Labeling the future: ______ salmon
New research on labeling seafood that is cultivated from cells
A new technology emerges, now what do we call it? The first mobile phone was dubbed a “shoe” phone when it launched because it was only the size of a shoe box. In the early 1800s, Ányos István Jedlik was among the first to devise the motor that would later propel the electric car. He called it a lightning-magnetic self-rotor.
Today, we are on the cusp of another breakthrough in technology, this time in seafood and meat production, and we are being asked to select a name early in a technology’s lifespan. Indeed, no cultivated meat, poultry, or seafood products have been brought to market in the United States, yet an important debate is underway to determine appropriate nomenclature that is consistent with labeling laws in this country.
FDA recently called for comments on how we might label seafood made using cellular agriculture technology. The debate about what to call these products has been swirling for years, with terms ranging from cultured, clean, cultivated, cell-based, and others. Wildtype wades into this debate by submitting consumer testing conducted in partnership with The Yale Center for Customer Insights (YCCI). Research methods and underlying data are described in this post to further a data-driven public discourse on how we should think about labeling a product that looks and tastes like conventional seafood, but was not harvested from a fish.
This is a long article, so we’ve put the main insights up front in case you don’t have time to read the whole thing.
Insight 1: Based on a social media analytics review of more than 16 terms, there was no single term that a majority of people used to refer to this technology. This is still an industry in search of a name that sticks.
Insight 2: None of the tested terms were as effective at communicating the source of the seafood as the short phrase used in our control condition.
Insight 3: The short phrase used in our control condition was found to be at least as appealing as several of the terms tested.
Insight 4: While the inclusion of the short phrase in our control condition may communicate the source of the seafood most effectively, practical labeling considerations may require a shorter term. If a term had to be selected today, Wildtype would advocate for using “cell-cultivated.” It performed best among the terms tested in communicating the source of seafood, while maintaining higher levels of appeal.
The research team at YCCI, led by Ryan Whalen and including Jennie Liu and Professor Ravi Dhar, set out to answer two questions:
- Is there data to support that a “common or usual” name exists today that describes cell-cultivated seafood?
- If a term were to be selected today, which term best meets the requirements of a common or usual name, while maximizing appeal to potential consumers?
Question 1: is there a leading term in public discourse?
As a starting point, the YCCI team conducted a search to measure the use of various terms in public discourse over the course of two years. Content sources for these searches included Twitter, blogs, forums, news articles, Reddit, and Tumblr. The key terms analyzed included:
- Cellular agriculture
- Cell-based seafood
- Lab-grown seafood
- Cellular aquaculture
- Clean seafood
- Cultured seafood
- Cultivated seafood
Figure 1 below illustrates the results of the YCCI team’s sentiment analysis that was conducted using Brandwatch’s Consumer Research platform, encompassing a period of January 1, 2019 to December 14, 2020. Brandwatch is a social listening platform that enables the measurement of positive and negative sentiment associated with each term based on the language, context, and emojis used in a post.
Figure 1: Mentions of various terms & sentiment (2019–2020)
Other terms were analyzed, but received fewer than 200 mentions over this two-year period. These terms were excluded from further analysis.
Figure 1 suggests that based on Brandwatch data, no term currently accounts for a significant majority of the discussion in various online forums surrounding this technology. While cellular agriculture accounted for the plurality of mentions associated with this new technology, it is an umbrella term coined by the organization New Harvest that encompasses not only cell-cultivated seafood products, but also cell-cultivated meat and poultry products. Moreover, it is not commonly used as an adjective or modifier in front of seafood names such as shrimp, tuna, or salmon.
The distribution of the remaining terms is wide and does not show any clear leader. “Lab-grown seafood” carries significant negative sentiment: 39% vs. an average of 5% negative sentiment among the other terms and is not an acceptable term for further consideration. “Clean seafood” is potentially problematic for at least two reasons. First, it may be confused with the act of cleaning a fish to prepare it for sale or consumption. This meaning was a common association of the phrase. Second, as a modifier, the term “clean” might imply that other sources of seafood could be considered unclean, and hence would be disparaging to conventional seafood sources. Finally, “cellular aquaculture” is a trademarked term and is therefore not appropriate for further consideration as a common or usual name.
Considering these data, Wildtype agrees with FDA’s view that foods comprised of or containing cultured seafood cells are not yet in the marketplace and, therefore, do not have common or usual names established by common usage. We now turn to the question of what data-driven guiding principles should inform the selection of terms suitable for a name or statement of identity for cell-cultivated seafood.
Question 2: What would we pick if we had to select a term today?
The team at YCCI then set out to test names based on earlier research that examined several terms that might meet FDA’s requirements for a common or usual name. Two nationwide surveys were conducted in early 2021. The first survey was conducted between January 7–10, 2021 in the United States. Survey attributes include:
- Sample size: 2,455 respondents who have personally purchased tuna or salmon in the past three months
- Gender split: 49% female, 49% male, 2% other/prefer not to answer
- 46% purchased both tuna and salmon (36% salmon only, 18% tuna only)
- Excluded from the analysis were respondents who failed quality assurance tests and whose families work in seafood, market research, or advertising industries
The study’s control (i.e., that should provide a clear understanding of the source of the seafood) as depicted in figure 2 below was a short description of the technology against which test terms such as “cell-based” and “cell-cultured” would be compared. The control included an image of a plain package of either salmon or tuna, the word “salmon fillet” or “grilled tuna” depending on which treatment the respondent was given, a two-sentence description of the technology, and a customary nutrition label associated with conventional salmon or tuna.
