Executive Summary

May 21, 2026

Balance between innovation and data protection in e-commerce

Mariana Catharina de Souza; Jeffson Veríssimo de Oliveira

Summary prepared by the ResumeAI tool, an artificial intelligence solution developed by Instituto Pecege focused on synthesis and writing.

E-commerce has consolidated itself as a transformative force in the global economic landscape since its emergence in the late 1990s, driven by the expansion of connectivity and the democratization of access to the World Wide Web (E-Commerce Brasil, 2022). This evolution has redefined consumer relations by eliminating geographical barriers and allowing products and services to reach diverse audiences in real-time. However, the acceleration of digital transformation has brought complex challenges, especially regarding information security and user privacy. The constant need for innovation to generate profit and notoriety in the market requires organizations to adopt advanced data collection and analysis strategies, often creating an ethical and operational dilemma between commercial efficiency and consumer protection. Trust emerges as a fundamental element for encouraging transactions in virtual environments, as the perception of risk can inhibit the adoption of new technologies and harm brand image (Corbitt et al., 2003).

The protection of personal data in Brazil gained an essential regulatory framework with the enactment of the General Data Protection Law, Law No. 13,709/2018, which establishes rigorous guidelines for the processing of information by public and private entities (Brasil, 2018). The central objective of this legislation is to ensure that the data subject maintains control over their information, requiring transparency and informed consent. Despite the legal advancement, daily practice reveals that many consumers still accept terms of use and privacy policies without proper understanding, motivated by haste or the complexity of the legal language used by platforms. This scenario of vulnerability is aggravated by the sophistication of cybercrimes, which exploit technical and behavioral flaws to carry out large-scale fraud and leaks. Therefore, the search for a balance between digital innovation and security becomes imperative for the sustainability of the online sales ecosystem.

Technological development allows companies to build detailed customer profiles, using this information for targeted advertising, personalized product design, and dynamic pricing strategies. While these practices can improve the user experience, they also raise concerns about constant surveillance and the misuse of sensitive data. The integration of privacy and security must be seen not only as a legal obligation but as a strategic value that strengthens the relationship between brand and customer (Bioni, 2019). Research into how developers, managers, and consumers perceive these risks is fundamental to proposing models that guide the creation of procedures capable of protecting privacy without compromising technological advancement and business competitiveness in cyberspace.

To conduct this research, an exploratory and qualitative methodological approach was adopted, aiming to understand the phenomenon in its multiple perspectives and identify patterns of behavior and perception (Borges, 2000). The theoretical foundation was built through a systematic literature review, consulting scientific articles, theses, and technical documents on open-access portals. Primary data collection was structured in three distinct stages, using the Google Forms platform to ensure the agility and organization of responses. All participants were informed about the nature of the study and provided free and informed consent, ensuring the anonymity and confidentiality of the information provided during the research process.

The first stage consisted of applying a structured questionnaire to consumers of online stores, with the aim of mapping the sociodemographic profile and perceptions about the sharing of personal data. The collection instrument comprised 14 multiple-choice questions and four open-ended questions, covering dimensions such as the level of distrust, knowledge about privacy policies, and digital security habits. The analysis of this data followed the proposed thematic categorization, allowing us to identify how variables such as gender, age, and education influence the user’s willingness to provide information in exchange for commercial benefits (Bardin, 2016). The focus was on understanding the motives that lead to the acceptance of terms of use and the identification of previous negative experiences, such as financial losses or undue exposure of data.

Simultaneously, a survey was conducted with developers of e-commerce systems and applications to obtain a technical and operational perspective on the topic. The script included 10 semi-structured questions that explored the security culture in organizations, data collection methods, and ethical dilemmas faced during the development of new functionalities. We sought to understand how security is integrated into the software lifecycle and what the main obstacles are to aligning the agility required by the market with the demands of current legislation. This stage was crucial for confronting the user’s perspective with the technical possibilities and limitations of the sector, providing input for the discussion on the viability of more robust protection solutions.

The third stage of the methodology involved the application of the adapted Delphi Technique, conducted in two rounds with industry experts, including the owner of a digital marketing company and a senior developer. The objective was to reach a consensus on the most critical risks and the most effective mitigation strategies for the current e-commerce scenario. In the first round, the experts listed and classified the main security risks, describing their impacts on consumers and businesses. In the second round, participants brainstormed solutions, evaluating the implementation feasibility of each proposal on a scale of one to five, considering factors such as cost, technical complexity, and legal compliance. The systematization of these responses allowed for the construction of a model of practical recommendations for the sector (Asana, 2025).

