **	0.1293	0.2992 **	0.1221	0.1944	0.1412	0.1492 *	0.0914
CA_PU	0.0229	0.0927	0.0437	0.0911	0.0955	0.0958	0.0549	0.1567	0.1120	0.2180	0.0824	0.0796
Y0_3	−0.1389 ***	0.0382	−0.1398 ***	0.0410	−0.2083 ***	0.0404	−0.2333 ***	0.0535	−0.2447 ***	0.0710	−0.2150 ***	0.0334
Ymore_6	0.2756 ***	0.0724	0.2878 ***	0.0956	0.3840 ***	0.1030	0.4746 ***	0.1159	0.4949 ***	0.1462	0.3946 ***	0.0693
DO	0.1120 ***	0.0389	0.0591	0.0410	−0.0094	0.0382	−0.0582	0.0558	−0.0264	0.0719	−0.0367	0.0356
Mono_V	−0.1065	0.0798	−0.0953	0.0728	0.0427	0.1002	−0.0289	0.1364	0.0053	0.1259	0.0035	0.0607
ALC%14	−0.2073 ***	0.0359	−0.2358 ***	0.0408	−0.2704 ***	0.0419	−0.2878 ***	0.0489	−0.3391 ***	0.0730	−0.2877 ***	0.0337
ALC%15.5	0.3715 ***	0.1044	0.3538 ***	0.0877	0.4358 ***	0.1051	0.6279 ***	0.1306	0.7352 ***	0.1622	0.5086 ***	0.0713
Red	−0.1529	0.1072	−0.0266	0.0824	−0.0435	0.0996	−0.0269	0.1166	−0.0359	0.1674	0.0100	0.0656
Eco_Practices	0.1461 ***	0.0545	0.0557	0.0493	0.0127	0.0493	−0.0233	0.0643	−0.0405	0.0961	−0.0283	0.0384
TARD	−0.0270	0.0849	−0.0344	0.0917	0.0021	0.0722	0.0222	0.0887	−0.1153	0.1180	−0.0174	0.0719
ALB	−0.0130	0.0484	0.0291	0.0487	~0.0000	0.0426	0.1420 **	0.0674	0.1475 *	0.0841	0.0722 *	0.0421
Imp_oth	−0.0157	0.0520	−0.0112	0.0517	−0.0176	0.0513	−0.0214	0.0537	−0.0683	0.0818	−0.0430	0.0391
AF_LE	0.1192 **	0.0579	0.2151 ***	0.0682	0.1396 *	0.0755	0.0729	0.0821	0.0870	0.1038	0.1606 ***	0.0562
AF_AC	−0.0240	0.0669	0.0582	0.0711	−0.0668	0.0745	−0.1289	0.0882	−0.0387	0.1053	−0.0066	0.0569
AF_oth	0.0871	0.0649	0.1709 **	0.0674	−0.0266	0.0746	−0.0497	0.0963	−0.0130	0.1195	0.0723	0.0596
IND	0.0934	0.0606	0.0635	0.0683	0.0517	0.0924	0.2522 ***	0.0848	0.1990 **	0.0911	0.1067 **	0.0537
OLDV	0.1407 **	0.0543	0.0488	0.0632	0.0411	0.0510	0.0583	0.0848	0.1621 *	0.0851	0.0392	0.0520
Pack_ast_cass	0.0257	0.0941	0.0856	0.1134	0.1623	0.1523	0.3614 **	0.1716	0.3738	0.3177	0.2726 **	0.1088
AW_U95	−0.0697	0.0657	−0.0862	0.0573	−0.0807	0.0573	−0.2296 ***	0.0791	−0.0386	0.1629	−0.1161 **	0.0525
AW_M95_LSD	0.1785 *	0.0950	0.3304 **	0.1350	0.3159 ***	0.0751	0.2423 **	0.1190	0.2128	0.2254	0.3201 ***	0.0807
AW_U90_95	−0.0466	0.0944	0.0138	0.0897	−0.0258	0.0586	−0.0852	0.0826	0.0698	0.1310	−0.0393	0.0616
AW_M90_95_LSD	0.0399	0.1133	0.0882	0.1274	0.0124	0.1292	0.2221	0.2565	0.7359 **	0.3685	0.1753	0.1265
AW_M90_95_HSD	0.0019	0.0648	0.0250	0.0612	0.0603	0.0713	0.0776	0.0898	−0.0356	0.0839	0.0291	0.0503
AW_U90	−0.0401	0.0996	0.0088	0.0716	−0.0689	0.0719	−0.1811 **	0.0770	−0.1928 **	0.0896	−0.0943	0.0606
AW_M90_LSD	−0.0445	0.0599	−0.0046	0.0681	−0.0829	0.0681	−0.1156	0.0747	−0.1235	0.0911	−0.0921 *	0.0540
AW_M90_HSD	0.0921 *	0.0547	0.0794	0.0623	0.1255 **	0.0576	0.0702	0.0737	0.0287	0.0948	0.0858 *	0.0490
_cons	2.5107 ***	0.1684	2.4359 ***	0.1689	2.6741 ***	0.1922	3.0398 ***	0.2765	3.1491 ***	0.3240	2.6741 ***	0.1331
Observations	656
Pseudo R2	0.3145		0.3231		0.3853		0.4362		0.4646
R2											0.5980

