Small Product Changes, Big Impact? What a New Portfolio Study Found

In short: A new Nature Food analysis finds wide differences within similar product groups. Changing a small share of a portfolio could make a large difference, although the model has important limits.

This sounds like a small lever with a big impact, but it is not a measured consequence of a prescription change that has already been implemented. The study calculates scenarios: weaker-rated products are computationally replaced by products with average values from the same category and manufacturer range. For consumers, therefore, the main message is not that brand A is always better than brand B, but that the composition can vary considerably within comparable products.

What data the analysis uses

The team linked product information from the UK online market with sales data from 2019. A total of 15,673 unique products, 396 brands and 44 manufacturers were identified. Together, these companies accounted for 51.5 percent of the surveyed UK packaged food and beverage market.

The nutritional value assessment was based on the British Nutrient Profiling Model. It takes into account energy, sugar, saturated fatty acids, salt, protein, fiber as well as the proportion of certain favorable ingredients. For the environmental assessment, greenhouse gas emissions, land and water use and eutrophication were combined to a combined value. Both assessments were based on 100 grams of product.

The environmental values mainly covered agricultural production. Processing, transport and storage of the finished product were not completely included. This is an important limitation when comparing products with different cooling, packaging or processing costs.

Why similar products can perform very differently

For example, in the case of ready-to-eat foods, the range ranged from plant-based products to meat-rich variants; in yoghurt, for example, sugar, fat, salt and energy influenced the nutritional evaluation; chocolate ingredients were also able to increase the modelled environmental impact. Such differences show that the name of a category reveals little about the individual product.

The researchers also found no manufacturer consistently having the best nutritional and environmental values across all categories. One company could perform relatively well in one category and significantly weaker in another.

What ‘ten percent of the portfolio’ actually means

In the scenario, the researchers gradually replaced particularly poorly rated products with products with medium values within the same manufacturer and product category. Thus, chocolate was not assumed to be replaced by vegetables or a ready-to-use dish with loose pulses. Modeling remained closer to the existing range.

For twelve of the 44 manufacturers, more than half of the total modeled environmental potential could be achieved if at most ten percent of the product line were improved. This was true for 17 manufacturers. However, the absolute improvement was limited in relation to the major societal goals. Authors therefore stress that additional fundamental changes in product ranges, demand and nutritional patterns are needed.

Three limitations easily lost in the headline

• Older market data: The product and sales basis is from 2019, which is useful for methodical analysis but does not automatically reflect current recipes or sales.
• Category-only sales: Within a manufacturer's category, it had to be assumed that individual products were sold similarly, so that bestsellers and niche products may be weighted inaccurately.
• Model rather than intervention: The study shows potential based on established assumptions. It does not measure whether customers are buying a reformulated product, whether the taste is accepted or whether companies are actually switching.

In addition, a composite environmental value is not a complete sustainability balance sheet; working conditions, animal welfare, biodiversity, packaging, payable and regional impacts cannot be fully represented in a single value.

What the study says about clean label—and what it does not

The work evaluates nutrient profiles and modelled environmental effects, not the length of the ingredient list. A shorter list is therefore neither a research objective nor an automatic quality characteristic. Reformulation can mean reducing salt, sugar or saturated fatty acids, increasing fiber or changing the main ingredients. It may also require technical ingredients to keep flavour, durability or texture stable.

For a fair product comparison, several information should therefore be considered together: nutritional table, ingredient order, portion reality, allergens and use. Marketing terms such as natural, fit or climate-friendly do not replace this information.

Four practical checks in the supermarket

1. For comparison only, the nutritional score is particularly useful within the same product group or for the same use, for example between two cereals or two frozen dishes.
2. Look at 100 grams: sugar, saturated fatty acids, salt, protein and fiber can be compared better with the same reference size than with differently defined portions.
3. Check the main ingredients: In the list of ingredients, ingredients are generally in descending order of weight, showing whether vegetables, whole grains or pulses are characteristic or appear only as small additives.
4. Think of a simple basic product: A finished dish can be practical, but for some meals, unprocessed or little processed building blocks such as potatoes, vegetables, lentils, beans, natural yoghurt, or grains are easier to adapt to your needs.

At Kochzauber, recipes for ingredients, vegetarian dinners and one-pot dishes offer ideas on how to combine such building blocks without complicated special products. (Vegetarian meals)

Why manufacturers and retailers have greater leverage

A single purchase does not change the entire range. Manufacturers decide on recipes, product development and the selection of what is offered. Retailers, with their own brands, placement, price quotes and availability, additionally control which alternatives are readily available.

The methodology could serve as a recurring scorecard in the future according to the research team, but would require up-to-date, product-accurate sales and environmental data. Without transparent timelines, it is difficult to determine whether announced targets actually lead to better sales and less impact.

Conclusion: The portfolio matters, not just the label

The new analysis shows that there is a lot of scope already within familiar product groups. Some manufacturers could use a large part of their modelled potential on a few particularly weak products. This makes reformulation and assortment control a useful lever.

However, the number of ten percent is not a guarantee of success and is not a German market forecast. The concrete comparison of similar products remains the most reliable for everyday life. And the bigger task for the policy and the industry is: to disclose current data, to measure progress over years and not only to improve individual flagship products.

Sources and status

• Hammond et al.: Large nutritional and environmental improvements can be achieved across the product portfolios of major UK food and beverage manufacturers, Nature Food, published on 16 September 2026, DOI 10.1038/s43016-026-01420-2.
• University of Oxford via EurekAlert: Food & drink manufacturers can greatly improve sustainability and healthiness of their products with small changes, research report dated 16 September 2026. (Food & drink manufacturers can greatly improve the sustainability and healthiness of their products with small changes)
• Oxford Martin School: Overview of the study published on 16 September 2026. (Overview of the publications for the study)
• National Centre for Nutrition: Nutrient score: nutritional comparison at a glance, suggestions for meaningful comparisons within product groups. (Nutrient score: a nutritional comparison at a glance)

Author: Kochzauber editorial team. Information and update status: 17 September 2026. The article orders a British model study and does not broadcast its results to the German market.