Executive Summary

For this project, we examined different European regions to determine where the retailer, Whole Foods, should strategically expand their market. We analyzed market segments based on certain qualities of the retailer that consumers value, such as store atmosphere, price, and service quality. We divided our data into two segments: segment 1, which comprises 30% of regions in 7 European countries, where price is the most important factor when buying groceries. Since Whole Foods is known for its premium pricing, we decided segment 2 would be the best segment to target. Segment 2 values service and atmosphere the most, while price is the least important factor for these consumers. Besides, Whole Foods should enter Germany and France first because they contain 55% of regions of segment 2.

Introduction & Background

Whole foods are known for having very pricey but high-quality organic products, and it is one of the biggest grocery retailers in the United States. The target customers for Whole Foods are affluent, health conscious and care about the way their food gets produced. Whole foods understand the growing demand for organic grocery products, so they are looking to expand their stores internationally. Whole Foods would like to better understand the shopping preferences of different consumer segments to determine which areas they will be the most profitable. We analyzed store image data across seven countries: Belgium, France, Germany, Italy, Netherlands, Portugal, and Spain, to determine which segments share similar shopping preferences and identified regions that better align with Whole Foods positioning strategy.

Data and Methodology

The data used in this analysis is from a survey. The survey collected the ratings of the store image and three drivers of the store image - service, atmosphere, and price - from 137 households in 105 regions in seven countries on the food retail stores they visit most often. The rating is on a 1-7 Likert scale. (Table 1)

We analyzed the data file in GLIMMIX and used Microsoft Excel to display model estimation results and posteriors and region predictions. To identify market segments and give advice for Whole Foods on entering the market, we should decide: [1] the number, size, and distribution of segments; [2] the most important store image driver in each segment; [3] the best segment market for Whole Foods. We used linear regression to model the effect of service, atmosphere, and price on the overall store image. Afterwards, we determined the two segments based on the lowest BIC value and the optimal model based on the highest R2 value. According to the posterior membership probabilities shown in the Excel sheet, we divided each region into one of the two segments based on the cutoff of 0.5 probability. Besides, the entropy statistic helped us to determine that we had made a better separation of segments. Finally, we obtained the desired store image value of Whole Foods in these segments and identified the regions with the best opportunity to enter based on estimation results of the Mixture Regression Model and the calculations in the Excel.

Key Findings

Optimal Number of Segment: After running multiple regression models, we found that there is an optimal number for segment 2 from the results of the plot of statistics against the number of segments. (Table 4) Also, the numeric value of the model for different segments from the statistics tab for 1 to 5 Segments, including the BIC value. (Table 5) The observation indicates that segment 2 is selected based on the lowest value of Bayesian Information Criterion (BIC plot). With the optimum segment number of 2, we run again the trials of 2 segments to find the most suitable model with optimum Entropy and Model Fit (R2 square). The numeric values of the models are shown under the “statistics tab” in the software. The most favorable is 2 segment model and primarily has the lowest BIC, the highest entropy, as well as the highest R2 (Table 6)

Optimum Segment Selection: For the two segment multiple regression models, each parameter listed has a P-value less than 0.05, (Table 2) indicating each parameter is statistically significant. But the parameter values are different, where segment 2 has a comparatively lower parameter for price, this could be interpreted as segment 1 is more price sensitive than segment 2. While segment 2 has a relatively higher parameter value on service and atmosphere. Indicating the segment 2 area cares more about quality of service and shopping experience. Also, people who care about service and experience contribute roughly 70% of the segment size and people sensitive to price only had roughly 30%. (Table 2) segment 1 had a larger parameter, meaning they are currently more satisfied with existing stores, and would reduce the interest to have a new retail outlet. The overall predicted store image on a scale of 1-7 segment 2 is higher (Table 3), scale settings were based on average image for different parameters and the positioning statement for Whole Foods. Whole Foods wants to be the leader in quality food business, connected with parameters “atmosphere” and “service” The average store image for “atmosphere” and “service” are 5.47 and 5.54. With the mission being the leader, we round up to 6 each, another reason to not set a 7 because upper scale European competitors also have a strong local presence, most of them provide specialty and that also leads to second high price index. So, we would keep it as 2. To conclude, based on the analysis, due to segment size and satisfaction index of the current store, Segment 2 would be the most optimum choice for Whole Foods. The characteristics of this segment are caring for service, atmosphere but less price sensitive. (Table 3)