Note: * p < 0.10; ** p < 0.05; *** p < 0.01. ^a Bootstrap standard errors in parentheses; Replications = 1000. ^b Robust standard errors in parentheses; OLS = ordinary least squares, included as reference. Confidence intervals are based on the Stata default 95% confidence level. Random samples were drawn using seed 54321 to ensure full reproducibility of the results.

, allows us to examine how observable attributes, relating to winery, production, product, and premiums, differentially influence the price in different market segments. The heterogeneity of effects along the distribution was further verified through the Wald tests on the equality of coefficients (Table S2), which confirm significant differences for some key variables, supporting the validity of the quantile approach.

In relation to the attributes associated with the winery, it should be noted that the geographical location in the Puglia region does not represent a discriminating element in the determination of the price. In contrast, the Score_C shows positive significance from the median (0.2709 **) and upper (0.2992 **) quantiles, as well as in the OLS (0.1492 *). This result highlights how the perceived reputation of the manufacturer is not a uniform factor but emerges above all in the intermediate and upper segments of distribution, while an OLS analysis would have returned a positive average effect without capturing this heterogeneity.

As far as product characteristics are concerned, the vintage and the alcohol content emerge as systematic discriminating variables throughout the price distribution. The most recent vintages (Y0_3) and the lowest alcohol content (ALC% < 14%) show a significantly negative effect in all quantiles, with increasing coefficients along the distribution, while the OLS confirms the negative direction but returns only an average estimate of the effect (−0.2150 *** and −0.2877 ***). On the contrary, prolonged ageing (Y_more6) and high alcohol content (ALC% > 15.5%) are associated with a positive effect, which is also increasing in the different market segments, with consistent results also in OLS (0.3946 *** and 0.5086 ***). These results suggest that regardless of price range, the consumer perceives value in attributes that relate to greater wine structure and complexity.

As regards the focus of our analysis, the results show a differentiated improvement in information regarding production practices, which integrate sustainability, quality and product characterization objectives, along the price distribution. The environmental practices of organic, synergistic and biodynamic farming are significant and positively associated with price only in the first quantile (0.1461 ***), while the OLS does not detect any effect. Similarly, information on the use of wood as a refining material is valued in the medium–low segment (Q10–Q50) and OLS (0.1606 ***), indicating a positive effect that is attenuated in the premium segments. On the contrary, more technical information, such as the adoption of the sapling training system in the vineyard (ALB, significant at Q75 and OLS) and the use of indigenous yeasts for the initiation of fermentations (IND, significant at Q75–Q90 and in OLS), show a significant effect especially in the medium–high segments, indicating that these aspects are perceived as distinctive by more experienced consumers. Finally, the communication relating to the use of old vines has a non-linear effect: it is valued both in entry-level wines (Q10 = 0.1407 **) and in the premium segments (Q90 = 0.1621 *), while the OLS returns a positive but not significant coefficient, confirming the greater sensitivity of quantile regression to grasp these effects.

