Direct Sales Approaches, Visitors, and Profitability of Agritourism Operations in the U.S.

Abstract

This paper empirically investigates the influence of specific direct sales approaches in attracting visitors to an agritourism operation and its profitability using survey data from the U.S. This study further examines the mediating role of the number of visits to a farm in the relationships between specific direct sales approaches and profitability. Agritourism operations enhance economic viability and sustain the business by opening farms to visitors for education, recreation, entertainment, and direct sales of farm products and services. The goal is to invite visitors to a farm and enhance income. Previous studies in the U.S. show that on-farm direct sales, in general, show a positive association, whereas off-farm direct sales show a negative association with the profitability of agritourism operations, along with many other factors. Farmers consider U-pick, sales through a farm stand/store, and subscription farming or community-supported agriculture (CSA) (on-farm pick-up) as on-farm, and CSA (off-farm delivery) and selling at a farmers’ market as off-farm direct sales approaches. However, which specific approach attracts visitors to a farm and generates profitability is not known. Multivariate analysis using the recently collected data from a U.S. national survey of operators reveals that on-farm direct sales such as a U-pick and a farm stand/store attracted significantly more visits to an agritourism operation, which ultimately yielded higher profitability. In contrast, the selling of produce at farmers’ markets attracted significantly fewer visits to the farm and reportedly reduced profitability. These results are adjusted for other factors including various agritourism experiences offered to the visitors. Moreover, as theoretically expected, the number of visits mediated the effects of specific direct sales (particularly a U-pick and farm stand sales) on profitability. This evidence has implications for agritourism operators, policymakers, and extension educators engaged in starting, expanding, and promoting direct sales via agritourism operations for their economic viability and sustainability.

1. Introduction

Agritourism, adding tourism to agriculture, has been an important source of diversifying incomes for many small and medium-sized U.S. farmers to remain competitive in farming (Bhandari et al., 2024; Bowen et al., 1991; Chase et al., 2018; Hollas et al., 2021). Agritourism operators open their farms and invite visitors to offer various experiences and attractions, such as products and services, education, active involvement on the farm, recreation and entertainment (Chase et al., 2018). In exchange, these farmers benefit from directly selling their products and services to visitors (Hollas et al., 2021; Tew & Barbieri, 2012). Thus, an agritourism owner’s/operator’s goal is to increase income by attracting more visitors and selling products and services to them (Barbieri, 2009; Barbieri et al., 2008; Brandth & Haugen, 2011; Tew & Barbieri, 2012).

Conceptually, the number of farm visits is one of the keys to securing a higher income for agritourism operations (Bhandari et al., 2024). Offering various attractions at the farm increases visitors’ flow as they enjoy the benefits of attractions. These attractions may range from direct sales of a farm’s fresh and value-added produce to education, experiential experiences and classes, and enjoyment from festivals and entertainments presented at the farm. The increased flow of visitors will increase direct sales, resulting in higher profits without competing with the large farms in the open market. Thus, an operation also generates additional income from visitor fees, fees for educational classes, entertainment and events, and experiential hands-on activities. Thus, any activity that attracts visitors to a farm is expected to enhance revenue, and thus, the profitability from agritourism. Direct sales, the selling of products and services directly to consumers, is one of the important marketing approaches used mostly by small farmers to boost their profitability (Schmidt et al., 2023). Direct-to-consumer sales create short supply chains and help increase revenue by reducing marketing costs associated with packaging, handling, and transportation costs (Azima & Mundler, 2022). According to the 2022 U.S. Census of Agriculture, the trend of selling produce directly to consumers is growing over the years. The economic value of food sold directly to consumers was estimated at 3,263,074 thousand dollars in 2022, which increased from 2,805,310 thousand dollars reported in 2017. Similarly, the number of farms that sold food directly to consumers increased from 21,570 to 27,981 in 2022. This is the value of edible products, which include value-added products sold directly to consumers at farmers’ markets, on-farm stores or farm stands, roadside stands or stores, U-pick, CSA (community-supported agriculture), and online marketplaces (National Agricultural Statistics Services, 2022).

There are several approaches to direct sales. Some of the commonly used approaches are selling produce at the farm through U-pick (or also known as pick-your-own, PYO), a farm stand or a farm store, adopting a community-supported agriculture (CSA) or subscription farming, online marketing, and selling at farmers’ markets (Chase et al., 2018; Schmidt et al., 2023). These techniques are commonly referred to as on-farm direct sales as the customers purchase various products and services directly from the seller at the farm. On the other hand, selling produce at the farmers’ market, through online marketing, and CSA, referred to as off-farm direct sales, are designed to pick up deliveries off the farm. Thus, these varying types of direct sales approaches may have differing levels of visitor attractions to the farm and, therefore, will have differential links with profitability.

Enhancing profitability though a diversification of revenue sources to agritourism is not free of challenges. Diversification of activities depends on factors such as the access to suitable land for various activities and the availability of human resources to manage the farm and visitors. The location of the farm is equally crucial. Moreover, revenue generation also depends on its capacity to attract, accommodate, and manage visitors. Producers should seriously consider the local zoning and permitting requirements, liability issues, demand for activities, reactions from neighbors and communities about the visitor crowds, and vehicle movements. On top of this, they should also understand heterogeneity in visitor motives and their spending behaviors. Not all the visitors visit a farm to purchase farm produce. They often visit agritourism operations for other experiences such as for education, recreation, entertainment, and learning through hands-on experiences. Some of them visit to spend a night(s) in a farm stay, for recreation, for enjoyment or entertainment, and or for gaining first-hand experiences of various farm activities.

A previous study using the survey data from Maryland, however, did not reveal the direct significance of the various direct sales approaches, such as selling produce at a farm market and offering U-pick, on the number of visitors and on the profitability (Bhandari et al., 2024). Although the relationships were positive initially, the significance was lost after controlling several other factors, particularly the number of employees and the length of establishment, perhaps suggesting that the number of helping hands and the goodwill developed over time could have influenced the number of visits and thus the profitability. This study, however, was based on data from a small number of cases in one state of the U.S., and the various direct sales approaches were not clearly defined and measured.

Another study using the national large-scale survey data of agritourism operators in the U.S., however, revealed that on-farm direct sales, in general, show a positive and independent association with the profitability of an agritourism operation (Hollas et al., 2021). On the other hand, farmers who sold their produce directly to customers off the farm, indeed, reported a loss, which is contrary to the expectation. These scholars consider both CSA off-farm delivery and selling at farmers’ markets together. They point towards draining of resources, e.g., the laborers away from farms, and also opine that perhaps the off-farm sales are geared more towards promoting products and the farm as a whole. They also found that the number of visits did not significantly contribute to profitability, adjusting for other factors. Meanwhile, they believe farms open longer invite more visitors, often not charging any fees. However, we believe it could be because this study utilized the raw number of visits to explain profitability, with a highly skewed distribution. Moreover, this study did not examine the role of on-farm or off-farm direct sales on the number of visits to a farm. This study also did not examine the mediating role of visitors in the relationship between the direct sales approaches used and profitability. Hollas and colleagues found that other factors, such as a farm receiving visitors from over 50 miles, owner’s experience, entertainment on the farm, and several motivation factors, significantly and positively increased profitability. On the other hand, being a female operator significantly reduced profitability, net of other factors.

Farmers invite visitors to their farms and sell their products and services to customers through a variety of direct sales approaches such as U-pick, at a farm stand/store, subscription farming, community-supported agriculture (CSA), and at a farmers’ market. In this study, we argue that direct sales approaches should influence the number of visitors, which ultimately should influence profitability through increased direct sales. Moreover, we also argue that the relationships between direct sales, the number of visitors, and profitability depend on specific on-farm and off-farm direct sales approaches, and therefore, should not be considered together. Considering the background, this paper answers the following: Does a specific approach of (on-farm or off-farm) direct sales invite significantly more or less visits to a farm, independent of other factors? Does the profitability reported by agritourism operators vary based on the types of direct sales approaches used? Does the number of farm visits mediate the relationship between (on-farm and/or off-farm) direct sales and profitability?

