Whole Foods Heat Map Tracks Stores at Risk of Unionization Using Diversity and Demographics

On April 20, 2020, Business Insider revealed that Amazon-owned Whole Foods had created an interactive “heat map” system that tracked and scored all 510 of its stores based on their risk of unionization. The sophisticated surveillance tool calculated risk scores using over two dozen metrics including employee racial diversity, turnover rates, “loyalty” scores, proximity to union offices, and local economic conditions like poverty rates and unemployment. The leaked documents stated the system’s purpose was to enable “resources to be funneled to the highest need locations, with the goal of mitigating risk by addressing challenges early before they become problematic”—corporate euphemism for suppressing union organizing before it could develop. The revelation exposed how Amazon deployed data analytics and algorithmic surveillance not just to optimize logistics, but to predict and prevent workers from exercising their legal right to organize.

The Heat Map System Architecture

The heat map system represented a sophisticated application of Amazon’s data analytics capabilities to union suppression. Each of Whole Foods’ 510 stores received an overall “risk score” calculated from three main categories of metrics:

1. External Risks The system tracked factors outside the store that might correlate with unionization:

• Size of local union membership in the area
• Distance in miles between the store and the closest union office
• Number of charges filed with the National Labor Relations Board (NLRB) alleging labor law violations in the region
• A “labor incident tracker” logging organizing and union activity in the area
• Percentage of families in the store’s zip code living below the poverty line
• Local unemployment rate

2. Store Risks The system analyzed internal store characteristics correlated with unionization likelihood:

• Employee racial diversity levels
• Average employee compensation
• Total store sales volume
• Rates of workers’ compensation claims
• Employee turnover rates
• Internal “loyalty” scores

3. Team Member Sentiment The system incorporated data from internal employee surveys measuring worker satisfaction and engagement. According to the leaked documents, “sentiment” data was “likely to be the first score to improve based on your efforts,” suggesting it was most responsive to anti-union interventions.

Diversity as Union Risk Factor

One of the most disturbing findings from the leaked documents was the system’s treatment of workforce diversity as a union risk factor. According to Business Insider’s reporting based on the internal documents and interviews with five people with knowledge of the system, stores with lower racial diversity showed higher unionization risk scores.

This correlation likely reflected Amazon’s analysis that more homogeneous workforces—particularly in areas with strong labor traditions—were more likely to organize collectively. However, the implication that Amazon was tracking racial diversity as a metric for union risk raised troubling questions about whether the company might actively seek to increase diversity not for equity or inclusion purposes, but as a union suppression tactic.

Labor researchers have noted that employers sometimes deliberately cultivate workforce diversity with high turnover rates and language barriers precisely to make collective organizing more difficult. The heat map’s inclusion of diversity metrics suggested Amazon may have been systematizing this approach, using data analytics to identify which stores’ demographic compositions made them more vulnerable to unionization.

Poverty and Economic Precarity as Risk Indicators

The system’s tracking of local poverty rates and unemployment as risk factors revealed Amazon’s understanding that economically precarious workers might be more likely to organize for better wages and conditions. Stores in areas with higher poverty rates or higher unemployment showed higher risk scores—suggesting Amazon recognized that workers with fewer economic alternatives might be more motivated to collectively bargain for improvements.

The cynical logic was clear: workers facing economic desperation were risks to be monitored and managed, rather than workers deserving better compensation and working conditions. Instead of responding to economic precarity by improving wages or benefits proactively, Amazon’s system enabled the company to identify stores where such conditions might lead to organizing, and deploy suppression tactics preemptively.

Proximity to Union Offices as Threat Indicator

The heat map’s tracking of physical distance to union offices demonstrated Amazon’s concern that existing labor infrastructure could facilitate organizing. Stores located closer to union offices received higher risk scores, reflecting the company’s recognition that unions with nearby physical presence could more easily provide organizing support, resources, and expertise to workers.

This metric revealed Amazon’s view of unions as external threats to be geographically monitored rather than as legitimate representatives of worker interests. The system essentially created a buffer zone around union offices, triggering heightened surveillance and intervention at stores that fell within range of union support networks.

“Funneling Resources” to High-Risk Locations

The leaked documents stated the heat map system’s explicit purpose: “This early identification enables resources to be funneled to the highest need locations, with the goal of mitigating risk by addressing challenges early before they become problematic.”

The corporate euphemism “mitigating risk” clearly referred to preventing unionization. The phrase “addressing challenges early before they become problematic” indicated the system was designed for preemptive intervention—Amazon would deploy anti-union resources to high-risk stores before any actual organizing occurred, suppressing potential union activity at the earliest possible stage.