Finally, with regard to awards, the results show that the possession of multiple awards with high scores (>95), when shared by multiple sources (AW_M95_LSD), represents the main price driver, being significantly associated with a positive effect in the lower-middle and intermediate segments (Q25–Q75) and also in the OLS (0.3201 ***), while it loses relevance in the premium segment (Q90). In the higher segments of distribution, a positive effect emerges for wines awarded scores between 90 and 95 and low disagreement (AW_M90_95_LSD), significant only at Q90 (0.7359 **), suggesting that even lower scores, if shared by multiple sources, can constitute a signal of value in high-end products.

Single premiums, both with high (AW_U95) and medium (AW_U90) scores, do not represent added value, showing significantly negative effects in the medium–high and premium segments (Q75 and Q90), a result also confirmed by the OLS for AW_U95 (−0.1161 **). This evidence indicates that the market rewards the presence of multiple prizes shared between different evaluation sources, which are considered more reliable signals than the assignment of a single prize, even if prestigious.

4. Discussion

The present study analyzed the role of several observable attributes in the determination of the price of Apulian Primitivo wine marketed online, highlighting how the effect of these attributes varies significantly along the price distribution. Simultaneous quantile regressions made it possible to grasp the different sensitivity of consumers to quality signals depending on the market segment. Product attributes, such as high alcohol content and prolonged ageing, are confirmed as price drivers throughout distribution, valued across the board by consumers. The elements related to the winery affect the intermediate segments of the market, where the perceived reputation of the producer acts as a signal of reliability for consumers. Production choices show a differentiated valorization: green practices and Geographical Indications emerge as strategic elements only in the lower segments, while the information related to the adoption of more technical and identity practices (bush-trained system and indigenous yeasts) is rewarded in the medium–high and premium segments. Finally, among the awards, only multiple awards shared between multiple sources are confirmed as price drivers.

With regard to the product characteristics, the results of the analysis are confirmed in the study by Gonçalves et al. (2021) [32], which reveal a systematic appreciation of vintage and alcohol as price determinants for all ranges and markets analyzed (France, Germany, Italy, and Australia). A higher alcohol content and prolonged ageing are rewarding attributes throughout the price distribution, consistent with the peculiarities of Primitivo, a grape variety historically appreciated to produce full-bodied and structured wines, for which a high alcohol content is a consistent indicator of perceived quality.

A second element of interest concerns the reputation of the winery, measured through the score attributed by online users. The result shows a positive and significant effect in the intermediate price segments (Q50 and Q75), suggesting that in these ranges the perceived reputation of the manufacturer represents a trustworthy signal for consumers. Unlike the findings of Di Vita et al. (2015) [34], who emphasized the role of producers’ reputation in the lower market segments, our results indicate that this attribute is valued primarily in the middle range, while losing relevance in the premium tier.

An interesting result concerns the role of Geographical Indications. The results obtained differ from what has emerged in much of the previous literature. Di Vita et al. (2015) [34] show that origin certifications determine an increase in price premiums in the high segments of the market, acting as signals of perceived quality. Chandra and Moschini (2022) [52] confirm their relevance also for New World wines, where the indication of the grape variety has always been considered more influential. Instead, in our case, the designation of origin is valued exclusively in the entry-level segments, while it does not emerge as a discriminating factor in the medium–high price ranges. It can be hypothesized that this result is linked to the specific context of the Apulian Primitivo. Being the subject of a relatively recent phase of re-evaluation and repositioning, it is possible that the DOC does not yet represent, in consumer perception, a signal of sufficient quality to justify a premium price in the higher segments. This result appears consistent with the observation by Panzone & Simões (2009) [53], according to which the ability of Geographical Indications to generate a surcharge depends to a significant extent on the established reputation of the production region. It follows that, to amplify their effects and foster sustainability practices, “bottom-up” actions are needed, embedded in product specifications and governed at the local level [54].

In the case of Primitivo, it can therefore be assumed that the DOC, although formally consolidated, mainly performs a function of guaranteeing minimum quality, enhanced in the lower segments, but not yet recognized as a distinctive element in the premium segments.