This study is important primarily because attracting visitors to promote direct sales is the goal of each agritourism operation, to enhance profitability by selling products and services to them. In this way, agritourism operations reduce the marketing costs associated with intermediaries and increase their marketing margins. This is the reason why agritourism operations diversify their income sources. They make profits by directly selling their products, such as wine, beer, and fresh produce (e.g., strawberries, cherries, apples, blueberries, honey and other value-added products), by charging registration fees for events, classes and hands-on training, and demonstrations, by organizing entertainment events such as festivals), and more. More importantly, these operations add extra dollars to their pocket by avoiding going through intermediaries or other supply chain channels. We further expand the evidence by investigating which specific approach of on-farm direct sales, such as a U-pick and farm stand/store, community-supported agriculture (CSA) pick-up, or off-the-farm direct sales, such as through farmers’ markets or community-supported agriculture (CSA) delivery, attracts more visitors to a farm and whether or not the number of visits results in higher profitability, by theoretically and empirically testing the mediating role of the number of visits in the relationships between the direct sales approaches and profitability. This is the significant contribution of this study. Unlike the previous studies (Bhandari et al., 2024; Hollas et al., 2021), first, we investigate the influence of specific direct sales approaches on the number of visits, and then, on the profitability, while controlling for many other theoretically important confounders.

To investigate the issue, we used the national survey of agritourism operators’ data (n = 1263) from the U.S. The results revealed that, on average, a farm operation received slightly over 7800 (5.95 logged) visits, as self-reported, with a skewed distribution ranging from a minimum of 1 to a maximum of 1,300,000 visits. Nearly one-quarter (24.6%) of the operations reported that they incurred no profit, or incurred a loss, and 12% of them reported a net profit of less than $1000. Nearly 28% of them reported a profit between $1000 and $9999, another 28% reported a net profit between $10,000 and $100,000, and the remaining 7.6% reported a net profit of $100,000 or more.

Twenty-nine percent of the operations reported that they offered a U-pick to the visitors. Slightly over half (57%) of them reported they had a farm stand or a farm store. Only 13% of them revealed that they had on-farm community-supported agriculture (CSA)—the members picked up their delivery from the farm. Twenty-nine percent of them reported sales through a farmers’ market, and nearly 10% reported that they delivered produce to members at various locations to their CSA subscribers.

Our findings from the multivariate analysis show that a net of controls, a U-pick and farm stand/farm store sales significantly and positively attracted more visits than those that did not use these approaches. On the other hand, those who sold their produce at the farmers’ market significantly reduced the number of visits to the farm. However, those operators who subscribed their produce through community-supported agriculture (CSA) to the customers who picked up at the farm or through off-site delivery were not significantly associated with the number of visits. In addition, while the U-pick and farm store sales significantly increased profitability, these effects were lost when the number of visits was adjusted, perhaps suggesting the mediating role of the number of visits in the relationship between direct sales approaches and profitability. Indeed, the further analysis confirmed the evidence of a mediating effect. These findings have implications for agritourism operators, policymakers, and extension educators engaged in beginning, expanding, and promoting agritourism to enhance the economic viability and sustainability of farms.

2. Empirical Reviews and the Conceptual Framework

Agritourism operators open their farms to invite visitors and offer diversified attractions or activities, such as the sale of products and services, education, active involvement on the farm, recreation, and entertainment to supplement their incomes (Chase et al., 2018). While several studies have examined various factors contributing to an agritourism operation’s profitability and performance (Bhandari et al., 2024; Hollas et al., 2021; Jin et al., 2022; Lucha et al., 2016), factors including the offering of various direct sales approaches, which influence visits to an agritourism operation, are scant. Because the goal of an agritourism operation is to increase profitability by attracting visitors, it is important to understand factors that influence visits to the farm.

Abelló and colleagues examined factors that were associated with the number of visits to farmers’ markets (Abelló et al., 2014). Based on the information collected from the visitors, the travel distance, market promotions such as entertainment and education, food events, and other factors such as the age and education of operators significantly influenced the frequency of their visits to farmers’ markets. Other scholars reported the significance of the quality and freshness of the product to attract consumers (Brown, 2003; Wolf et al., 2005). While these factors are not directly related to visits to an agritourism farm, similar variables may have a role in attracting visitors to an agritourism farm. Below, we provide information about how direct sales may have an influence on inviting visitors and profitability, net of all other factors, to a farm. Then, we provide a conceptual framework and hypotheses used in this study.

2.1. Direct Sales Approaches

Selling products and services directly to consumers, also referred to as ‘direct sales’, is a common marketing approach used by agritourism operators (Schmidt et al., 2023). The producers or farmers invite visitors to the farm and sell products and services directly to them (Schmidt et al., 2023; Simona et al., 2022). This approach offers one of the shortest paths to reach consumers and increases profit margins for the producers by bypassing intermediaries and receiving the full consumer dollar share by saving direct and indirect costs accrued by various marketing channels in the value chain. Producers or sellers have control over the sales process and distribution and have flexibility in pricing, promotions, and product offerings. More importantly, these sales approaches allow direct interaction between a seller and a buyer.

Agritourism operators commonly organize a U-pick (pick-your-own, PYO) or sell products at a farm stand or a farm store, adopt community-supported agriculture (CSA) or subscription farming, and sell produce online and at farmers’ markets (Chase et al., 2018; Schmidt et al., 2023; Veeck et al., 2006, 2016). Farmers who produce fresh fruits and vegetables on their farms organize U-pick, where customers or visitors harvest the type and quantity of produce of their interest by themselves. Producers offer various options to pickers or visitors. For example, some of them provide various sizes of boxes or containers to the visitors with a fixed price, and visitors pick as much as they want. In other cases, a per-unit price is fixed and pickers pay the price for the amount they have harvested. A U-pick has been a popular method of direct sales that invites visitors to their farms and allows visitors to self-experience picking fresh products. While producers save labor costs, they may lose produce as pickers may eat produce (such as blueberries, strawberries) or damage crops by over-picking or picking immature produce as waste. In one of the agritourism traveling workshops organized by the UMES Extension, a host farmer stopped their U-pick operation for fruits (apples and peaches) due to a significant amount of damage and switched to producing value-added products and selling fresh produce (fruits, peaches) at farmers’ markets, through CSA, and with pre-order online sales for pick-up. The owner–operator reported that “Some people would pick too much and then dump them,” and he was not keen on losing products (Stephens, 2024).

Other farmers sell produce at the farm stand or at a roadside stand, such as a table, stall, or tent nearby their farm. Other farmers sell at the farm store, where they keep commodities including their own produce. The opportunity to buy off-the-farm commodities invites more visitors, and the sellers also benefit from selling products, including their own.

Another growing approach of direct sales is through subscription farming or community-supported agriculture (CSA). CSA is a community of individuals or families who pledge support (subscribe) to a farm operation so that the farmland becomes the community’s farm. Both the growers and consumers provide mutual support and share the risks and benefits of food production locally. CSA members pay subscriptions upfront to the producers, providing regular income support to the producers for continued investment to produce. The subscription acts as a binding contract to both producers and consumers for their commitment to buy produce from the farm or to supply produce to the consumers on a timely basis. Thus, CSAs are advantageous to both producers and consumers (Myers, 2010; Roos, 2025; Woods et al., 2017). While some subscriptions involve members coming to the farm to pick their subscription, other farmers deliver produce to a different location, and members pick up their produce from an off-farm location without visiting a subscribing farm. While profit-making and sustaining farms are the important goals, CSAs are also important in building community connections, developing relationships among producers and consumers, and alleviating food and environmental security.

Farmers’ markets have been the most common platforms, where farmers sell their produce, such as fruits, vegetables, crafts, baked goods and more, at a local marketplace. The marketplace provides space for several (two or more) vendors that sell produce directly to customers. Farmer sellers transport their commodities to the marketplace on the day the marketplace is open and sell their produce. Thus, visitors buying produce at the farmers’ markets do not necessarily require visiting a farm. However, this is an important venue to promote farms and encourage visitors to come to their farms. Selling at farmers’ markets also helps build networks among producers and customers.

Farmers are increasingly using online portals to sell their produce to customers (O’Hara & Low, 2020). Farmers who are tech-friendly use methods such as Facebook, e-mail, and websites for business marketing and selling products. According to O’Hara and Low, the online marketplaces promote farms to directly compete with customers, making their products and services available to the customers on the internet, which reduces the costs for search and transportation as compared to in-person direct-to-consumer (DTC) transactions. While an online marketplace can be a strategy for income diversification, this will be less desirable for agritourism farms that have a goal of attracting visitors to their farms. However, such marketplaces will be useful to customers who are distantly located.

Direct marketing techniques can be grouped as on-farm and off-the-farm direct sales. A U-pick, a farm stand/farm store, and CSA (on-farm pick-up) are grouped as on-farm direct sales. On-farm direct sales are designed to attract visitors and offer visitors the first-hand experience of picking produce as they like. Such a U-pick operation offers fresh products such as vegetables, fruits, flowers, and even pumpkins as per the demand from consumers or pickers. More importantly, a U-pick is designed to offer the real-world hands-on experience to visitors. Similarly, a farm stand or a farm store also attracts visitors to the farm and sells them fresh produce, often healthy and environmentally friendly products such as free range meat and eggs, milk, value-added products such as jam, jelly, pickles, and tomato sauces unique to the local setting. Those visitors who promote local produce often are motivated to buy such produce directly from the farm—U-pick, a stand or a farm store, and through CSA subscription. On the other hand, selling produce at the farmers’ market, online marketing, and CSA, designed to pick up deliveries off-farm, are known as off-the-farm direct sales. These off-the-farm direct sales approaches do not directly encourage visitors to visit the producers’ farms, but may influence them to visit farms at a later date. Thus, these varying types of direct sales approaches clearly have differing levels of visitor attraction. However, it is not clear whether they will have a differential link to profitability.

