Key takeaways

5 minute Podcast

Project slide deck

The problem: Veganism fails at the supermarket

Imagine you have just turned vegan and are doing your first grocery shop. What used to take 10 minutes now takes half an hour. You read ingredient lists line by line, hesitate over unfamiliar additives, and still leave unsure whether you made mistakes. When you get home, you discover that one product contains eggs. Frustration sets in. The experience feels mentally exhausting rather than empowering.

This is not a niche story. For many people, veganism does not fail because of ethics or motivation. It fails because everyday tasks become cognitively demanding. Every shopping trip requires sustained vigilance. This constant friction quietly erodes motivation, until quitting feels like relief.

Behavioral science has a name for this: high friction environments suppress otherwise motivated behavior.

Convenience is the weak link

Large-scale data confirms this pattern.                

Faunalytics (2014, 2025) finds that 84% of people who attempt plant-based diets eventually quit, with inconvenience among the most commonly cited reasons. Cross-cultural consumer research shows the same result in both Western and Eastern contexts. Bryant (2019) identifies lack of convenience and ease as the most negatively perceived attributes of vegetarian diets.

Recent data from the China Vegan Survey 2024 reinforces this conclusion: inconvenience is the leading barrier preventing flexitarians from adopting a vegan lifestyle.

In short, veganism is often not a moral problem, but a usability problem.

The solution: Turn the supermarket vegan with a click of a button

Now imagine the same online supermarket, but with a simple toggle at the top of the page: Vegan Mode: ON.

With one click, all non-vegan products disappear. What remains is the full assortment of vegan options across every category: staples, snacks, sauces, frozen foods, household items.

Nothing about the user’s motivation has changed. The environment has.

This is a classic choice architecture intervention: instead of asking people to exert more willpower, we redesign the environment so the desired behavior becomes the default.

Strong consumer validation

Survey data from the China Vegan Survey 2025 shows overwhelming support for a vegan filter among Chinese vegans and flexitarians:

• 94% say it would make vegan shopping more convenient
• 95% say it would save time and effort
• 91–92% say it would support both adoption and long-term adherence
• 91% say it would make them more likely to shop at that supermarket

This suggests the Vegan Filter is not only theoretically helpful, but strongly desired by its target users.

How the Vegan Filter could work

Core mechanism and partnership model

The Vegan Filter is designed as a collaborative intervention with online supermarkets, not a third-party workaround. The core idea is to work directly with the retailer and provide them with high-quality product classification data that enables a vegan filter within their existing platform.

Once integrated, the supermarket can offer an optional “Vegan Mode” that hides all non-vegan products, allowing users to browse the full store as if it were entirely vegan.

Importantly, this collaboration does not require changes to inventory, suppliers, pricing, or product availability. It only adds a product-tagging layer that plugs into existing filtering or category APIs already used by large e-commerce platforms.

Data creation: two complementary approaches

1. Paid expert review (best suited for pilots with retail partners)
In collaboration with a partner supermarket, a small paid team reviews the full product catalog (around 10,000 items) to determine vegan status based on ingredient lists and manufacturer information.
This curated dataset is then handed directly to the supermarket for integration and is periodically updated to account for reformulations and new products.

2. Community-powered aggregation (scalable with retailer oversight)
To scale beyond pilots, vegan product data can be aggregated from the vegan community through a dedicated reporting platform. To ensure reliability for retail partners, classifications are assigned confidence scores based on user agreement, contributor reliability, and historical accuracy. Only high-confidence data is shared with supermarkets.

In practice, a hybrid model is likely most effective: paid review enables fast, high-accuracy deployment with a supermarket partner, while community aggregation supports long-term scaling across platforms and regions.

Why collaboration matters

Direct collaboration with supermarkets is what makes the Vegan Filter uniquely powerful:

• It embeds the intervention at the point of purchase
• It reaches all shoppers
• It allows immediate, platform-wide impact at very low marginal cost

This is supported by consumer data. In both the China Vegan Survey 2024 and 2025, most respondents report that vegan certification would make them more likely to prefer supermarkets and online supermarkets over competitors.

This partnership-based approach turns a simple filter into a scalable behavioral intervention. For retailers, this reframes the Vegan Filter from a values-based feature to a customer acquisition and retention tool.

Why It Could Work

A 2024 randomized controlled trial published in PNAS, involving 2,359 shoppers, found that grouping vegan products into a separate category increased plant-based selection by 25%(Katare & Zhao, 2024).

A vegan filter goes even further. Instead of adding a category, it removes all non-vegan options with a single click, reducing search costs and decision fatigue across the entire shopping journey.

If modest categorization effects already produce large behavioral shifts, a full-store vegan mode is likely to have even stronger effects.

Importantly, grocery shopping in China is already highly digital. In the China Vegan Survey 2025, 69% of respondents report buying groceries online often or always.

Impact estimation

We estimate that The Vegan Filter could cut the convenience barrier roughly in half by addressing the “supermarket barrier,” one of the largest friction points for new vegans.

Assuming:

• 16% baseline vegan retention (Faunalytics, 2016)
• 33% of dropouts driven primarily by inconvenience (China Vegan Survey 2024)

This intervention could plausibly increase long-term vegan retention from 16% to 30%, effectively almost doubling the sustained vegan population in affected areas.

Even partial success would translate into very large downstream effects.

Why this matters for effective altruism

The vegan filter sits at the intersection of behavioral economics, tech infrastructure, and animal advocacy. It is:

1. Neglected

Although inconvenience is one of the main reasons people abandon vegan diets, very few online supermarkets offer a comprehensive vegan filter across the entire store. Most existing solutions are fragmented, leaving a clear and underexplored gap at the point of purchase.

2. Tractable

The Vegan Filter is technically simple to implement. It relies on product tagging and filtering infrastructure already used by large e-commerce platforms and does not require changes to inventory, pricing, or supply chains. Pilot implementations can be deployed incrementally with minimal engineering effort.

3. Large scale

Once integrated, a vegan filter can be rolled out across an entire platform with near-zero marginal cost per additional user. A single partnership with a large online supermarket can reach millions of shoppers repeatedly, making it easy to expand across regions and platforms without proportional increases in cost or complexity.

Even a modest increase in vegan retention (say, 10%) across large e-commerce platforms could translate to tens of millions fewer animals consumed annually - for a tiny marginal cost.

This could be one of the most cost-effective interventions to reduce animal suffering, especially if open-sourced and replicated globally.