Similarly, environmental sustainability practices (organic, biodynamic and synergistic farming) are also valued exclusively in the entry-level segment, thus acting as basic signals rather than differentiating factors in the premium segments This result differs from Fanasch & Frick (2020) [51], according to which these certifications constitute credible signals, capable of generating price premiums; however, it is confirmed by the Abraben et al. (2017) [38] results, which note that the premium price associated with organic practices and certification tends to decrease in the higher quality segments, where such information appears to be overwhelmed by attributes perceived as more qualifying. Consistently, while the organic label enjoys basic recognizability, consumers display limited trust and a lack of awareness regarding the production method [55], making evidence-based communication essential to effectively engage premium segments.

Furthermore, according to Vecchio et al. (2023) [56], consumers’ limited knowledge about sustainable practices, especially biodynamic and natural, can contribute to limiting their signalling effectiveness, especially in markets where there is no clear and standardized communication. The result relating to the enhancement of selected production practices, such as the use of wood in ageing, the sapling system, and indigenous yeasts, associated with a premium price in the medium–high and premium segments, deserves relevance. To our knowledge, no previous hedonic study has specifically analyzed the value attributed to the communication of such practices, despite the fact that Nunez et al. (2024) [28] highlight that elements related to the production process are progressively entering among the attributes considered in the definition of price functions. Yet, as the authors themselves point out, these techniques are generally considered in their function as objective determinants of cost or sensory quality, without attention to their enhancement as communicative tools. Our results suggest that, when properly communicated, the adoption of identity practices, such as the bush-trained planting typical of the Apulian territory or the use of indigenous yeasts, takes on a discriminating role in the higher market segments, probably by their ability to suggest an image of authenticity, traditional craftsmanship and a strong connection to the territory. This market recognition could be an incentive for the continuation and enhancement of traditional production practices, characterized by lower yields but capable of contributing to the conservation of the rural landscape, the protection of biodiversity, and the enhancement of the link between grape variety and terroir. This consideration is consistent with the findings of Petrontino et al. (2022) [57], who analyzed Italian consumers’ preferences for an innovative Apulian wine made with native cultivars and enriched with nutritional information on the label. They confirm that consumers tend to connect the importance of native cultivars to the evocative appeal of the wine-growing landscape. The promotion of typical varieties, accompanied by additional quality information on the label, therefore represents a concrete opportunity for the Apulian wine sector, both in terms of competitive positioning and local income generation. In this perspective, a clearer and more standardized communication of these practices could represent a strategic lever to strengthen the positioning of Primitivo wines in the premium segments, highlighting both the identity value and the sustainable dimension of these production choices. Moreover, as emphasized by Chaminade (2020) [58], place-based actions are paramount to accelerating sustainability transformations, making them a strategic priority for the Apulian wine sector and the premium positioning of Primitivo wines.

Finally, quality awards selectively influence pricing. Only multiple awards, awarded from multiple sources and with high ratings (AW_M95_LSD), represent significant distributed drivers, except for premium wines for which the value of shared awards prevails over more moderate scores (AW_M90_95_LSD). On the contrary, a single award or recognition, even with high scores, is penalizing in the higher segments. This evidence helps to explain the variability of the empirical results observed in the literature, where the impact of expert evaluations on prices is positive but not unequivocal. Even if meta-analyses [48,49] confirm an average positive effect of quality assessments, several studies have shown heterogeneous effects, more marked in red wines, in “primeur” products or in the presence of consensus opinions by several experts [59]. Increasing heterogeneity in assessment methods, sensory paradigms, and scoring scales could contribute to the variability of observed empirical results [28]. In this context, our results confirm the importance of consistency between evaluation sources as an element of strengthening the signal of perceived quality.