2.2. The Conceptual Framework

The number of visitors to a farm is a key to increasing income for an agritourism operation. Although the number of visits per se does not result in profitability, it is the number of visitors that purchase products and services that results in an increased volume of purchases and increased revenue collection. In addition, the revenue from farming also increases through visitor fees. Various products for direct sales could be fresh produce, such as fruits, vegetables, and flowers, and value-added products, such as processed honey (raw honey is not a value-added product), processed meat, wine and beer (only if fermented and brewed on the farm). Similarly, an operation may generate extra income if it charges any fee for the visitors. Some operations offer educational classes or courses to visitors for a fee. Some agritourism operations with a farm stay benefit from hospitality services such as lodging and food. Additionally, those operations that organize festivals and entertainment events generate income by selling tickets to visitors.

Direct-to-consumer sales create short supply chains, offer an opportunity for producer farmers to set their own price, and reduce the costs associated with middlemen (Azima & Mundler, 2022). Direct sales also encourage consumers to visit local farms and purchase fresh foods. Thus, it is theoretically expected and hypothesized that any activity that offers on-the-farm direct sales will attract more visitors to a farm. An increased number of visitors will likely increase direct sales, which is expected to positively influence the profitability, adjusting for other potential confounders. The profitability through enhanced direct sales increases as a result of reduced marketing costs and increased market margins due to the absence of intermediaries (Azima & Mundler, 2022). According to Paul (2019), direct farm sales lead to better economic outcomes, higher prices, and enhanced revenues by selling farm products directly to consumers. Thus, overall, it is hypothesized that the agritourism operations with on-the-farm direct sales, such as U-pick or sales through a farm stand/store, or selling produce through the subscription of community-supported agriculture (CSA), will (a) attract more visitors to the farm, and (b) increase the profitability of a farm. On-the-farm direct sales will have a complementary effect on profitability. While direct sales are expected to increase revenue, and thus the profitability through reduced marketing costs and increased market margins in the absence of middlemen, they also increase revenue by increasing the number of visitors, which produces an increased volume of sales. Thus, theoretically, we expect that the number of visitors will have a mediating (both direct and indirect) effect on profitability through direct sales (e.g., direct sales approaches → number of visitors → profitability) (Preacher & Hayes, 2004). Conversely, those operations that sell off-the-farm products, such as at farmers’ markets or CSAs, which deliver food to other locations off the farms, clearly discourage the number of visits to the farm. However, whether there will be any direct or indirect influence on profitability is not known, because the aim of selling produce through CSA or at the farmers’ markets is also to increase revenue.

We control the effects of many other confounders that potentially influence the number of visits and ultimately profitability. These measures are the length of operation, rural or urban location, offerings of other services, such as a class, hands-on skills, and recreation through organizing festivals and entertainment events, the number of days of operation, and the season of operation, which potentially influence the number of visits to the farm (Barbieri et al., 2008; Bhandari et al., 2024; Hollas et al., 2021). Theoretically, the length of operation provides an operator the opportunity to learn and improve from the past weaknesses and improve visitor-friendly products and services (Bhandari et al., 2024; Hollas et al., 2021; Nickerson et al., 2001). This will likely affect the number of visits because of experience of handling and managing products, services, and visitors, reputation (goodwill), and the accumulation of greater assets for investment (Barbieri & Mshenga, 2008). Barbieri and her colleagues found that the length of operation significantly increased farm income, which is likely due to the increased number of visitors. Moreover, the numbers and types of offerings, marketing, and promotion play equally important roles in inviting visitors to a farm and in enhancing the profitability of an agritourism farm.

The number of staff members available at the farm could be another factor, as human resources are important for managing and delivering products and services to visitors. The Agritourism and On-Farm Direct Sales Survey reported that labor shortage was among the top two challenges facing agritourism operations in the U.S. (Chase et al., 2021). Of the total respondents, 89% reported that labor was somewhat or very challenging. The Maryland Agritourism Operator’s Survey also revealed a labor shortage as a farm management problem (Ejiogu et al., 2023). Barbieri and Mshenga (2008) and Bhandari et al. (2024) provide evidence that the number of employees on a farm significantly and positively influences an agritourism operation’s performance. Thus, if a farm has enough staff to efficiently and effectively manage and deliver products and services on time, it is expected to attract more visitors. The distance to the farm from the more extensive population base, the source of visitors, is another important factor (Hollas et al., 2021). A convenient and nearby location for an agritourism operation for people will attract more visitors. With this, a farm located in a rural area may attract fewer visitors due to limited access to many visitors and amenities and a longer travel distance, unlike in urban areas.

3. Materials and Methods

3.1. Data

This paper used the National Agritourism and On-farm Direct Sales Survey data collected by a group of scholars from the University of Vermont, Vermont, VT; Oregon State University, Corvallis, OR; the University of California, Davis, CA; and West Virginia University, Morgantown, WV in 2020 (Chase et al., 2021). This is a national survey of farms that were open to visitors for product sales and/or experiences. The survey was administered online, starting in November 2019 and ending in February 2020. A link to the survey was shared among farmers and ranchers through email, social media, and newsletters throughout the U.S.

First, the survey confirmed by asking the very first question, “Do you have visitors on your farm or ranch (paid or unpaid)? Examples could include farmstands, U-pick, CSA, tours, overnight stays, events, hunting, and any other experiences that bring visitors to your farm/ranch.” This study utilized data from those farms that answered ‘Yes’ to this question. The survey was responded to by 1834 operations from all 50 U.S. states. These operations reported that they invited visitors to their farms and or ranches for various reasons, such as to buy products from their farmstands or for U-pick, community-supported agriculture (CSA), tours, overnight stays, events, hunting, and to have any other experiences at the operation. For more details, refer to Chase et al. (2021); Hollas et al. (2021).

In this study, of the total 1834 cases interviewed, excluding 571 cases with missing values, the remaining 1263 cases with valid responses were included in the analysis. While the number of missing cases raises a concern of sample bias, this is the only national survey from the U.S. However, to gain confidence in our results, we broadly explored if the missing cases were random or systematic by comparing mean differences in certain important variables, such as the net income, number of visits, land size, number of days open, and length of operation. None of the mean differences between those cases with missing values and those used in the analysis were statistically significant, providing us the confidence in our results that the missing cases were random rather than systematic. A significant mean difference between the two groups could have excluded a significant group of cases from the data, resulting in biased results.

3.2. Measures

3.2.1. Outcome Measures

There are two outcomes: (a) the number of visits, and (b) the net income from agritourism, a measure of profitability. The first measure is the number of visits to the farm. It was measured by asking, “Approximately how many visits (paid and unpaid) took place on your farm/ranch in 2018? Count the number of visits, not visitors, so that one person who visited 10 times in 2018 would be 10. A tour bus of 50 people would be 50 visits.” Respondents provided an estimated number of visits. As the range of the number of visitors greatly varied (with a minimum of 1 and a maximum of 1,300,000) and had a highly skewed distribution, this outcome variable was transformed using the natural log to normalize the data.

The second outcome, the self-reported net profitability, was measured by asking, “How much profit (net income) do you estimate your agritourism enterprise(s) generated in 2018?” Several options were provided, including the farm did not make a profit or operated at a loss. The response was measured on an ordinal scale (in categories). Because the frequency distribution for a few categories was small, we collapsed those categories.

3.2.2. Explanatory Measures

The survey first asked if the farm/ranch offered experiences to visitors in 2018. There were 6 categories of experiences offered: (1) on-farm direct sales, (2) accommodation and lodging, (3) education, (4) entertainment/events, (5) outdoor recreation, and (6) off-farm direct sales. If a farm/ranch provided experiences other than the six indicated above, they were asked to respond as other. In this study, we included cases that reported either on-farm direct sales or off-farm direct sales.

If a farm/ranch reported on-farm direct sales offered in 2018, they were further asked if they offered (a) a U-pick, (b) a farm stand/farm store, (c) community-supported agriculture (CSA) on-farm pickup, and/or (d) other. Similarly, if they offered off-farm direct sales, they were asked if they (a) sold their produce at a farmers’ market and/or (b) provided community-supported agriculture (CSA) delivery. The responses were recorded as offering an experience (coded 1) or did not offer an experience (coded 0).

