A company may describe its product as universal and report strong average performance, even though part of its audience regularly faces additional costs, inconvenience, or greater risk. The problem is often not weak demand, but the original user model around which the product's features, standards, and evaluation criteria were developed.
For decades, many industries relied on a particular set of assumptions: a male body, a direct commute between home and work, continuous employment, and a predictable income. When these scenarios are treated as the universal norm, women's experiences remain underrepresented in research and testing, and the product's limitations only become apparent after launch.
For businesses, this mistake distorts data on quality, risk, and market size. A company may interpret a customer's decision not to use a product as a lack of interest, when the real barriers are its design, conditions of use, or evaluation method. Automotive safety, urban mobility, and bank lending demonstrate how a broader view of the user experience can help correct systemic imbalances and provide a more accurate understanding of an audience the business previously assumed it was already serving.
How a Limited Experience Becomes the Universal Norm
Product development almost always begins with a simplified user model. A company determines who will use the service, what problem that person needs to solve, the conditions in which they will use it, and the features they require. This approach allows the company to limit the scope of its research and make initial decisions more quickly. However, the quality of the final result depends on how accurately the team understands the boundaries of the scenario it has selected.
When initial tests, observations, and datasets primarily describe one group, the resulting conclusions continue to shape the product even after its audience expands. They form the basis for technical standards, medical recommendations, transportation systems, and customer evaluation criteria. Over time, the origins of these rules are no longer questioned: the team works with an established norm and rarely returns to the question of whom it originally represented.
Simply including women among research participants does not solve the problem. When results are combined into a single average, consistent differences between groups can disappear within the overall figure. A high overall conversion rate does not show whether registration is equally easy for different users, just as an average injury risk does not reveal which passengers receive less effective protection from a safety system.
Companies should therefore identify in advance which characteristics may affect the outcome and account for them when collecting and analyzing data. Sex may be one relevant factor, alongside age, income, health, employment type, and conditions of use. Breaking down the data does not provide an immediate explanation, but it helps reveal areas in which a successful overall metric conceals a recurring problem.
The consequences of an overly narrow user model are particularly visible when health and safety depend on the quality of the product.
When the Standard Fails to Account for Differences Between Bodies
For decades, the primary model of an adult used in automotive crash tests was a dummy representing an average-sized man. A dummy representing a smaller woman was also used, but it covered only a limited range of body sizes and did not reproduce all the physiological differences that can affect injury patterns.
In a 2023 report, the U.S. Government Accountability Office found that the dummies in use represented too narrow a range of bodies, did not sufficiently account for physiological differences between men and women, and could not collect certain data on lower-extremity injuries. The tests themselves also had limitations: results were affected by the dummy's position inside the vehicle, its seating location, and the speed of the collision. As a result, the tests did not provide complete information about the protection offered to different groups of passengers.
Statistics from real-world crashes confirm that the consequences are not distributed equally. In January 2026, the U.S. National Highway Traffic Safety Administration published a study covering 150 injury patterns. In 26% of them, the risk for women was statistically higher. In frontal crashes, women's overall injury risk was 46% higher than men's, their risk of lower-extremity injuries was 62% higher, and their risk of injuries to the feet and ankles was 128% higher.
The severity of an injury is affected by age, body mass, driving position, vehicle design, and other circumstances, but all of these factors are part of real-world use. A manufacturer sells the same vehicle to people of different heights and body types, so testing should cover a corresponding range. Customers should not have to match the dimensions of a standard crash-test dummy to receive the advertised level of protection.
In November 2025, NHTSA published technical specifications for the advanced THOR-05F female crash-test dummy, which more accurately reproduces the body's response to a collision and contains more than 150 sensors. Automakers can already use it in their own testing, although formally incorporating the model into safety ratings and mandatory standards will require an additional regulatory process.