Figure 2: Control (salmon)
The survey was conducted as a monadic study where each respondent would only see one treatment: salmon buyers would receive a salmon prompt and tuna buyers would receive a tuna prompt. Respondents who indicated that they purchased both were randomly assigned one of two fish. Seven terms were tested across both types of fish:
*Novari is a coined term (derived from the Latin verb “novo” or “to make new”) and was added to the study to assess how a term with no previous obvious connection to seafood source (conventional or cultured) would fare in terms of consumer understanding and appeal.
The study, therefore, was a 16-cell design as summarized in figure 3 below, where each respondent was randomly assigned to only one cell. The survey size allowed for approximately 150 respondents per cell for each fish, giving the survey adequate statistical power.
Figure 3: Survey one grid design
As an example of what a respondent would have seen while taking the survey, figure 4 shows one cell from the above study design: “bio-crafted” + tuna. The survey also built on earlier nomenclature studies by ensuring that a “don’t know/unsure” option was included in the response options, allowing for a clearer measure of consumer understanding by reducing guessing. An option for selecting “plant-based fillet of fish” was also added as a decoy response as well as a further quality control check to ensure that consumers could articulate the difference between plant-based and cell-cultured technologies.
Figure 4: Sample question for a tuna buyer
In addition to the question above, which measures consumer understanding, additional survey questions assessed product appeal and intent to purchase. A final question asked consumers for their understanding of whether an allergy to seafood should preclude them from eating cell-cultured seafood products.
The key results of the phase I survey are found below in figure 5. In terms of consumer understanding, the control was the most effective at communicating that the product shown was neither wild-caught nor farm-raised. 93% of tuna consumers and 97% of salmon consumers correctly distinguished the control from conventional seafood.
Figure 5: Phase I summary survey results*
*For appeal and likelihood to purchase (3rd and 4th rows of each table above), percentages represent the number of respondents who selected the top two choices on a five-point scale: “somewhat appealing” and “very appealing” and “somewhat likely” and “very likely” to purchase
Review of findings: Phase I survey
Data on the “% misunderstanding” row aggregates the percentage of respondents who believed the product was either a) wild-caught, b) farm-raised, or c) plant-based. Appendix 1 provides a more detailed breakout of these data as well as the percentage of consumers who responded that they didn’t know or were unsure. A commonly-used benchmark limits the percent of people who misunderstood a particular concept to 15% or less, net of control (not absolute). By this standard, only 3–4 terms could be considered for further testing:
The <15% misunderstanding threshold eliminated the highest performing terms with respect to appeal and intent to buy, namely Novari-cultured, cultivated, and bio-cultivated.
The critical insight from the phase I survey is that none of the names tested were nearly as effective in helping consumers identify the correct source of these products as was the short control phrase.
A hypothesis that emerged from the phase I data was that the word ‘cell’ in a name could potentially help increase understanding among consumers about the source of the seafood. A second hypothesis concerned the lack of understanding for the two terms containing the word “cultured.” Many respondents (25% for tuna and 35% for salmon) confused the terms with farm-raised fish, which may be due to the association of the word cultured with conventional aquaculture techniques in shellfish such as mussels and clams.
Based on the phase I findings, four names (plus the control) were selected for a second survey:
Bio-crafted and cell-based were carried over from the phase I study as the best performing terms after the control. Cell-cultivated is a new test term designed to assess whether the addition of the word ‘cell’ would serve to minimize the confusion associated with ‘cultivated.’ The modifier ‘cultured’ was dropped from Novari to assess the impact on understanding and appeal.
Figure 6 below summarizes the demographics of the second survey, conducted in mid-February 2021, in relation to the phase I survey. Demographics were comparable between the two surveys across most dimensions. No respondents participated in both phases of the survey.
Figure 6: Phase I & II demographic attributes
The results of the Phase II survey are summarized below. As in phase I, the control significantly outperformed the four test terms with respect to consumer understanding. 99% of tuna respondents and 97% of salmon respondents successfully distinguished the control from conventional seafood.
With the exception of Novari, the control also performed on par or better than the tested terms with respect to consumer appeal (somewhat/very appealing and somewhat/extremely likely to purchase). Excluding Novari, respondents who saw the control were on average 15% more likely to say that they were likely to purchase the product. Additional summary tables for the phase II survey are available in Appendix 2.
Figure 7: Phase II summary survey results
We embarked on this study to find a term that satisfies FDA’s requirements of a common or usual name without sacrificing consumer appeal. While some terms (e.g., cell-cultivated) performed better than others in this respect, our conclusion is that among the terms tested, a single word or a compound word on its own is inadequate.
Requirements enumerated in 21 CFR 102.5 (US Code of Federal Regulations) require that a common or usual name “shall accurately identify or describe, in as simple and direct terms as possible, the basic nature of the food or its characterizing properties or ingredients.” Because general awareness of cell-cultivated foods is currently low among US consumers, a short phrase such as that contained in this study’s control has the potential to be more effective at achieving the outcome desired from a common or usual name.
As cell-cultivated products are brought to market, awareness among consumers will likely rise, allowing for emergence of a common or usual name that allows consumers to quickly distinguish between these products and conventionally-produced seafood. In the meantime, our young industry would be better served by adopting a short phrase such as that found in the control in this study to help consumers understand how these products are made.
Our intent in publishing these data is to encourage further consumer and academic research on these important topics. The raw data sets may be made available to researchers upon request by contacting us at firstname.lastname@example.org