What “resources” meant in practice likely included:

• Anti-union managers or consultants sent to high-risk stores
• Increased surveillance of worker communications and behavior
• Mandatory anti-union meetings or training for workers
• Targeted improvements in working conditions or benefits designed to reduce organizing pressure
• Enhanced HR presence to quickly address worker grievances before they could motivate collective action

The system’s predictive approach meant workers at high-risk stores would face intensified anti-union pressure even if they had never considered organizing, purely because algorithms had identified their store’s characteristics as correlating with unionization potential.

Timing: COVID-19 Pandemic Context

The Business Insider revelation came in April 2020, during the early months of the COVID-19 pandemic when Amazon warehouse workers were protesting dangerous working conditions and inadequate safety measures. Just weeks earlier, Amazon had fired Christian Smalls for organizing a COVID safety protest at the Staten Island JFK8 facility, and leaked memos had revealed Amazon executives’ plan to smear Smalls as “not smart or articulate.”

The heat map revelation during this period demonstrated that even as workers faced deadly pandemic conditions, Amazon was focused on preventing them from organizing collectively to demand better protection. The system’s existence showed that Amazon’s aggressive response to COVID-era organizing was not improvised, but drew on existing surveillance infrastructure designed to predict and prevent worker organizing.

Workers’ Compensation Claims as Union Risk Metric

The inclusion of workers’ compensation claims as a risk factor revealed particularly cynical logic: stores where workers were injured on the job at higher rates showed elevated union risk scores. Instead of treating high injury rates as a problem requiring improved safety measures, Amazon’s system flagged them as indicators of potential organizing.

The logic presumably was that workers who had been injured, or who worked alongside injured coworkers, might be more motivated to organize for better safety conditions. But Amazon’s response was not to improve safety proactively—it was to monitor these stores for organizing and deploy suppression tactics.

Company Response and Denial

After Business Insider published the story, Amazon did not directly address the heat map system but issued a generic statement about its commitment to workers. The company did not deny the system’s existence or contest the accuracy of the leaked documents, instead attempting to shift focus to its wage and benefit offerings.

Labor advocates and worker rights organizations condemned the heat map as evidence of systematic, algorithmically-enabled union busting. The revelation that Amazon was using sophisticated data analytics to predict and prevent organizing seemed to many observers to violate the spirit if not the letter of labor laws protecting workers’ right to organize.

Legal and Ethical Questions

The heat map system raised significant legal questions under the National Labor Relations Act, which prohibits employers from surveilling workers’ union activity or interfering with organizing. While Amazon might argue the system analyzed publicly available data and workplace metrics rather than surveilling individual workers’ organizing, labor law experts suggested the system’s purpose—preventing workers from exercising their legal right to organize—could constitute illegal interference.

The system also raised questions about algorithmic bias and discrimination. If the system treated workforce diversity, poverty rates, or injury rates as risk factors justifying increased anti-union intervention, it might be perpetuating discrimination by targeting stores with particular demographic or economic characteristics for enhanced union suppression.

Significance: Surveillance Capitalism Applied to Labor Suppression

The Whole Foods heat map represented a disturbing evolution in corporate union busting: the application of sophisticated data analytics and predictive algorithms to suppress worker organizing. Rather than responding to actual organizing efforts, Amazon had created systems to predict where organizing might occur and intervene preemptively.

This approach transformed union suppression from a reactive to a proactive enterprise. Workers at stores flagged as high-risk would face anti-union pressure before they had even considered organizing, based purely on algorithmic predictions derived from demographics, economics, and workplace characteristics.

The heat map also demonstrated how Amazon’s core competencies in data analytics, logistics optimization, and predictive modeling were being deployed not just to improve customer experience or operational efficiency, but to prevent workers from collectively bargaining for better conditions. The same technological sophistication that enabled Amazon to deliver packages with remarkable speed was being used to surveil and suppress its workforce.

The system revealed that Amazon viewed worker organizing not as a legitimate expression of labor rights, but as a risk to be measured, monitored, and mitigated through algorithmic intervention. Workers were not people with rights to organize, but data points generating risk scores that triggered suppression protocols.

Broader Context of Amazon Anti-Union Infrastructure

The heat map was part of Amazon’s comprehensive anti-union infrastructure that included:

• 2018: Leaked training video teaching managers to identify “warning signs” like workers using terms like “living wage”
• 2020: Firing Christian Smalls and smearing him as “not smart or articulate” after organizing COVID safety protest
• 2021: Spending millions on anti-union consultants and engaging in illegal tactics to defeat Bessemer union drive
• 2022: Refusing to negotiate with Staten Island’s JFK8 union despite workers voting to organize

The heat map system provided the analytical foundation for these tactics, enabling Amazon to identify which stores required most intensive anti-union intervention and to deploy resources accordingly. It represented the bureaucratization and systematization of union suppression, transforming ad hoc anti-union responses into a data-driven, predictive operation managed through sophisticated surveillance technology.