In conclusion, the results suggest that in the online market for Apulian Primitivo, the enhancement of observable attributes follows a stratified logic: in premium segments, specialized and reputational signals are prioritized, while in lower segments, more standardized communicative attributes prevail. Communication strategies operate as key signals convert technical attributes into perceived value and purchase intention across distinct market segments, reinforcing the promotion of sustainability and quality. In particular, potential solutions include the adoption of Quality Assurance Programs (QAP) [60], which—if effectively communicated—can establish recognizable cues that amplify the value of sustainability, especially among price-sensitive segments and for attributes that are otherwise difficult to interpret. Alternatively, brand-building and e-commerce co-creation initiatives [61] can leverage digital storytelling to elevate the visibility and appeal of identity- and reputation-based practices in premium segments, thereby consolidating the layered segmentation logic observed.

Limitations and Future Research

This study presents some limitations that may offer directions for future research. First, it focuses exclusively on Apulian Primitivo labels sold through e-commerce, excluding traditional mass-market channels and specialized wine shops. A future extension of the sample to include different sales channels could not only enhance the representativeness of the analysis but also enable comparisons between value dynamics across distribution contexts.

Secondly, unlike surveys that use historical data series [37,46,51], the cross-sectional nature of the dataset does not allow for capturing any changes over time in the perception of the analyzed attributes. Considering that Primitivo is going through a phase of commercial repositioning, it would be interesting to understand if and how the value attributed to practices such as identity or sustainable practices may change over time.

A third consideration can be developed regarding the attributes considered in the model. Although they cover important aspects related to intrinsic quality, sustainability, and reputation, they cannot include some dimensions of perceived value already recognized in the literature. As also highlighted by recent review about wine hedonic pricing studies [28], several extrinsic and intrinsic variables have been identified as price determinants in the hedonic models applied to wine and, among these, there are characteristics not included in the data collection of the present survey, such as: the sensory and chemical aspects of the product, the climatic and pedological conditions of the territory of origin, the exclusivity of the reference (e.g., reserves or limited editions), the quantity produced, the context and occasion of consumption, and the size of the company.

Finally, although the hedonic approach is useful in analyzing the relationship between observable attributes and price, it does not allow for direct investigation of individual preferences and cognitive processes underlying purchasing decisions, especially considering that some sustainability variables introduced as dummies have not yet been certified or validated through direct comparison with consumers. In this regard, integrating stated preference method, such as choice experiments or willingness-to-pay (WTP) analysis, could provide a more comprehensive understanding of the value consumers attribute to different production and communication practices.

5. Conclusions

This study examined how quality, sustainability, and winemaking identity influence the price of Apulian Primitivo wine sold online. The application of a quantile hedonic model made it possible to analyze market segmentation, revealing how the value attributed to the various aspects examined changes along the price distribution.

The results provide a joint answer to the two research questions. They show that artisanal and sustainable practices affect the price of Primitivo wines marketed online (RQ1), but that their relative weight varies depending on the market segment (RQ2). In detail, traditional product attributes, such as alcohol content and ageing, are universally recognized signs of quality. On the other hand, environmental practices and Geographical Indications serve as basic signals, reassuring consumers in the cheaper segments, but fail to differentiate the product in premium markets. It is precisely in these segments that the distinctive value of identity attributes is noticed, such as bush-trained system farming and the use of indigenous yeasts, which represent a local production tradition perceived as a sign of authenticity and craftsmanship.

From a strategic point of view, the Apulian Primitivo seems to have a communication that is still too focused on certifications and generic sustainability statements. Producers could instead give greater prominence to identity practices as a lever of differentiation in high-spending segments, also through a clearer and more standardized narrative of artisanal practices. In addition, strengthening the collective reputation of DOC Primitivo could help consolidate the appellation’s role as a premium brand.

To conclude, this work contributes to expanding hedonic models on wine by introducing variables related to sustainability and production identity, and it demonstrates the usefulness of quantile regression to analyze the heterogeneity of the online market. More generally, the results also offer insights into the role of the digital economy in sustainable development, showing how the greater transparency of information in e-commerce channels can affect the enhancement of production practices and, indirectly, purchasing behaviour.

The results suggest that sustainability, if not linked to a recognizable territorial identity, risks being seen as a basic requirement rather than a distinguishing factor. The real competitive key for Primitivo lies in the ability to combine sustainability and territorial identity in a cohesive narrative, capable of attracting the most experienced and quality-conscious consumers.