Choice of the Optimal Regions: We estimate the two segments of the data collected from across 105 regions and 7 countries by Posterior Membership Probability (Table 8). According to the calculation of 2-segment model, the demonstrations of two segments are explained as below:

Findings of Segment 1: The estimation result indicated that 30% of regions in seven European countries fall into segment 1, representing 25 regions of seven European countries that are positioned under segment 1(Table 9). Under Segment 1, prices are the most important, estimated value for 0.297, followed by service (0.269) and atmosphere (0.157) (Table 2). Based on estimation of posterior membership probability for segment 1 (Table 8), the regions that are under Segment 1 are distributed as 7 regions are in Germany, 6 in Spain, 3 in Belgium, Italy and Netherland, 2 regions in France and 1 region in Portugal (Table 7).

Findings of Segment 2: The result indicated that 70% of regions in seven European countries belong to segment 2. 80 regions are positioned under Segment 2(Table 9). Under the scenario of segment 2, Service has the most proportion of coefficient (0.355), followed by atmosphere (0.303) and then price takes the least (0.24) (Table 2). According to the posterior membership probability (Table 8). Most of the regions of segment 2 are in Germany, accounting for 30 regions. For other countries’ regions distribution, 14 regions are in France, 10 in Italy, 9 in Netherland and Spain, 5 in Belgium, and 3 in Portugal (Table 7). From the findings we can conclude grocery stores in Segment 1 have high sensitivity to pricing; Segment 2 prefers better service and atmosphere which is suitable for Whole Foods’ target market.

Conclusions and Recommendations

We recommend Whole Foods should target segment 2 regions, where consumers are less price sensitive and focus more on service quality and shopping atmosphere. As we have analyzed above, Whole Foods have a competitive advantage in service and atmosphere. Successfully rolling out stores near those customers will greatly help Whole Foods’ strategic expansion in Europe. Germany and France have 55% of the regions in segment 2(Table 7), which Whole Foods’ store images are more favorable than other competitors. We recommend Whole Foods to build distribution centers close to regions’ population centers, especially along the France/Germany border where Alsace, Koln, Unterfranken, Stuttgart, and Freiburg (Table 9) are located nearby for optimal logistics and open stores afterwards in those 2 countries.

To launch Whole Foods successfully in Europe, we recommend improving and investing more on service quality and shopping atmosphere for our targeting segment and minimizing the less favorable impact of high prices. Below are our recommendations:

Service: Since European countries are usually densely populated compared with the US, Whole Foods can partner with local delivery companies to offer express delivery service. People tend to shop more and more online during this pandemic, having express delivery service will also provide people with a safer shopping experience. Staff should also receive adequate training to communicate this service to customers visiting the offline stores.

Atmosphere: Whole Foods stores within the US have similar store layout and interior decoration. European countries are culturally more diverse than the US, we recommend Whole Foods to decorate their stores with consideration of local preferences while maintaining their overall decoration theme.

Price: Whole Foods Market is known for its premium pricing, although many of the regular customers of Whole Foods are not price sensitive, to attract customers from competitors, Whole Foods should offer more sales on it products and at the same time, promote its organic, healthy and environment conscious concepts to customers to justify its higher price.

Limitations: Store images are dependent on various variables, not limiting to service, atmosphere, and price. Further studies should be conducted to identify other factors i.e., store location, advertising.