3.2.3. Controls

Several factors may influence the number of visits to a farm and its profitability. Thus, the following potentially confounding factors were controlled in the analysis, and their measurement is discussed below.

Offering of various experiences, such as accommodation and lodging, education, entertainment/events, and outdoor recreation, other than those that offered direct sales (on-farm and/or off-farm), also directly influenced the number of visits to and profitability of an agritourism farm/ranch. Whether an agritourism operation offered these experiences—accommodation and lodging, education, entertainment/events, and outdoor recreation—was controlled in the analysis. The responses for each category of experience type were recorded as offering an experience (coded 1) or did not offer an experience (coded 0).

The length of operation is measured by asking, “What year did you begin offering agritourism, including on-farm direct sales?” The response was recorded as the year of offering agritourism. Since the data was collected in 2019, the year was subtracted from 2019 to get the number of years of operation. The next factor controlled is the total number of days per year the farm/ranch was open to visitors, measured by asking, “About how many days per year is your farm/ranch operation open to visitors?” In addition, we also used whether the farm/ranch operated all year (all four seasons, coded 1) or for only part of the season (coded 0).

The percentage of visitors who travelled 50 miles or more was another variable controlled. This item was measured by asking, “Approximately what percentage of these visits were from people who traveled 50 miles or more (one-way) from their homes?” The responses were guestimates and measured in percentages. A farm’s distance from a city of at least 50,000 people (miles) is one of the common measures influencing the number of visits and profitability. This item was measured by asking, “How far is your farm/ranch from a city of at least 50,000 people?” The responses were recorded as (i) located in a city with a population of 50,000 or more, (ii) less than 5 miles (a reference category), (iii) 5–9 miles, (iv) 10–29 miles, (v) 30–49 miles, and (vi) 50 miles or more. Considering the distribution of responses, the following groups were created: (i) less than 10 miles, (ii) 10–29 miles, (iii) 30–49 miles, and (iv) 50 miles or more. The next variable is the geographic region of the operation, measured as located in the (a) Northeast, (b) Midwest, (c) Southern, or (d) Western. The Northeast cluster was used as the reference category.

4. Analysis

We used version 29.0.0.2.0(20) of SPSS for data analysis. Descriptive statistics of the measures from 1263 agritourism operations used in this study were calculated. We provided frequencies, percentages, means, and standard deviations where appropriate. We also examined Finally, as the dependent variables of interest were scale variables, we used the multiple linear regression (ordinary least squares, both log-linear and linear log models) technique to examine the associations between the outcomes and the variables of interest (Yang, 2023). Since the first outcome variable, the number of visits, was log-transformed, we used the log-linear model (Equation (1)). We used the linear log model for the second outcome, as the outcome was considered a continuous (actually an ordinal) measure and one of the predictors (the number of visits) was log transformed (Equation (2)).

Predicted log(Y′) = β0 + β1X1 + … + βnXn + ε

(1)

Predicted Y′ = β0 + β1 × log(X1) + … + βnXn + ε

(2)

where Y′ = predicted value of the dependent variables (the number of visits (logged) and the profitability (net income) levels); β0 = intercept; β1 + … + βn = regression coefficients (beta coefficients); X1 + … + Xn independent (or explanatory) variables; ε = error term.

To examine objective 1, we estimated three regression (log linear) models to explain the number of visits by types of on-farm and off-farm direct sales below. Model 1 provides the unstandardized regression coefficients from the ordinary least squares technique to examine the associations between direct sales and the number of visits without adjusting for any other confounders. In model 2, we estimate these relationships, controlling for other firmographic and geographic characteristics, such as the length of operation, total number of days per year open to visitors, an operation’s location from a city of at least 50,000 people, and the geographic region of the operation. In model 3, in addition to model 2, we further control the effects of other experience types, such as accommodation and lodging, education, entertainment events, and outdoor recreation, that are equally important in attracting visitors to an agritourism operation.

To examine whether profitability is independently explained by the types of on-farm and off-farm direct sales (objective 2), we estimated four regression models using the linear log regression equation. Model 1 provides the unstandardized regression coefficients from the ordinary least squares technique to examine the associations between direct sales and profitability without adjusting for any other controls. Similarly, in model 2, we estimated these relationships by netting out the effects of other firmographic and geographic characteristics, such as the length of operation, the total number of days per year open to visitors, an operation’s location from a city of at least 50,000 people, and the geographic region of the operation. In model 3, in addition to model 2, we further adjusted the effects of other experience types, such as accommodation and lodging, education, entertainment events, and outdoor recreation, that are equally important in attracting visitors and, hence, the profitability of an agritourism operation.

Note that, as the outcome variable was measured on an ordinal scale, initially, we estimated the ordinal logistic regression models. However, the results from the ordinal logistic regression and OLS provided consistent results, so we used the outcome variable as a continuous measure and estimated the OLS (linear log models), for a comparative examination with Hollas et al.’s (2021) study, which used the same data and outcome.

In general linear regression, for a categorical (or a dichotomy) explanatory variable, if the regression coefficient is positive, this will be interpreted as the outcome variable of interest (here, the number of visits and the profitability), which is increased or is higher based on the unit of the regression (β) coefficient as compared to the reference category (that is coded 0), net of other factors. Similarly, if the regression coefficient is negative, this implies that the outcome measure of interest (here, the number of visits and the profitability) has decreased or is lower based on the unit of the regression (β) coefficient as compared to the reference category (that is coded 0), net of other factors. For the log-transformed dependent variable (logged number of visits, for a log linear model, Table 4), a positive coefficient is interpreted as the number of visits higher by e^β units (or (e^β − 1) × 100 percent) and a negative coefficient is lower by e^β units (or (e^β − 1) × 100 as percent). For the log-transformed predictor or explanatory variable (the number of visits as log transformed and the profitability as the linear outcome—a linear log model, Table 5), a positive coefficient is interpreted as the profitability higher by β/100 units and a negative coefficient is lower by −β/100 units ((−β/100) × 100 as percent).

Similarly, in general linear regression, for a continuous explanatory variable, a positive coefficient indicates that when other things remain the same, a one-unit increase in the variable of interest increases the number of visits or the profitability by a given unit (β coefficient). Conversely, for a negative coefficient, a one-unit increase in the explanatory variable of interest decreased the outcome measure of interest (here, the number of visits and the profitability) by a given number (β coefficient). For the log-transformed dependent variable (logged number of visits, a log linear model, Table 4), a positive coefficient is interpreted as a one-percent increase in the predictor variable increasing the number of visits by e^β units ((or (e^β − 1) × 100 percent), and a negative coefficient is interpreted as a one-percent increase in the predict variable decreasing the number of visits by e^β units ((or (e^β − 1) × 100 percent). For a log-transformed predictor or explanatory variable (profitability as the linear outcome and the number of visits as the log-transformed predictor, Table 5), a positive coefficient is interpreted as a unit (percent) increase in the number of visits increasing the profitability by β/100 units, and a negative coefficient is interpreted as a unit increase in the number of visits decreasing the profitability by −β/100 units (or (β/100) × 100 as percent).

To examine objective 3, the mediating role of the number of visits, the results from the nested model using OLS regression (linear log model) may be used. However, we further examined the mediating role of the number of farm visits in the relationships between specific approaches to on-farm and off-farm direct sales, with impacts on profitability, to gain our confidence in our results. We used the SPSS PROCESS macro developed by Andrew F. Hayes (Copyright 2013–2025 by Andrew F. Hayes) in version 29.0.2.0(20) of SPSS. We considered the net profitability as the outcome variable (y), the number of visits as the mediating variable (m), and specific approaches to direct sales (such as a U-pick, a farm stand, community-supported agriculture—on-farm pick-up; farmers’ market and community-supported agriculture—off-farm pick-up) as predictors (x_s). As discussed earlier, we expect that various on-farm and off-farm direct sales approaches will influence profitability directly, and at the same time, indirectly through the number of visits (mediation effect).

Both direct and indirect effects are provided. As suggested by Zhao et al. (2010), a positive indirect effect suggests a complementary effect (both mediated and direct effects exist and point to the same direction), whereas a negative indirect coefficient suggests a competitive effect—both mediated and direct effects exist, but point in the opposite directions. In addition, when a mediated (indirect) effect exists, but no direct effect, it is the indirect-only effect, and when a direct effect exists but no indirect effect, it is referred to as a direct-only mediation effect. When both the effects do not exist, there is no mediation effect.