The analysis of sociodemographic data revealed that the consumer sample profile is predominantly female, representing 66.7% of respondents, while the male demographic comprises 33.3%. Regarding age distribution, there is a significant concentration of young adults, with 75% of participants aged between 25 and 34 years. The education level is high, as 75% have completed higher education or postgraduate studies, suggesting an audience with access to information but who still face barriers in understanding digital security technicalities. As for platform usage time, 33.3% spend between 30 minutes and one hour per day on shopping websites, evidencing a deep integration of e-commerce into the daily routine of these individuals.

The results indicate a clear willingness of consumers to share personal data in exchange for immediate financial advantages. Approximately 54.2% of respondents stated that they provide information to obtain discounts and low prices, while 29.2% do so to participate in loyalty programs or receive personalized recommendations. This trend reflects how convenience and economic benefit often outweigh privacy concerns. However, distrust remains a determining factor in purchasing behavior: 91.7% of participants admitted to having abandoned a transaction due to a lack of trust in the platform. The main reasons for this abandonment include suspicious website design (70.8%), negative social media reviews (54.2%), and a lack of clarity in privacy policies (45.8%).

Experience with security incidents is a reality for part of the sample, with 16.7% reporting financial losses due to data breaches or unauthorized purchases. This data reinforces the perception of risk, which is the biggest concern for 87.5% of people when providing information online. Despite this apprehension, there is a critical gap in digital literacy: 58.3% of respondents admit to accepting terms of use without prior reading to expedite access, and 45.8% have only a superficial understanding of how cookies work and the destination of their data. The lack of knowledge about the rights guaranteed by legislation is evident, which shows that the existence of the law does not, in itself, guarantee effective protection if there is no user education (PROCON, 2021).

On the technical level, developers reported that data collection is a fundamental pillar for innovation, enabling the creation of intelligent and personalized solutions that increase commercial efficiency. The data collected ranges from basic contact and financial information to technical navigation data, such as IP address and location. However, professionals recognize that the balance between innovation and compliance is a constant challenge. The rush to launch new products to the market and the high costs of implementing robust security systems are pointed out as the main obstacles. The adoption of practices such as “Privacy by Design” and the integration of security into the software development lifecycle are seen as necessary paths, but they still face cultural resistance within organizations.

The application of the Delphi Technique allowed for the identification of three categories of priority risks: digital fraud and scams, data breaches, and cyberattacks. For experts, the impacts of these risks are multidimensional, affecting consumers’ financial health, privacy, and psychological well-being, in addition to causing severe reputational damage and legal sanctions for companies. As consensual solutions, the adoption of constant multi-factor authentication (MFA), the use of behavioral biometrics, and the implementation of real-time transaction analysis supported by artificial intelligence stood out. The feasibility of these solutions was positively evaluated, with a score of 4.50 out of five, indicating that the technology is available, but its effectiveness depends on a change in business mindset that ceases to view security as an expense and begins to see it as an essential investment.

The discussion of the results points to the need for “invisible security,” which protects the user without creating bureaucratic barriers that harm the shopping experience. The strategic use of advanced technologies, such as device monitoring and end-to-end encryption, can offer proactive defense layers against automated threats that grow annually in the sector (Axur Platform [AXUR], 2025). Furthermore, vulnerability management and strict access control, based on the concept of least privilege, are fundamental measures to mitigate the risk of internal and external leaks (International Business Machines Corporation [IBM], 2024). Transparency in customer communication should be prioritized, transforming complex privacy policies into didactic and accessible information.

The future scenario of e-commerce requires companies to invest in more robust budgets for cyber defense, following the global trend of increased spending on information security (Gartner, 2025). The resilience of businesses in the digital economy will depend on organizations’ ability to anticipate threats and educate their consumers, creating an environment of mutual trust. The integration of protection practices from the conception of digital products not only ensures compliance with legislation but also consolidates itself as a competitive differentiator in a market increasingly attentive to privacy risks. The challenge lies in converting security from an operational cost into a central strategic value for the sustainability and growth of online sales platforms.

