Food Data
Mean Cuisine: Why One Average Cannot Describe Plant-Based Meat
A large blinded tasting found a real average gap between plant-based and animal products, but also huge differences by format and product. The interesting story lives inside the average.
An average is excellent at making a crowded table look tidy. It is less talented at telling you whether any particular bite will be delicious. A 2026 peer-reviewed study of plant-based and animal meat products makes that distinction unusually clear: the broad average showed a sensory gap, while some leading products and formats came surprisingly close to their comparison samples.
The researchers ran blinded, randomized tastings in restaurant locations in San Francisco and New York City. Between November 2024 and January 2025, 2,684 participants evaluated products one at a time across 14 categories. The study included 121 plant-based products, 14 high-selling animal benchmarks, more than 11,000 plant-product evaluations and over 800,000 data points. Participants scored overall liking, flavour, texture, appearance, similarity and purchase intent on standardized scales.
First, the wide-angle view: on a seven-point liking scale, animal benchmarks averaged 5.7, the leading plant-based product in each category averaged 5.0, and all tested plant-based products averaged 4.3. Every animal benchmark led its category average. That is a meaningful result, but stopping there would hide most of the useful detail.
The gaps varied sharply by format. The leading unbreaded plant-based filet was only 0.1 point behind its animal benchmark in overall liking. The gaps were 0.2 for nuggets and 0.3 for burgers. At the other end, the leading plant-based bacon and steak samples trailed by 2.0 and 1.5 points. In other words, “plant-based meat” did not behave like one single product category in the tasting room.
Individual opinions were wider than the averages, too. For the leading unbreaded filet, 48 percent of participants rated the plant-based sample the same as or better than the animal benchmark. The figure was 47 percent for the leading burger, followed by 42 percent for breaded filets, 41 percent for nuggets and 40 percent for breakfast sausages. A near-tie for many tasters can sit inside an average that still favours one side.
The study also looked at which descriptions moved with higher or lower liking. Savoriness, a cleaner aftertaste, juiciness and tenderness emerged as high-value areas, while off-flavours, chemical notes, dryness, crumbliness and mushiness were linked with weaker scores. These are associations within this dataset, not a universal recipe for making every product better, but they offer more practical direction than the vague instruction to “make it taste more realistic.”
For a shopper, the calm takeaway is not that every alternative has caught up, or that none has. It is that the specific format and product matter. Disliking one steak-style product tells you little about a nugget, burger or breakfast sausage from another maker. Even within a category, the leading product often performed well above the plant-based average.
There are boundaries around the finding. The researchers tested selected animal-style products in two US cities, prepared them according to manufacturer instructions and allowed condiments when appropriate. The study measured sensory response and purchase intent, not nutrition, price, ingredient simplicity or environmental impact. Products also change, so a benchmark is a snapshot rather than a permanent league table.
One especially useful feature is that the team released the sensory data publicly. That lets other researchers inspect the category differences instead of relying on a single headline number. The result is not a simple victory speech or a gloomy verdict. It is something tastier for curious readers: a map showing where the field is close, where it is not, and why the mean is only the beginning of the meal.
Sources
You may also like these