5. Results and Discussion

As described earlier, in this study, we examined whether the specific types of on-farm and off-farm direct sales independently influence the number of visits, and ultimately the profitability of an agritourism operation in the U.S. For this purpose, we control the types of offerings or experiences offered at the farm along with many other potential confounders. First, we provide the descriptive statistics of various measures—outcomes, predictors, and other controls used in the analysis.

Table 1 provides the results from a descriptive analysis. The results reveal that the average number of visits to a farm operation, the first outcome variable, is slightly over 7800 (5.95 logged), with a highly skewed distribution ranging from a minimum of 1 to a maximum of 1,300,000 visits. The second outcome variable is the self-reported profitability of an agritourism operation. Of the total, nearly one-quarter (24.6%) of operations reported that they incurred no profit, or the operation incurred a loss. On the other hand, 12% of them reported a net profit of less than $1000. Nearly 28% of them reported a profit between $1000 and $9999, another 28% reported a net profit of $10,000 to below $100,000, and the remaining 7.6% reported a net profit of $100,000 or more.

Table 1. Descriptive statistics of variables used in the analysis (n = 1263).

Variables	Mean/%	SD
Dependent variables
Number of visits (min 1, max = 1,300,000)	7840.0	46,237.65
Number of visits logged (natural) (min 0, max = 14.08)	5.96	2.52
Profitability of an operation (ordinal scale, min 1, max 8)	4.03	2.44
Profitability levels (percent)
1. No profit or the operations incurred a loss	24.5
2. Less than $1000	12.0
3. $1000–$2499	10.1
4. $2500–$4999	7.7
5. $5000–$9999	10.0
6. $10,000–$24,999	13.9
7. $25,000–$99,999	14.3
8. $100,000 or more	7.6
Explanatory variables
On-farm direct sales
U-pick (yes = 1)	29.1
Farm stand/farm store (yes = 1)	56.8
Community-supported agriculture (CSA) (yes = 1)	12.7
Off-farm direct sales
Farmer’s market (yes = 1)	28.9
CSA pick-up or delivery (yes = 1)	9.8
Controls
Length of operation (years) (min = 0, max = 93)	13.42	13.55
Experience types
Accommodation and lodging (yes = 1)	21.3
Education (yes = 1)	58.6
Entertainment events (yes = 1)	50.4
Outdoor recreation (yes = 1)	28.3
Total number of days per year open to visitors (min 1, max = 365)	181.32	132.40
% of visits who traveled 50 miles or more one-way (min = 0, max = 100)	35.30	30.72
Location from a city of at least 50,000 people (miles)
Less than 10 miles	20.9	-
10–29 miles	27.1	-
30–49 miles	21.2
50 miles or more	30.2
Geographic region of the operation
Northeast (yes = 1)	22.6	-
Midwest (yes = 1)	21.9
Southern (yes = 1)	29.5
Western (yes = 1)	26.0	-

Source: National Agritourism Survey, 2019.

Table 1 also provides the distribution of specific approaches of on-the-farm and off-the-farm direct sales to examine their influence on the number of visits to and the profitability of an operation. Among on-the-farm direct sales, of the total, 29 percent of the operations reported that they had a pick-your-own (U-pick) operation offered to the visitors. Similarly, slightly over half (57%) of them reported that they had a farm stand or a farm store. On the other hand, only 13 percent of them revealed that they had community-supported agriculture (CSA) where the members picked up their delivery from the farm (on-the-farm CSA pick-up).

The sale of produce at the farmers’ market and the delivery of produce to the members of community-supported agriculture (CSA) were the two off-the-farm methods of direct sales assessed. Of the total, 29 percent reported selling through a farmers’ market, and nearly 10 percent reported that they deliver produce to their members at various agreed-upon locations (off-the-farm pick-up).

The average length of operation was slightly over 13 years. While nearly three percent of the operations had just started agritourism operations in the survey year (2019), slightly over one percent (1.3%) had started about 50 years ago. While nearly three-fifths of operations (58.6%) provided educational activities, such as tours, classes, hands-on practices, and so on, slightly over half of them (5.4%) offered entertainment events such as festivals, concerts, and more. Twenty-eight percent of them provided outdoor recreation activities, and slightly over one-in-five operations provided accommodation and lodging. Nearly 18 percent of them offered only one activity (no diversification, mostly those offering on-farm direct sale), slightly over 25 percent (25.3%) offered a combination of two experiences, nearly 27 percent of them offered a combination of three activities, and nearly 30 percent of them provided four or more activities (diversified operation) to the visitors.

On average, an operation was open to visitors for about 181 days (6 months), with a minimum of 1 to a maximum of 365 days (year-round). Slightly over one-third (35.3%) of visitors traveled to the operation from 50 miles or more. Slightly over one-fifth (20.0%) of operations reported that they were located within 10 miles of a city with a population of 50,000 or more. On the other hand, slightly over 30 percent of them were located within 50 miles or more of a city with a population of 50,000 or more. About 23 percent of them were in the Northeast region, whereas nearly 30 percent of them were in the Southern region. The results show that about an equal proportion of the operations were distributed in all four different regions.

5.1. Direct Sales and the Number of Visits

Our first research question is, (i) Does a specific approach of on-farm or off-farm direct sales influence the number of visits to an agritourism operation differently? Results from Table 2 (panel i) reveal the average number of visits to operations that offer various on-farm and off-farm direct sales. For example, on average, an operation with on-farm direct sales, in total, received 9097 visits (logged 6.148) in a year. As expected, those farms that sold their produce directly to customers off the farm reported fewer visits to their farm (7564).

Table 2. Number of visits by direct sales approaches.

Direct Sale	Average Number (Std. Dev.)	Average Number (Logged) (Std. Dev.)
On-farm (total, n = 1015)	9097 (51,124)	6.148 (2.511)
U-pick (n = 368)	15,798 (76,950)	6.735 (2.681)
Farm stand/store (n = 717)	11,309 (59,911)	6.353 (2.537)
CSA on-farm pick-up (n = 160)	13,499 (106,761)	5.746 (2.514)
Off-farm (total, n = 572)	7564 (59,868)	5.792 (2.485)
Farmers’ market (n = 365)	7570 (71,150)	5.674 (2.444)
CSA pick-up or delivery (n = 124)	13,290 (117,425)	5.370 (2.291)

Source: National Agritourism Survey 2019.

Specifically, those that offered U-pick received the most—15,798 (logged 6.735)—visits, followed by 13,499 (logged 5.746) visits at community-supported agriculture (CSA, on-farm pick-up) and 11,309 (logged 6.353) visits at the farm that had a stand/store.

On the other hand, an operation with a farmers’ market, on average, received 7564 (logged 5.792) as compared to 13,290 (logged 5.370) visits at CSA (pick-up or delivery). Note that the normalized (logged) numbers are slightly different from the raw numbers due to a highly skewed distribution, making the use of logged numbers more sensible.

Note that these numbers are just an indication of the trend of visitors and are not mutually exclusive. A farm may have used one or more types of direct sales methods to sell its products to visitors. For instance, a farm that has a pick-your-own operation may also sell produce at the farm stand/store, at a farmers’ market, or through subscription to community-supported agriculture (CSA).

Results from Table 3 reveal the relationships between the types of on-farm and off-farm direct sales and the number of visits to an agritourism operation. The results from three nested equations (models) are designed to present the net effects after adjusting for potential confounders. As described earlier, Model 1 provides the relationship between the types of direct sales and the number of farm visits without adjusting for the effects of other controls. In Model 2, we adjust for the effects of a series of controls. As explained in the theoretical reviews, these controls are expected to confound the relationship between the types of direct sales and the number of farm visits. Finally, in Model 3, these relationships are further adjusted, controlling for the effects of other agritourism experiences that are equally important in attracting visitors to an operation. Now, let’s discuss the results.

5.1.1. U-Pick and the Number of Visits

As shown by the results, an operation with a U-pick received a significantly greater number of visits (0.935 logged number) (unstandardized β = 0.935, p < 0.001; model 1) as compared to those that did not offer a U-pick, net of other types of direct sales. In other words, those farms with a U-pick operation invited 2.55 (e^β = e^0.935 units) more visitors than those that did not offer a U-pick, net of other sales approaches. Additionally, when we adjusted for the effects of other confounders, such as the length of operation, total number of days open to visitors, farm’s location from a city of at least 50,000 people, and the geographic location of a farm, the direction and the strength of the association remained, except a slight decrease in magnitude (unstandardized β = 0.877, p < 0.001, model 2). Furthermore, we controlled other agritourism activities, such as accommodation and lodging, education, entertainment, and outdoor recreation, that potentially influence the relationship between the number of visits and the types of direct sales. Even after adjusting for the effects of other agritourism activities, direct selling of produce through U-pick still attracted significantly more visits to the operation than those without it (unstandardized β = 0.808, p < 0.001, model 3). These results provide us with confidence in our results.