It is concluded that the objective was achieved by identifying that the balance between digital transformation and data security in e-commerce depends on an integrated approach involving technology, legislation, and education. The investigation showed that, although consumers value immediate benefits, the lack of trust is the main obstacle to loyalty and purchase completion, highlighting that security is an inseparable component of the user experience. At the macro level, the effectiveness of data protection requires a robust regulatory framework and incentives for companies to adopt high standards of governance. At the micro level, it is fundamental to promote an organizational culture of security that prioritizes privacy from the development of systems, combined with awareness campaigns that empower consumers to act proactively in the digital environment. The integration of these pillars allows technological innovation to continue driving the market without compromising the fundamental rights of citizens, transforming data protection into a pillar of sustainability for the contemporary digital economy.

Bibliographic References:

ASANA. 2025. O que é um fluxograma? Aprenda os símbolos, tipos e como interpretá-los. Disponível em: <https://asana.com/pt/resources/what-is-a-flowchart>. Acesso em: 3 jun. 2025.

AXUR. 2024. Threat Landscape 2024/2025: Relatório de Ameaças Cibernéticas. Disponível em: <https://cmsb.nortonlifelock.com/sites/default/files/2025-02/Q4%20Gen%20Threat%20labs%20Whitepaper%20%283%29.pdf>. Acesso em: 12 jun. 2025.

Bardin, L. 2016. Análise de Conteúdo. Edições 70, São Paulo, SP, Brasil.

Bioni, B.R. 2019. Proteção de Dados Pessoais: A Função e os Limites do Consentimento. Forense, Rio de Janeiro, RJ, Brasil.

Borges, N.B. 2000. Métodos e técnicas de pesquisa em psicologia. Revista Estudos de Psicologia (PUC-Campinas) 17(2): 80-81.

Brasil. 2018. Lei nº 13.709, de 14 de agosto de 2018. Lei Geral de Proteção de Dados (LGPD). Diário Oficial da União, Brasília, 15 ago. 2018. Seção 1, p. 1.

Corbitt, B.J.; Thanasankit, T.; Yi, H. 2003. Trust and E-Commerce: A Study of Consumer Perceptions. Electronic Commerce Research and Applications 2: 203-215.

E-Commerce Brasil. 2022. Evolução do e-commerce. Disponível em: <https://www.ecommercebrasil.com.br/artigos/evolucao-do-e-commerce>. Acesso em: 12 jun. 2025.

Fundação de Proteção e Defesa do Consumidor de São Paulo [PROCON-SP]. n.d. Proteção de Dados e LGPD. Disponível em: <https://www.procon.sp.gov.br/protecao-da-dados-e-lgpd/>. Acesso em: 19 mar. 2025.

Gartner. 2025. Gartner Forecasts Worldwide End-User Spending on Information Security to Total $213 Billion Dollars in 2025. Disponível em: <https://www.gartner.com/en/newsroom/press-releases/2025-07-29-gartner-forecasts-worldwide-end-user-spending-on-information-security-to-total-213-billion-us-dollars-in-2025>. Acesso em: 28 set. 2025.

International Business Machines Corporation [IBM]. 2024. Segurança da Informação. Disponível em: <https://www.ibm.com/br-pt/think/topics/information-security>. Acesso em: 28 set. 2025.

Executive summary from the Final Course Work of the Specialization in Project Management of the MBA USP/Esalq

To learn more about the course, click here and access the MBX Academy platform

Who edited this article

Most recent

You may also like

October 16, 2026

Tax reform and technology companies: adaptation of controlling to the new model of tax credits

The Brazilian Tax Reform, instituted by Constitutional Amendment No. 132/2023 and regulated by Complementary Law No. 214/2025, altered consumption taxation, replacing taxes with the Tax on Goods and Services (IBS) and the Social Contribution on Goods and Services (CBS), under the principle of full non-cumulativeness. The study aimed to describe the main challenges perceived by controlling professionals in technology companies in adapting to the new model of non-cumulativeness and tax credit calculation. Applied research, descriptive, with a mixed approach (qualitative-quantitative), was conducted through a field survey. Data were collected from 15 controlling or fiscal management professionals, using a structured questionnaire. The responses were analyzed by frequency distribution and thematic content analysis. The results revealed that professionals perceive the Reform as an event with multifunctional impact, demanding revisions in Enterprise Resource Planning (ERP) systems, credit calculation criteria, team training, and interdepartmental integration. The limitation of credit utilization on labor costs emerged as a specific structural challenge for the sector. The strengthening of the controlling function’s strategic role in the tax transition was observed, especially in pricing, tax planning, and contracting structure, contributing to the debate on the managerial and operational implications of the Tax Reform.

Keywords: controlling; tax credits; technology companies; non-cumulativeness; Tax Reform.