The automotive example shows that formal compliance with a technical standard does not necessarily guarantee the same level of safety across the actual range of customers. When the standard itself is built around a limited body model, a product may successfully pass all required tests while still providing less protection to a substantial portion of its audience.
A similar problem existed for many years in medicine, where the model of the future patient is established even earlier-during the selection of cells, laboratory animals, and research participants. The U.S. National Institutes of Health acknowledges that basic and preclinical biomedical research was traditionally conducted primarily on male cells and animals. This practice may have concealed the effect of biological sex on disease development, drug response, and treatment outcomes, because findings obtained from one group were applied much more broadly.
Today, women account for approximately half of the participants in NIH-supported clinical research, but numerical equality solves only part of the problem. Biological sex must be considered when defining the research question, choosing a model, collecting data, and analyzing results. Otherwise, differences may be discovered too late or remain hidden within an average figure.
Since 2016, the NIH has required researchers to consider sex as a biological variable in studies involving humans and vertebrate animals. Scientists proposing to study only one sex must provide a meaningful justification based on scientific literature, preliminary data, or the specific nature of the research. The absence of previous evidence of sex-based differences is no longer considered a sufficient argument, because such evidence cannot be obtained without an appropriate research design.
The conclusion for businesses, medical organizations, and research teams is the same in both cases: an average result may appear to confirm a product's effectiveness or safety while concealing substantially different outcomes among individual groups. The later these differences come to the attention of developers, the more expensive they are to correct-and the greater the likelihood that part of the audience has already experienced the consequences of incomplete testing.
In urban planning, the universal norm is not a particular type of body, but a particular pattern of movement.
When a Service Is Designed Around a Direct Commute
Transportation systems have traditionally been designed around travel between home and a place of paid employment: in the morning, a passenger travels from a residential area to a business district and returns home in the evening. This kind of journey is easy to predict because it has a clear direction and identifiable periods of peak demand, around which transportation capacity, staffing, and budgets can be organized.
Trips made by women are more likely to combine several tasks within a single journey: commuting to work, accompanying children, helping relatives, shopping, and visiting medical facilities. The World Bank notes that women in developing countries make more frequent and shorter trips on average, stop more often, and combine several errands, while men's journeys are more likely to remain direct and linear.
Differences in travel patterns directly affect the quality and cost of a service. Charging separately for every leg increases the cost of a trip chain, infrequent service outside peak hours extends travel time, and a poorly lit transfer point may make an evening journey entirely inaccessible.
The statistics, however, will show only a low proportion of female passengers. A transportation operator may interpret this as weak demand, even though some potential customers have stopped using the service because of its schedule, transfer costs, or safety conditions.
This creates an important distortion for businesses: the company measures completed use of the product but does not see the activities customers abandoned because of barriers built into the system. Demand is evaluated based on the behavior of people who managed to adapt to the existing conditions, while the potential audience remains outside the analysis.
Vienna began accounting for gender differences in urban planning based on data about actual mobility. Research showed that women in the city make almost 50% more caregiving and household-related trips than men. Half of these journeys are made on foot and are often combined with paid employment as part of a complex trip chain. Traditional planning placed greater emphasis on vehicle traffic, leaving pedestrian routes as a longstanding weak point.
Based on these findings, the city began paying greater attention to sidewalks, crosswalks, transit stops, and the lighting of spaces used during everyday journeys. The municipal lighting department launched a dedicated program to assess 200 parks. During modernization projects, it considered pedestrian paths, areas beneath overpasses, and locations where architectural features blocked the light and created poorly visible spaces.
These solutions benefit a much wider range of residents: parents with strollers, children, older adults, passengers with disabilities, and anyone whose day does not fit into a direct journey between home and the office. Analyzing women's travel patterns broadens the very definition of everyday mobility, allowing the transportation system to account for real-world usage scenarios that previously remained at the margins of planning.
Financial products are built around a different kind of norm: a continuous career history, regular income, and a long credit record.