Table 3. Unstandardized regression coefficients from the ordinary least squares (log linear regression) technique to examine the associations between direct sales and the number of visits, net of other controls (n = 1263).

Variables	Model 1	Model 2	Model 3
Independent Variables
On-farm direct sales
U-pick (yes = 1)	0.935 ***	0.877 ***	0.808 ***
Farm stand/farm store (yes = 1)	0.946 ***	0.821 ***	0.688 ***
CSA on-farm pick-up (yes = 1)	−0.045	−0.003	−0.059
Off-farm direct sales
Farmer’s market (yes = 1)	−0.639 ***	−0.493 **	−0.451 **
CSA delivery (yes = 1)	−0.461 †	−0.488 †	−0.418
Controls
Length of operation (years) (mean) (min = 0, max = 93)	-	0.027 ***	0.027 ***
Total number of days per year open to visitors (mean)	-	0.002 ***	0.002 ***
Farm’s location from a city of at least 50,000 people (Ref = less than 10 miles)	-
10–29 miles	-	−0.154	−0.133
30–49 miles	-	−0.593 **	−0.517 **
50 miles or more	-	−0.622 ***	−0.448 *
Geographic region of the operation (Ref = Northeast)	-
Midwest (yes = 1)	-	−0.091	−0.183
Southern (yes = 1)	-	0.045	−0.028
Western (yes = 1)	-	0.338 †	0.283
Experience type
Accommodation and lodging (yes = 1)	-	-	−0.462 **
Education (yes = 1)	-	-	−0.145
Entertainment events (yes = 1)	-	-	1.165 ***
Outdoor recreation (yes = 1)	-	-	0.135
Intercept	5.385 ***	4.938 ***	4.559 ***
Regression sum of squares	636.52	1052.380	1490.538
Residual sum of squares	7398.12	6982.256	6544.099
Residual degrees of freedom	1257	1249	1245
ANOVA F-ratio	21.630 ***	14.481 ***	16.681 ***
Adjusted R-square (%)	7.6	12.2	17.4

^† p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001.

Table 3. Unstandardized regression coefficients from the ordinary least squares (log linear regression) technique to examine the associations between direct sales and the number of visits, net of other controls (n = 1263).

Variables	Model 1	Model 2	Model 3
Independent Variables
On-farm direct sales
U-pick (yes = 1)	0.935 ***	0.877 ***	0.808 ***
Farm stand/farm store (yes = 1)	0.946 ***	0.821 ***	0.688 ***
CSA on-farm pick-up (yes = 1)	−0.045	−0.003	−0.059
Off-farm direct sales
Farmer’s market (yes = 1)	−0.639 ***	−0.493 **	−0.451 **
CSA delivery (yes = 1)	−0.461 †	−0.488 †	−0.418
Controls
Length of operation (years) (mean) (min = 0, max = 93)	-	0.027 ***	0.027 ***
Total number of days per year open to visitors (mean)	-	0.002 ***	0.002 ***
Farm’s location from a city of at least 50,000 people (Ref = less than 10 miles)	-
10–29 miles	-	−0.154	−0.133
30–49 miles	-	−0.593 **	−0.517 **
50 miles or more	-	−0.622 ***	−0.448 *
Geographic region of the operation (Ref = Northeast)	-
Midwest (yes = 1)	-	−0.091	−0.183
Southern (yes = 1)	-	0.045	−0.028
Western (yes = 1)	-	0.338 †	0.283
Experience type
Accommodation and lodging (yes = 1)	-	-	−0.462 **
Education (yes = 1)	-	-	−0.145
Entertainment events (yes = 1)	-	-	1.165 ***
Outdoor recreation (yes = 1)	-	-	0.135
Intercept	5.385 ***	4.938 ***	4.559 ***
Regression sum of squares	636.52	1052.380	1490.538
Residual sum of squares	7398.12	6982.256	6544.099
Residual degrees of freedom	1257	1249	1245
ANOVA F-ratio	21.630 ***	14.481 ***	16.681 ***
Adjusted R-square (%)	7.6	12.2	17.4

^† p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001.

5.1.2. Direct Sale at a Farm Stand/Farm Store and the Number of Visits

Our results also revealed that an agritourism operation that sells produce directly to customers in a farm stand/store also attracted significantly more visits to their farm (unstandardized β = 0.946, p < 0.001, model 1) as compared to those that did not have a farm stand/store, net of other sales approaches. Those farms with a farm stand or a farm store invited 2.58 (e^β = e^0.946 units) more visitors than those that did not offer a U-pick, net of other sales approaches. As per expectation, these results, although the magnitude slightly decreased, hold true even after adjusting for other controls (in models 2 and 3).

5.1.3. Community-Supported Agriculture and the Number of Visits

Some farms offer CSA subscriptions to customers to maintain a regular supply of produce. Some other farms sell CSA memberships to customers while offering many other attractions. In addition, some CSA members pick up produce from the farm, whereas others want it delivered off-site.

In both cases, the evidence suggests that an operation with CSA on-farm pick-up, off-site pick-up, or delivery reduced the number of visits to the farm. However, the results are statistically not significant, implying no difference from those who did not have this offer. This is true in all three models assessed (Table 3). The size of the relationships is also as expected—an operation with off-site delivery of CSA products receives fewer visits to the farm compared to on-site delivery or pick-up, although both are statistically not significant.

5.1.4. Direct Sales (Off-Farm) at the Farmers’ Market and the Number of Visits

The evidence from Table 3 clearly shows that if an operation sells produce at a farmers’ market, the operation receives a significantly fewer number of visits to the farm than those that do not sell there. The evidence is that the number of visits to the operation is reduced by 0.639 units (logged) (unstandardized β = −0.639, p < 0.001, model 1) if an operation sells at a farmer’s market. Simply put, an agritourism operation that sold its produce at a farmers’ market had about 1.89 (e^β = e^−0.639 units) fewer visits than that which did not use a farmers’ market as a sales outlet, net of other direct sales approaches used. This result continues to hold true even after adjusting for the effects of several other variables in the equation (in models 2 and 3). This result is plausible because off-the-farm sales do not necessarily attract visitors to a farm. However, farmers’ markets could play a complementary role by showcasing and promoting farmers’ produce and their farms among large groups of customers (Hollas et al., 2021). This could also provide an opportunity to develop networking with fellow farmers and customers.

Now, let us examine the theoretical validity of relationships. As theoretically expected, the length of operation (number of years of establishment of the operation) statistically significantly increased the number of visits to an operation. Similarly, the number of days the operation was open had a positive impact on the number of visits. On the other hand, if the farm is located further from a city of at least 50,000 population, it significantly reduces the number of visits to the operation. However, these results did not significantly vary by geographic region. More interestingly, the direction and the magnitude of the relationships remained with a marginal decline even after adjusting for the effects of other experience types (model 3). These results provide us with great confidence in the reliability of the instrument and the validity of our results.

5.1.5. Direct Sales, Number of Visits, and Profitability

A previous study, using the same data, provided evidence that operations that directly sold their produce at the farm reported significantly higher profitability (Hollas et al., 2021). In contrast, the same study revealed that operations that sold their produce off-farm (i.e., at the farmers’ market and/or CSA delivery) reported significantly lower profitability. However, it is not clear what specific approach of on-farm or off-farm direct sales generates more profits than others and why. Motivated by that research, this paper dove deeper and examined the next research questions: (ii) Does a specific approach of on-farm or off-farm direct sales independently influence the profitability of an agritourism operation? and (iii) Does the number of farm visits mediate the relationship between on-farm and/or off-farm direct sales and the profitability of an agritourism operation?

Table 4 (panel i) provides the losses or profits experienced by operations by types of direct sales. The results show that slightly larger proportions of those who offered off-farm direct sales reported either no profit or incurred a loss (24.1%) compared to 22.1% of those who offered on-farm direct sales. In particular, fewer proportions of those who had a U-pick (20.9%) and a farm stand/store (21.8%) reported either no profit or incurred a loss, as compared to 28.1% of those who offered community-supported agriculture (CSA) on-farm pick-up. On the other hand, a similar proportion of those who sold their produce at the farmers’ market (22.7%) and via CSA pick-up or delivery (23.4%) did not make a profit.

Table 4. Direct sales, profitability, and the number of visits.

