The project objective is to assess farmer profitability in the Northeast region from selling a new product, retail frozen local foods, by estimating key returns and costs, and then subtracting costs from returns. We fill the knowledge gap by: 1) estimating consumer demand for locally processed products; 2)developing non-proprietary standard operating procedures for two safe, high-quality, popular retail frozen fruit and vegetable products, and: 3)estimating processing/production costs using scale processing pilots and costs data from community partners. This information will fill the knowledge gap that hinders adoptability of retail frozen local sales in the Northeast.
Problem, Novel Approach and Justification:There is an opportunity for farmers to meet growing demand for local foods and increase farm profitability by entering a new market for retail sales of frozen value-added products. In particular, farmers could capitalize on opportunities provided by recent investments in regional food processing facilities by freezing produce for retail sales in winter. Like many new marketing opportunities, farmers do not have access to information they need to determine profitability in the new market: whether the returns from processing and selling retail frozen produce are greater than the costs or producing safe, high-quality food products that meet food safety regulations.
Farmers have indicated that they are interested in the market, but need better information on: processing/production costs; food safety/quality assurances; consumers’ demand/pricing; and packaging/marketing. We propose to combine original research in this proposal with existing supply-chain information provided by community partners to fill the information gap so farmers can determine whether this could be a profitable market for their operation.
Hypothesis and Research Plan:Our overall research hypothesis is that local produce can be profitably grown and processed (frozen) for off-season retail sales. Two additional hypotheses are important to that overall hypothesis. First, we hypothesize that consumers have higher willingness-to-pay for locally produced and processed frozen foods, and second, that the costs of producing safe, high-quality locally grown and processed frozen foods will not exceed consumers’ willingness-to-pay.
To test our hypotheses, we estimate returns using a demand survey to estimate consumer willingness-to-pay for product attributes, including labels, packing, origin information, pricing, past purchases, and demographics. We develop two safe, high-quality prototype products, and produce them at-scale to assess processing/production costs. We supplement this cost research with marketing data from community partners to estimate total costs, then subtract cots from returns to assess profitability.
Outreach Plan:Research results will be shared through outreach efforts utilizing networks and programs established within UMass Extension. We provide additional knowledge regarding local value-added production opportunities to Northeast farmers and other agricultural partners and service providers. We deliver 10 outputs through a diverse outreach platform, including technical presentations, extension workshops, webinars, and peer-reviewed and extension publications.
Project Objective:The project objective is to assess farmer profitability in the Northeast region from selling a new product, retail frozen local foods, by estimating key returns and costs, and then subtracting costs from returns. We fill the knowledge gap by: 1)estimating consumer demand for locally processed products; 2)developing non-proprietary standard operating procedures for two safe, high-quality, popular retail frozen fruit and vegetable products, and: 3)estimating processing/production costs using scale processing pilots and costs data from community partners. This information will fill the knowledge gap that hinders adopt-ability of retail frozen local sales in the Northeast
Our overall research hypothesis is that local produce can be profitably grown and processed (frozen) for off-season retail sales. Two additional hypotheses are important to that overall hypothesis. First, we hypothesize that consumers have higher willingness-to-pay for locally produced and processed frozen foods, and second, that the costs of producing safe, high-quality locally grown and processed frozen foods will not exceed consumers’ willingness-to-pay.
RETURNS-CONSUMER DEMAND SURVEY (WTP)
a) Target Population: Primary household shoppers in the Northeast surveyed through MTurk (Mechanical Turk, described below).
b) Methods: Returns from sales of regionally frozen retail products will be estimated via consumer willingness-to-pay market research. In 2018, UMass researchers completed a literature review of fresh and value-added origin-identified consumer products. Following this literature review, UMass researchers proposed a draft willingness-to-pay survey that included the product attribute categories that are both relevant to the practical needs of the Farmers interested in selling retail frozen produce and are of cutting-edge academic interest. The survey design includes a conjoint choice experiment, as well as a series of questions regarding past and future shopping practices, and demographic information.
The UMass research team vetted the survey design, first with the team from the FCCDC. The purpose of the vetting was to explain the design elements and intended outcomes, share the instrument design process and how the draft survey was informed by the current literature, and to receive feedback and make changes to the design, if needed. The FCCDC team made a number of requests for changes. Based on these requests, the UMass team returned to the literature to determine the best path forward. Using additional literature and the FCCDC’s suggestions, the UMass team made a number of changes to the survey design.
The UMass research team then vetted the survey design with the Advisory Board. The Advisory Board similarly made a number of useful suggestions, and requested information regarding the proposal design. The UMass team proposed an additional meeting with interested Advisory Board members, after the survey draft and design was complete.
Conjoint Choice Experiment
Attribute Levels Baseline Research
Price & Package Size Attributes. To develop the conjoint choice experiment, FCCDC staff collected blueberry and spinach product data from 27 regional stores, including price and package size information. UMass researchers will analyze this data to determine the optimal package size and price ranges for the choice experiment attributes.
Product Packaging – FCCDC staff solicited product packaging samples from different companies to be used in photographs of products for the choice experiment. Because the FCCDC operates on a relatively small scale, many packaging companies have not been interested in providing samples, particularly for opaque packaging. The cost of product packaging is very high.