October 16, 2026

High precipitation forecast in Rio Grande do Sul for the prevention of hydrometeorological disasters

In April 2024, Rio Grande do Sul faced the greatest climate tragedy in its history, highlighting the need to improve strategies for forecasting and preventing hydrometeorological disasters. The study aimed to develop a predictive model based on Long Short-Term Memory (LSTM) neural networks to classify high precipitation events in the state. Historical data from the National Institute of Meteorology (INMET) between 2013 and 2025, collected from 44 automatic weather stations, were used. The methodology included variable selection by the Random Forest algorithm, which identified total hourly precipitation and dew point temperature as the features with the highest predictive power, and modeling with a two-layer LSTM architecture, trained with SMOTE application for handling class imbalance. The model achieved an AUC-ROC of 0.9373 and a recall of 82.4% for extreme events, defined as precipitation above the 95th percentile of the historical distribution (8.60 mm/hour). The capacity to identify the vast majority of high precipitation occurrences in advance was demonstrated, confirming the viability of using LSTM neural networks for forecasting extreme hydrometeorological events. The work offers potential contributions to climate monitoring systems and public policies for natural disaster prevention in Brazil.

Keywords: Hydrometeorological disasters; LSTM; Extreme precipitation; Artificial Neural Networks; Rio Grande do Sul.

October 16, 2026

Spatial Analysis of the Age of Condominiums in São Paulo as a Strategic Tool for Administrators

The housing densification in São Paulo, where 29.4% of the population resides in apartments, has established a highly competitive condominium management market that demands spatial data-based prospecting strategies. The study aimed to map the urban evolution of the São Paulo capital through spatial analysis algorithms and building age indicators for the development of competitive intelligence tools. The methodology comprised the extraction and integration of public CNPJ data via the KNIME platform, resulting in a database of 32,040 condominiums, 98% of which were geocoded with the Google Maps API. Macro-spatial analysis identified that the East and South Zones led recent real estate growth, while peaks in registrations in 1984 and 1985 were associated with changes in tax legislation and not with real construction cycles. In parallel, micro-spatial analysis used the BallTree algorithm to compare each condominium with its 20 closest neighbors, revealing that 29% of the units represented points of local renewal in consolidated areas. The results were consolidated in an interactive dashboard that allows strategic visualization of the market at different levels of granularity. It was concluded that the integration of data science to the territory allowed administrators to replace generic approaches with personalized prospecting tailored to the lifecycle and competitive reality of each condominium, optimizing the allocation of marketing resources and the effectiveness of commercial teams.

Keywords: BallTree Algorithm; Spatial Analysis; Urban Data Science; Condominium Management.

October 16, 2026

Criteria and usability of documents in the process of choosing didactic materials

A study analyzed pedagogical documents that guide the choice of teaching materials in the National Program for Books and Teaching Materials (PNLD), considering their relevance for teaching practice and school management. The objective was to analyze the clarity, organization, and applicability of the PNLD 2023 Digital Guide – Early Years and the Support Guide for the Analysis of Teaching Materials from the Reúna Institute, identifying their limitations, potential, and implications for pedagogical practice. A qualitative, exploratory, and descriptive research was conducted, based on documentary analysis guided by the content analysis technique. Aspects such as structure, navigability, organization of criteria, textual density, and reading time estimation were examined. The analysis revealed tensions between state regulation, the publishing market, and school autonomy, and the results showed significant differences between the documents. The PNLD Guide proved to be extensive, technically dense, and predominantly normative. In contrast, the Reúna Institute’s Guide was organized progressively and didactically, with graphic resources that favored understanding and collective use by schools. It was concluded that the structural and discursive characteristics of the documents directly influence their usability and can impact the degree of autonomy and participation of school teams in the process of choosing teaching materials.

Keywords: teacher training; teaching materials; PNLD; curriculum policies; decision-making process.