1. Profitability by Types of Direct Sales Approaches
Direct sale	Yes, profit (%)	No profit or incurred a loss (%)
On-farm (total, n = 1015)	77.9	22.1
U-pick (n = 368)	79.1	20.9
Farm stand/store (n = 717)	78.2	21.8
CSA (on-farm pick-up, n = 160)	71.9	28.1
Off-farm (total, n = 572)	75.9	24.1
Farmers’ market (n = 365)	76.3	22.7
CSA pick-up or delivery (n = 124)	76.6	23.4
2. Number of Visits and Profitability Status
Profitability levels	Average number (std. dev.)	Average number (logged) (std. dev.)
1. No profit or incurred a loss (n = 310)	2970 (15,761)	5.446 (2.305)
2. Less than $1000 (n = 151)	866 (3412)	4.549 (2.003)
3. $1000–$2499 (n = 128)	803 (2852)	5.094 (1.879)
4. $2500–$4999 (n = 97)	1335 (2770)	5.660 (1.923)
5. $5000–$9999 (n = 126)	3431 (8933)	6.449 (2.088)
6. $10,000–$24,999 (n = 175)	3084 (6252)	6.299 (2.308)
7. $25,000–$99,999 (n = 180)	11,466 (34,474)	7.055 (2.633)
8. $100,000 or more (n = 96)	58,148 (149,081)	7.969 (3.452)

Source: National Agritourism Survey 2019.

We also examined whether the number of visits differed by the amount of profitability reported by farms (Table 4, panel ii). The trend of the relationship (based on logged value) suggests a positive correlation between the two—farms reporting a higher level of profitability with a greater number of visits. In general, the trend of distribution shows an association between the types of direct sales, the number of visits, and the profitability of an agritourism operation.

Now, let us focus on the results of the multivariate (linear log regression) analysis to determine whether the profitability of an agritourism operation significantly differs with the specific method of on-farm or off-farm direct sales. Results from Table 5 (model 1) reveal that an agritourism operation with a U-pick and or a farm stand/store statistically significantly increases the chance of reporting profits. For instance, those farms that offered a U-pick activity to visitors had 0.586 points (i.e., ((β/100) = 0.586/100 = 0.00586 units) more on the profitability level (ordinal scale) than those that did not offer a U-pick, net of other direct sales approaches (model 1). These results are consistent and remain statistically significant after controlling for all other theoretically important confounders (model 2) and various experience types (model 3). However, we find a slight decline in the magnitude and the strength of relationships when we control for other factors.

Table 5. Unstandardized regression coefficients from the ordinary least squares technique (linear log regression) to examine the associations between direct sales and net profitability, adjusting for controls (n = 1263).

Variables	Model 1	Model 2	Model 3	Model 4
Independent Variables
On-farm direct sales
U-pick (yes = 1)	0.586 ***	0.426 **	0.409 **	0.212
Farm stand/farm store (yes = 1)	0.558 ***	0.441 **	0.385 **	0.216
CSA on-farm pick-up (yes = 1)	−0.362	−0.156	−0.185	−0.170
Off-farm direct sales
Farmer’s market (yes = 1)	−0.295 †	−0.125	−0.107	0.004
CSA pick-up or delivery (yes = 1)	0.121	0.169	0.212	0.314
Controls
Length of operation (years) (mean) (min = 0, max = 93)	-	0.047 ***	0.047 ***	0.041 ***
Total number of days per year open to visitors (mean)	-	0.001 *	0.001 *	0.001
Percent of visitors who traveled 50 or more miles (min = 0, max = 100)	-	0.009 ***	0.009 ***	0.009 ***
Farm’s location from a city of at least 50,000 people (Ref = less than 10 miles)	-
10–29 miles	-	−0.165	−0.167	−0.135
30–49 miles	-	−0.369 †	−0.349 †	−0.224
50 miles or more	-	−0.461 *	−0.396 *	−0.289
Geographic region of the operation (Ref = Northeast)	-
Midwest (yes = 1)	-	−0.078	−0.118	−0.074
Southern (yes = 1)	-	0.296	0.273	0.278
Western (yes = 1)	-	0.053	0.018	−0.052
Experience type
Accommodation and lodging (yes = 1)	-	-	−0.080	0.030
Education (yes = 1)	-	-	−0.132	−0.097
Entertainment events (yes = 1)	-	-	0.545 ***	0.260 †
Outdoor recreation (yes = 1)	-	-	−0.130	−0.163
Number of visits (logged)	-	-	-	0.245 ***
Intercept	3.664 ***	2.702 ***	2.592 ***	1.474 ***
Regression sum of squares	223.11	900.118	985.603	1377.899
Residual sum of squares	7279.63	6602.616	6517.131	6124.835
Residual degrees of freedom	1257	1248	1244	1243
ANOVA F-ratio	7.705 ***	12.153 ***	10.452 ***	14.718 ***
Adjusted R-square (%)	2.6	11.0	11.9	17.1

^† p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001. Source: National Agritourism Survey, 2019.

It is interesting to note that the direct selling of products to customers through a farmers’ market, CSA on-farm pick-up, CSA off-farm pick-up, or delivery were statistically significantly not associated with the profitability of an operation. Surprisingly, although statistically not significant, the direction of the relationship between the profitability and selling produce at the farmers’ market or via CSA delivery of produce on-farm was negative, while CSA off-farm delivery was positively associated with profitability.

Now, let us empirically investigate the next research question: whether the number of farm visits mediates the effect of direct sales approaches on profitability. A previous study by Hollas and colleagues (Hollas et al., 2021) showed that the number of visits did not have a statistically significant association with profitability. These scholars suggested that this could be possibly because “… many farms remain open longer and receive more visitors, without charging each visitor for their experience” (p. 9). However, as theoretically expected, our results using the same data, which used a direct sales approach, clearly reveal that the number of visits (log transformed) statistically and significantly increased (unstandardized β = 0.245, p < 0.001; model 4, Table 5) the profitability of an agritourism operation, net of all other controls. This result suggests that a one (percent) unit increase in the number of visits increased the profitability (in ordinal categories) of a farm by 0.00245 (which is β/100 = 0.245/100) units, net of all other factors. In addition, the model’s explanatory power increased from 11.9% (model 3) to 17.1% when the number of visits was added to the equation simultaneously, revealing this variable’s significance. Conversely, the magnitude of the relationship for all the direct sales approaches was reduced, and the relationship turned out to be statistically insignificant, suggesting that the number of visits mediated the effect of direct sales approaches on profitability. For instance, the regression coefficient for U-pick decreased from 0.586 (model 1) to 0.212 (model 4), as shown in Table 5, and the statistical significance was lost when the number of visits was used in the equation. We have similar observations for selling produce at a farm stand/farm store. The results imply that the farm direct sales approaches, particularly U-pick and a farm stand, first influenced the number of visits to a farm, which statistically significantly and positively contributed to profitability, suggesting the mediating role of the number of visits to influence net profits. These results are the net of all other potential confounders. However, the results for CSA and selling at a farmers’ market did not have a significant effect on the number of visitors (Table 4) and on the profitability (Table 5).

To confirm the mediating effects, as stated in the analysis section, we further examined the mediating effect of the number of visits on profitability (Table 6) using the PROCESS macro in SPSS. These results are based on simple mediation models (Preacher & Hayes, 2004). The results show a clear complementary effect of U-pick (direct effects = 0.3368, p < 0.05; and indirect effects 0.3356, p < 0.05) and a farm stand (direct effect = 0.2839, p < 0.05; indirect effect = 0.2816, p < 0.05). According to Zhao et al. (2010), these effects are complementary, as both the direct and indirect effects exist and point in the same direction, implying that the direct sales approaches, here U-pick and a farm stand, significantly and directly increase profitability. At the same time, they also positively and significantly influence profitability indirectly through the number of visits. However, selling at farmers’ markets (direct effect = 0.0023, p > 0.05; indirect effect = −0.1278, p < 0.05) suggests an indirect-only effect, implying that selling products to customers at a farmers’ market (off-the-farm) reduces profitability significantly. The CSA, on-farm delivery as well as off-farm delivery did not have a significant effect on the number of visits or on the profitability. These results are without adjusting for other variables. These results are perhaps reasonable as the CSA managers often have small number of subscriptions, not necessarily inviting other visitors to the farm. Or it could be that CSAs are designed to meet the demand for farm-fresh produce in local communities and to help build community relationships or to sustain family farms through mutual support.

Table 6. Mediation analysis—direct and indirect effects.

