Progress 05/15/16 to 05/14/17

Outputs
Target Audience:There are 3 target audiences for this project: SNAP consumers: 3,331 SNAP participants increased purchases of fresh fruits and vegetables with the project reaching a total of 8,688 people (based on average household size of 2.6 across CT and VT) through $64,100 in SNAP incentives. SNAP households in Vermont's Northeast Kingdom learned about the SNAP incentive program through a mailing from the VT Department of Children and Families. The VT participants included 40% of households with individuals over 60, 22% included minors and 3% included both seniors and minors. In Hartford, CT, SNAP participants learned about the program through in-store signage and promotion, and through community-based marketing efforts. Grocery stores: Six small and medium-format, independent grocery stores in VT and CT saw increased produce purchases by SNAP consumers during the first year of this project. Wholesome Wave, Green Mountain Farm-to-School and Hartford Food System provided grocery stores with 1) program implementation training; 2) support with adapting point-of-sale systems; and 3) materials for in-store and community-based marketing. CBOs/local food aggregators: Mission-driven CBOs/local food aggregators tested innovative farm-to-grocery SNAP incentive models for expanding affordable healthy food options in underserved communities. Local food aggregators received training and technical assistance on program implementation. Based on strong program data analysis, Wholesome Wave provided insights and direction to enable CBOs to shift all grocery stores to the qualifying purchase model, which drove significant incremental produce purchases among SNAP consumers. CBOs also received technical assistance to set the conditions for successful local food procurement throughout the remainder of the grant. Changes/Problems: Wholesome Wave and Green Mountain Farm-to-School had staff turnover mid-year which resulted in revising the timeline and roles and responsibilities, including moving tool creation and design into year 2. As well, the majority of the year 1 research analysis occurred early in year 2, after the first year of implementation and data collection was complete.We still met all objective milestones for year 1. Based on learnings, both successes and challenges, we are identifying and refining the tools and guides as we work toward a toolkit that can be easily replicated. Currently, we have refined: store typology; program model, including an approach to point-of-sale technology changes; evaluation tools; andmerchandising and marketing communication solutions. Thesewhich will be tested further in Year 2 to ensure effectiveness in both existing and new store rollouts. Moved incentive implementation in year 2 to August/September timing, to best take advantage of local production and to assess the full outcomes from the qualifying purchase program design prior to encouraging further investment in POS upgrades. What opportunities for training and professional development has the project provided? Nothing Reported How have the results been disseminated to communities of interest? Nothing Reported What do you plan to do during the next reporting period to accomplish the goals?The following activities are planned for the next reporting period: Adapt the program design in five stores in VT to launch the qualifying purchase model, based on the successes of that model at the store in Hartford, CT. As planned, all stores in the project will also undergo technology updates to improve and simplify processes, including point-of-sale system updates and integration of UPC codes. Wholesome Wave and the food hubs will train staff for ease of integration. Refine Local Food distribution strategy with Food Hubs, based on Year 1 data and feedback from participating stores. Determine barriers to Local Food distribution, tracking and in-store sales. Refine in-store communication plan for Year 2 roll-out, based on Year 1 experience, usage and feedback. Revise placement locations, messages and signage types to increase impact and drive participant knowledge and usage of the program. Expand successful year 1 external communication strategy at willing community locations, such as libraries, town buildings, etc - to help spread knowledge usage of the nutrition incentives. Refine store training to include POS elements, measurement and in-store communication changes. Expand training of store employees based on Year 1 learnings, to include - program knowledge of the qualifying purchase model, advocating for the program in-store and encouraging usage. Update measurement standards to reflect additional information gathered based on POS updates and new technology solutions. Any changes will be updated on the evaluation plan. In Year 2, add a new grocery store in Hartford's North End, which is in planning and concept, which will allow for further testing of materials and help refine the POS, merchandising, communication and training toolkits.