The draft includes:
i) Past purchasing behavior to establish a baseline from which to estimate changes in consumer demand due to the product availability, include:
1) Frequency of recent purchases of fresh and frozen produce labeled “local”, “regional”, or “Grown in USA”;
2) Recent sales outlet at which fresh and frozen produce were purchased: direct (farmers market, CSA, farm stand); super stores (“big box” stores); supermarkets (regional/national grocery store chains); local grocers; or cooperative grocers.
ii) Consumers’ WTP for frozen regional product attributes: A randomized conjoint choice experiment. Respondents are shown a choice set of four products and the choice to not purchase any of the products. Each choice set offers the respondent products with five randomized product attributes. The randomly selected attributes for each product in the choice sets include:
1) Price: standard price/12 oz. bag of blueberries
2) Production Origin: no information, local, regional, Northeast, domestic
3) Purchase Location: Direct (farmers market, CSA, farm stand); Super Store (“Big Box” store offering large household goods in addition to grocery items); Supermarket (Regional or National Grocery Store Chain); Cooperative grocer.
4) Processing Origin: no information, local, Northeast, domestic
5) Package Design: opaque sticker label; opaque printed label; clear sticker label; clear printed label
iii) Demographics: age, gender, income, education, household size and make-up, race/ethnicity, and location.
c) Data Collection: We will administer the survey based on the Dillman-Tailored Design Method. The choice experiment is developed using Qualtrics, and we field the survey using Amazon’s Mechanical Turk (MTurk). We screen respondents for primary household shoppers in the Northeastern US with high MTurk approval rates. The MTurk survey population is representative of average American consumers and is frequently used for consumer choice experiments. We include “attention check” questions and track response time to screen respondents for quality and likelihood that they accurately read survey questions. The survey takes 7-12 minutes to complete.
Survey Software Challenges – The UMass team encountered an unexpected challenge in the process of designing the draft choice experiment survey. The team had planned on using Qualtrics software to design the choice experiment and survey, but found that the current Qualtrics license held by UMass does not include choice experiment software. The team has solicited bids for the software and has considered the possibility of designing and administering the choice experiment without the software as a contingency, which would be more time consuming in both the design and data analysis. The team will make a decision by the mid-January regarding whether to shift the budget to cover the cost of new software.
a.) Treatment: We use iterative operational cost analysis to track processing, storage and marketing costs of scale-production to assess costs.
b.) Methods: Establish data collection protocols.
In Fall 2018, UMass researchers and FCCDC staff began to discuss the process of transforming food processing business management data into data that will allow researchers to analyze scale production of frozen locally grown and processed blueberries and spinach. The goal of these conversations was to identify the current state of data collection, the gaps in data collection, and establish the process through which the team would develop new data collection protocols both for the purposes of achieving the goal of this research project and to provide the FCCDC with the tools to make strategic decisions regarding frozen product development in the future.
The team began to map out the current characteristics of the data associated with frozen processed product. Initial discussions focused on three main categories of data: Labor (hiring, managing, payroll); Inputs (capital/fixed expenses, variable expenses); and Sales. The process began by discussing the timeline for procuring one hypothetical batch of product. During the discussion, the UMass researcher identified key expenses and noted the individual FCCDC staff associated with tracking and managing these expenses. From there, the team categorized the expenses, including those expenses that are currently tracked and those that are not but should be for cost analysis.
Different individuals within the FCCDC manage different aspects of each category. Each individual currently tracks information required for the task for which they are responsible. In some cases, the information required for internal purposes is sufficient for researchers’ purposes, but in other cases it is not. The team began preliminary conversations regarding what currently tracked information will need to be shared among FCCDC staff, what additional information will need to be tracked, and how to begin the process of tracking that information.
The team translated the identified expenses incurred into a list of the data needs, and discussed which FCCDC staff handled those data, or which FCCDC staff would need to begin tracking those data and how to begin that process. The Business Development Specialist (BDS) at the FCCDC agreed to begin discussions with other staff at the FCCDC to discuss the data needed and begin to develop spreadsheets, using the lists that the team had created and internal knowledge of the current FCCDC data tracking structure and process. The team agreed that the BDS would request regular meetings with other FCCDC staff to ensure that the discussion stayed on track, and that the UMass researcher could initiate and participate in those regular meetings if needed.
Since an important part of cost analysis is data from prior years’ production of wholesale frozen products, it was agreed that a good place to begin developing the new data collection would be to review past frozen processing batches and establish newly proposed protocols in the context of new wholesale production. There are a few reasons for this strategy. First, a review of past batches will allow researchers to confirm the scope of data collected and assess the quality of the data. Second, an investigation of past batches will allow a test case of how different FCCDC staff data management practices are currently organized, and identify potential barriers to coordinating additional future data tracking. Third, since wholesale frozen product is an established product line, the addition of new data collection protocols can be tested first with wholesale production, thus identifying potential unanticipated problems with the newly established data collection protocols for retail products. Finally, using concrete examples from the past could help anchor the new data collection protocols in a concrete example, hopefully bridging the gap between practical application and academic goals of new data collection.
c.)Data Collection: Once protocols are established, we will use both historical community partner retail sales data, and FPC costs data and prototype processing/production data, including inputs used at each stage for each production activity. We assess costs of each processing stage independently so facilities with different production capacities can swap out stages and independently assess their own costs, as needed.