When Creditworthiness Is Measured by a Linear Financial History
Banking decisions are based on income, assets, credit history, and business stability, which makes the criteria appear equally applicable to everyone. However, the traditional model of a reliable borrower is better at recognizing a particular professional and financial trajectory: continuous formal employment, predictable income, registered property, and a long-standing relationship with financial institutions.
Customers whose economic activity follows a different pattern may receive a lower assessment even when they operate stable businesses and consistently meet their debt obligations. This issue is particularly visible in lending to women-owned businesses.
According to 2024 data from the International Finance Corporation, women-owned businesses accounted for more than one-third of small and medium-sized enterprises in emerging markets, yet they received only 19% of the total value of loans issued by the financial institutions studied. Their average loan size was 28% smaller, even though their nonperforming loan rate was 3.6%, compared with 3.8% for the overall small and medium-sized enterprise portfolio.
These figures indicate a disconnect between the actual quality of the portfolio and the amount of financing available. The IFC attributes part of the problem to a lack of sex-disaggregated data: until a bank analyzes the performance of women-owned businesses separately, it cannot see their actual repayment discipline and continues to rely on outdated assumptions about risk.
Formally neutral criteria can produce inaccurate results when they are designed around a single economic trajectory. Career breaks, a shorter credit history, irregular income, or the absence of traditional collateral do not necessarily indicate poor creditworthiness, but a standard model may interpret them that way.
Data on cash flow, transactions, digital payments, and actual business performance can broaden the assessment. CARD Bank in the Philippines adopted this approach by introducing a new credit-scoring model for customers whom its previous methods had failed to evaluate accurately.
After piloting the system at six branches, the bank expanded it nationwide and issued more than 10,000 new loans, 93% of which went to women or women-led businesses. According to the bank, the new model also saved more than 140,000 person-days of work and reduced the number of manual data-entry errors.
CARD Bank retained its creditworthiness requirements but changed the way creditworthiness was measured. As a result, the institution gained access to customers whose economic activity had been poorly captured by the traditional model.
This case demonstrates the direct commercial value of a more accurate user model. The company corrects a bias in its assessment while simultaneously identifying a high-quality audience that its previous analytics had undervalued. In this context, a broader dataset does not weaken risk management. Instead, it allows risk to be assessed based on customers' actual behavior.
What Businesses Gain by Expanding Their User Model
The phrase "a product for women" is often associated with a separate product line, a specialized design, or an additional set of features. In the cases examined here, however, the issue is the quality of the core product, which performs its stated function less effectively for part of the audience because their experiences were not adequately considered during research, testing, and data analysis.
A similar mechanism operates in all three sectors. A company or government organization selects a limited user model, embeds it in standards and algorithms, and then evaluates demand and effectiveness within the system it has created. Women experience more injuries in vehicles, use a particular route less frequently, or receive smaller loans. The resulting figures are then interpreted as evidence of natural differences in behavior or risk. In reality, the data may be describing the consequences of decisions made during product development.
Accounting for women's experiences helps distinguish a genuine absence of demand from a situation in which customers are prevented from using a product by its design, evaluation criteria, or conditions of use. The company gains a more accurate understanding of the size of its market, the quality of its service, and the reasons some customers stop engaging with the product.
Correcting these limitations usually benefits a much broader range of users. Well-lit routes and convenient transfers help people traveling with children, older adults, and passengers with disabilities. Scoring based on actual cash flow helps entrepreneurs with short credit histories, while more diverse body models improve safety for everyone who does not match the dimensions of a hypothetical average user.
For businesses, examining women's experiences is a way to test their own product assumptions. It reveals whether a solution can function outside the familiar scenario, whether the company recognizes hidden barriers, and whether it can distinguish weak demand from a poorly designed offering.
Accounting for women's experiences is one of the most important ways to determine how fully a company understands its audience-and how justified it is in describing its product as universal.