Impacts
What was accomplished under these goals? Objective 1: The first year of the Wholesome Wave Farm-to-Grocery SNAP Nutrition Incentive program successfully met objectives and ran in 6 retail sites - 1 in Hartford, CT and 5 in Northeast Kingdom, VT. Two distribution models were tested, a Qualifying Purchase model in CT and a Voucher Distribution model in VT. The Qualifying Purchase model was activated when a customer paid for a purchase that was at least $10 by SNAP/EBT and included at least 1 piece of produce. The cashier then distributeda $5 voucher for free additional produce. In the Voucher Distribution model, vouchers with instructions weremailed to 1,115 households in Orleans and Essex Counties in VT for redemption of fruits and vegetables at participating retailers. The vouchers ranged in value from $25 to $75, based household size. The programs reached 3,311 SNAP participants and 8,688 people, based on an average household of 2.6 people. $105,000 in incentives were distributed, resulting in $64,100 in redemptions, 92% of the Year 1 goal. The remaining amount will be made up in Year 2, as we have a baseline for redemption activity from Year 1 with merchandising and communication insight. The CT redemption rate was 50%, and in VT the redemption rate was 86%. Theses were increases 64% and 38%, respectively, from their pilot programs. The program is allowing people to purchase the fruits and vegetables that they need and want. The Manager of the Kingdom Market said, "A lot of people can't afford the fruit and vegetables....I remember one family with 3 or 4 kids that came in and were all excited - you'd have thought they were in the candy aisle". Diving deeper into the results, it is clear that the Qualifying Purchase model in CT had positive effects on produce purchasing by SNAP shoppers. The program grew over the period, with redemptions increasing from month 1 to 4, indicating that as shoppers became familiar they used it more. 16% of participants used the vouchers once, 32% twice and 52% 3 times or more - showing the frequency of use was high. By tracking transactions through the POS system, we found that 75% of redemptions took place on the same day as issued - with participants going back into the store after receiving the voucher to use it, spending $15, which includes $7 on produce - $2 more than the value of the voucher. The 25% of participants redeeming later spent an additional $4 on produce. The data shows that on average participants spent additional on produce when the vouchers were used. Finally, we saw significantly higher SNAP sales triggering the incentives, a total of $856,500, of which $199,700 in produce was purchased, in Year one alone. This helps to set the baseline for Years 2 and 3 for expected spend. Objective 2: The project is currently on-track for this objective. In addition to an estimated $12,000 in local products purchased by the retailers, we built buy-in among the stores, which strengthened the local procurement relationships. We successfully set up the conditions for success with the participating food hubs - Hartford Food System in CT and Green Mountain Farm-to-School in VT. In VT, 4 stores signed a contract stating that they would purchase locally for 30% of the amount of their incentives, and the actual spend equated to about 40% of total incentives. In CT, our objective was to work with the Store Manager to engage with his distributors of local food. The Manager does purchase some local produce, but is not able to fully track the typical and any increased amount ordered. In the beginning of this program, he was not convinced that local food was an important part of the program, but due to the successes he's seen with the program, he has finalized this change for Year 2. Transaction reports show that there was an actual increase in produce sales, as the CT Store Manager confirmed: "...I know our customers are buying more fruits and vegetables, because we are buying more from our suppliers..." For Year 2, merchandising and communication will be overhauled with the introduction of new program materials and merchandising support. In CT, which is an urban area, the term local food may not be understood as intended, as 100% of people surveyed responded that they bought local produce. Our assumption is that the term was misunderstood and we will work to clarify this through in-store merchandising and revised survey prompts. Objective 3: For Year 1, this objective was successfully achieved. Sales statistics noted in Objective 1 show the increased SNAP and produce sales. The detail in Objective 2 shows the improvement in local food offerings. Wholesome Wave layered best practices into the set-up and developed training materials for set-up, operational processes, staff scripts, measurement reporting tools and guides. The Food Hubs used prior expertise running pilots and other programs and worked in sync with Wholesome Wave. The Food Hubs led the communications efforts with the retailers: making local farmer connections, producing signage and communication pieces, helping stores update their POS systems and gauging readiness for launch. Wholesome Wave made a total of three in-person visits to sites during the pre-launch period to help train, set-up and merchandise with the CBO partners and two additional visits during the post launch period, as well as numerous remote conferences. In the assessment of Year 1, store effectiveness was evaluated two ways: 1) interviews with each Store Manager and CBO Managers, and 2) the NEMS-S retail environment assessment tool which assessed the quality, availability and price of healthy food, including local produce, in-store marketing efforts, and accessibility for SNAP shoppers. These evaluations provided feedback on the challenges and successes of the stores and created a baseline of effectiveness and an initial roadmap for refinement of the program design and variables. A significant learning was on the effect of merchandising and in-store signage. The Manager of the Kingdom Market in VT said, "Merchandising has been a big change and we actually have seen a big change in sales due to it......[we] didn't expect to sell as much. People have made (positive) comments...... and it's fun doing it." Indications are that the in-store merchandising and signage can be even more effective. The merchandising planning will year 2 will be more detailed and directed, and the communications more clear and specific to the program. Objective 4: The analysis of Year 1 data allowed for a solid understanding of the areas of success and variables to refine. Plans are underway to capitalize on successes and make refinements as we work towards replicability. Year 1 data shows that the Qualifying Purchase model drives significant spend at each trip - $23 per redemption on average driven by the interactive and on-going nature of the Qualifying Purchase model (spend - receive incentive - redeem - repeat). Both models drove redemption and additional purchasing of fruits and vegetables, but the Qualifying Purchase model drives additional spend and produce purchasing at a higher level. For Year 2, we will: Capitalize on the success of the QP model and introduce this model to the VT market. Test training content during the deployment of the QP Model in VT and create tools. These will be further tested during new store additions in Year 2 and Year 3. Implement a suite of in-store marketing materials, which were created based on merchandising learnings from year 1. These will be rolled out in both markets and further ideas will be tested and refined during Year 2, leading to a replicable communication toolkit Local food procurement will be improved and redeployed to overcome the challenges observed in Year 1, and outcomes will be developed into a procurement model. These learnings and refinements are leading to better content for Year 2 and as they are proven, will become the tools and models for the future of this program.

Publications