Direct Sales Approaches	Visits	Profitability
		Total Effect	Direct Effect	Indirect Effect (Bootstrapped Confidence Intervals)
U-pick (on-farm)	1.0953 (p < 0.001)	0.6759	0.3368 (p < 0.05)	0.3356 * (BootLLCI 0.2242; BootULCI 0.4577)
Farm stand (on-farm)	0.9131 (p < 0.001)	0.5655	0.2839 (p < 0.05)	0.2816 * (BootLLCI 0.4577; BootULCI 0.3860)
CSA pick-up (on-farm)	−0.2437 (p > 05)	−0.2939	−0.2165 (p > 0.05)	−0.0774 NS (BootLLCI −0.2134; BootULCI 0.0598)
Farmers’ market (off-farm)	−0.4012 (p < 0.01)	−0.1255	0.0023 (p > 0.05)	−0.1278 * (BootLLCI −0.2274; BootULCI −0.0309)
CSA delivery (off-farm)	−0.6531 (p < 0.01)	−0.1335	0.0749 (p > 0.05)	−0.2084 (BootLLCI −0.3520; BootULCI −0.0721)

*: Statistically significant; NS: Statistically not significant.

In summary, the various types of direct sales approaches differentially attracted the number of visits to the farm. For example, the U-pick and a farm stand/farm store attracted significantly more visits to an operation, ultimately and significantly increasing the likelihood of profitability. Evidence further suggested that the number of visits plays a role in enhancing the profitability of agritourism operations. However, the other three approaches of on-farm and off-farm direct sales did not contribute significantly to the number of visits as well as profitability. Perhaps, as indicated by other scholars, these marketing approaches, CSA and farmers’ market sales, were used by agritourism operations as a marketing tool for the farm’s products and services, and thus, the revenues or profits from agritourism were indirectly influenced (Hollas et al., 2021; Tew & Barbieri, 2012). Moreover, CSAs are important to develop and strengthen community networking and relationships. And farmers’ markets are used by operators to expand their customer base, and develop networking with other peers and customers, rather than making profits. This is worth investigating, even though many agritourism operations use farmers’ markets as an important outlet to sell their farm-fresh produce, and more of them are increasingly attracted towards ‘subscription farming’ or community-supported agriculture to diversify incomes.

Several of the theoretically important variables, such as the length of operation (number of years of establishment), the number of days the operation was open, and the percentage of visits from 50 or more miles, significantly and positively contributed to profitability. On the other hand, as expected, if a farm’s location is farther away from a city of 50,000 or more people, the profitability of operations decreases. These results are consistent with the findings from previous studies (Barbieri et al., 2008; Bhandari et al., 2024; Hollas et al., 2021), providing us confidence in our findings.

6. Conclusions and Implications

Previous research using this same survey data provided evidence that on-farm direct sales increased the profitability of an agritourism operation, whereas the off-farm direct sales were significantly but negatively associated with the profitability (Hollas et al., 2021). This research, however, did not examine the relationships between specific direct sales approaches and profitability. Moreover, this research also did not examine the mediating role of the number of visits in shaping the relationships between specific direct sales approaches and profitability. Expanding this important research of Hollas and colleagues, in this paper, we examined whether a specific type of direct sale increases or decreases the number of visits to a farm and whether these direct sales approaches increase or decrease profitability differentially. Moreover, we also examined whether the number of visits to a farm mediated the relationships between specific direct sales approaches and profitability.

Our findings provide empirical evidence that (a) the number of visits to an agritourism operation depends on the types of direct sales approaches adopted, (b) the types of direct sales approaches influenced the profitability of an agritourism operation differentially, and, however, (c) the number of visits to a farm differentially mediated the relationships between specific direct sales approaches and profitability generated by a farm.

The adjusted results from the multivariate OLS regression (linear log) show that not all the approaches of direct sales considered in this study significantly influenced the visits to a farm as well as the profitability. For instance, (a) U-pick and a farm stand/store significantly increased the number of farm visits as well as the profitability, (b) selling of produce through a farmers’ market significantly reduced the number of visits to a farm, but not the profitability, and (c) community-supported agriculture (on-farm pick-up or off-site delivery) neither influenced the number of visits to a farm nor profitability. However, when the number of visits was simultaneously added to a profitability equation along with other variables, both the magnitude of the effect as well as the strength of association declined and turned out to be statistically insignificant, perhaps suggesting a mediating role of the number of visits in shaping the relationship between direct sales approaches, specifically, a U-pick and farm stand sales, and profitability. That is an interesting finding of this study.

On the other hand, selling produce off the farm at the farmers’ markets significantly reduced the number of visits to a farm, which is clear. This is because producers brought their produce and sold their produce directly to customers off the farm. Moreover, selling produce at a farmers’ market also did not show a significant influence on profitability. Despite this, we should not discount the significance of farmers’ markets for several reasons. First, farmers’ markets supply locally produced fresh produce to the customers’ tables, impacting positively on environmental and health outcomes. Second, a farmers’ market could be an important place to showcase and promote farmers’ produce to customers. Third, selling produce at the farmers’ markets provides farmers with networking opportunities with their peers and customers, which might indirectly influence customers to visit their farms and increase on-farm direct sales. A farmers’ market is an important venue for producers to sell their produce to customers if they have limitations in inviting visitors to their farms.

Moreover, the influence of community-supported agriculture (CSA) in attracting visitors and profitability is statistically not significant. It seems plausible as producer farmers offer subscriptions to a limited number of customers, mostly local. These farmers are not necessarily opening their farms to a wider public for several reasons, including liability issues. In addition, its effect on profitability is also not clear. However, CSAs have been important in building community relationships, networking, and at the same time offering fresh produce to the table for healthy outcomes, which indirectly offer many social, economic, and environmental benefits to the farmers and the entire community. These results have both theoretical as well as practical implications.

In summary, the common belief is that any method of direct sales would increase the profitability of an operation. However, this relationship was not examined previously for various approaches to direct sales specifically. This paper contributes to the existing literature with differential evidence on the relationships between the types of direct sales, the number of visits, and profitability. Secondly, this paper also provided evidence of the mediating role of the number of visits in shaping the relationships between the on-farm direct sales approaches (U-pick and farm stand sales) and profitability. Contrastingly, the previous research had found that the number of visits was not significantly associated with the profitability (Hollas et al., 2021). However, our results clearly indicated that the effect of direct sales approaches, and the other experience types, adjusting for all other potential confounders, indeed worked through the number of visits to influence profitability. Overall, increasing the number of visits to the farm, encouraging direct sales of a farm’s products and services, and providing various experiences and entertainment activities at an agritourism farm are quite important for generating profitable incomes and for sustaining their businesses for long-term economic viability.

These findings are relevant to agritourism operators who are planning to establish or expand their operations, to invite more visitors to their farms and generate revenue from them. More importantly, to those who are planning to sell their produce through community-supported agriculture or at the farmers’ market, this finding will be very helpful. However, more investigation is necessary about the significance of CSAs and farmers’ markets. In addition, adding entertainment opportunities would significantly increase the number of visits as well as profitability. From a policy perspective, it is important for the government to find supporting mechanisms for small and medium farmers, to encourage the invitation of visitors to their farms through on-farm direct sales approaches. Extension agents in the field will find these findings useful to generate more income, for their economic sustainability.

Despite the significant evidence, consider the findings carefully for several reasons. First, we used crude measures of the number of visits and profitability from the agritourism operations. These numbers do not directly come from a farm’s record but are self- reported by an operator. However, we believe these are the best reports measured thus far by any agritourism survey. Second, the large number of missing cases in the data raises a concern of sample bias. Although the exploratory analysis we performed as indicated above suggests that the samples were missing randomly and that the empirical findings from the multivariate analysis support the theoretical expectations, these results should be considered carefully. Third, this study failed to control other factors, such as the operation size, marketing and promotion, operational cost, commodity prices/fees, and motivation for farming, which may potentially influence the number of visits and profitability. In addition, we could not touch on online marketing as a direct marketing approach due to a limitation of the data. Moreover, addressing some methodological challenges, such as endogeneity and self-selection bias, could produce more robust results. Furthermore, results from longitudinal (panel) data could yield more conclusive results than those based on cross-sectional data used here, and, if possible, longitudinal panel data collection from agritourism operations and analysis. Even though this is the most recent and richest data collected from a variety of agritourism operations spread over all 50 states of the United States, future studies considering these limitations and augmented by mixed-method approaches (enriched by qualitative methods) and longitudinal (panel) data could generate conclusive evidence.

Based on the evidence, we expect that offering direct sales modes such as U-pick and selling at a farm stand/store will positively attract visitors, and the increased number of visits will increase direct sales, ultimately resulting in profitability. However, we must acknowledge that not all agritourism operations have the capacity to invite, accommodate, and manage as many visitors as they like. On the other hand, not all visitors have similar interests and motivations to visit an agritourism operation and purchase farm produce and services from there. More importantly, visitors’ spending behaviors are different. This raises important questions, such as how many visitors are required or how many direct sales are needed to break even? Does the farm have the capacity to invite, accommodate, and engage as many visitors as it likes? What about support from the community or neighbors? While these questions are beyond the scope of this study, we believe these are important topics for further investigation.