October 16, 2026

Automation of Tax and Accounting Auditing: Quality Verification Tool for the SICONFI Ranking

The transparency and integrity of accounting information are fundamental requirements for the fiscal compliance of public entities, evaluated by the National Treasury Secretariat (STN) through the Ranking of the Brazilian Public Sector Accounting and Fiscal Information System (SICONFI). The study aimed to develop a proactive audit tool for a federation entity, seeking to mitigate fiscal inconsistencies in Dimension 1 (Information Management) before the official result. The methodology was based on Data Science, using the Python language for automated extraction of the Accounting Balances Matrix (MSC) via the Transparent Treasury Application Programming Interface (API), and the Power BI software for building a compliance dashboard. The project focused on data integration to ensure the symmetry of deadlines and deliveries between accounting records and fiscal statements. The results demonstrated that the automation of the data flow allowed for the early identification of discrepancies, promoting information reliability and the efficiency of governmental internal control, with a structured architecture for expansion to other dimensions of the ranking. The application of the tool to data from the 2025 fiscal year demonstrated its full functionality, with 100% compliance in the 38 control checks of Dimension 1. It was concluded that the solution acted as an innovative and scalable model for the modernization of public management, delivering a tool capable of mitigating fiscal inconsistencies prior to the consolidation of the SICONFI Ranking.

Keywords: Data science; Governmental compliance; Internal control; Public management; Business intelligence.

October 16, 2026

Structured high-risk management model in supplementary health: operational efficiency and care coordination

The growing pressure for economic-assistance sustainability in Brazilian supplementary healthcare has driven the adoption of management models for higher-risk populations. The study aimed to describe and analyze the structuring and operationalization of a high-assistance-risk portfolio management model, as well as to descriptively compare the aggregated indicators of hospitalizations and emergency room visits of the same cohort of 100 beneficiaries between 2023 and 2025. A single, descriptive, retrospective case study was conducted, integrating qualitative document analysis and descriptive quantitative analysis of administrative data. The model was implemented in a vertically integrated health plan, with 100 high-risk beneficiaries identified by an electronic marker. The process involved daily capture of assistance events, categorization by specialty, distribution of cases to the responsible team, and a multidisciplinary follow-up protocol for 28 days after emergency room visits or hospital discharges. Document analysis revealed greater standardization of workflows, definition of responsibilities, and traceability of follow-ups in the reformulated model. In the comparison between periods, hospitalizations decreased from 166 in 2023 to 52 in 2025, a reduction of 68.67%, and emergency room visits fell from 238 to 130, a variation of 45.38%. The reductions occurred during the period when the model was already incorporated into institutional routine, indicating a favorable temporal association. However, due to the descriptive design, the absence of a comparison group, and the impossibility of adjusting for clinical and contextual factors, the observed variations could not be exclusively attributed to the model’s implementation. The study contributed by systematizing an organizational experience of managing a high-risk portfolio in supplementary healthcare, offering elements for analysis and adaptation in similar institutional contexts.

Keywords: Care coordination; High-risk management; Population management; Post-event monitoring; Supplementary healthcare.

October 16, 2026

Digital culture, automation and CRM usage: a case study in the real estate market.

Digital transformation has expanded the availability of Customer Relationship Management (CRM) systems, automation, and artificial intelligence in the sales process, but their strategic appropriation is not automatic. The study analyzed how organizational digital culture, sales automation, and CRM usage influence sales management and performance in a real estate market organization, investigating CRM usage practices, operational frictions, perceptions of brokers and leadership, and data-driven governance mechanisms. A mixed, applied, and descriptive approach was adopted, with structured questionnaires applied to 48 brokers and 4 leadership professionals, in addition to descriptive statistical analysis, construction of synthetic indices, and thematic analysis. The findings indicated that CRM is incorporated into the access routine, but not into the consistent production of data. Convergence was observed between brokers and leadership regarding registration weaknesses and fragmentation across parallel channels, although with distinct perceptions of the effects. Length of service was not linearly related to digital maturity, and the increase in the volume of opportunities was associated with greater difficulty in manual tracking. As a contribution, the GCOD (Governança Comercial Orientada a Dados – Data-Oriented Sales Governance) framework was proposed, which integrates CRM, analytical intelligence, sales orchestration, and continuous governance. It was concluded that sales efficiency depends on the organizational capacity to integrate technology, behavior, and governance to reduce frictions, preserve data quality, and orchestrate human and automated attention throughout the lead journey.

Keywords: Sales automation; CRM; Data culture; Sales governance; Digital transformation.

October 16, 2026

Applications of Artificial Intelligence and Computational Simulation in Pricing and Liquidity Management

The study, of an exploratory nature, investigated how Artificial Intelligence and Computational Simulation techniques can be applied to pricing with a focus on corporate liquidity management. The objective was to propose and illustrate a method of Adaptive Dynamic Cash-Oriented Pricing, in which pricing depended on its expected effects on sales and, subsequently, on the cash flow trajectory. To this end, a hypothetical case was developed with synthetic data on prices and sales, a predictive sales model as a function of own price and average competitor price, a cash flow projection model capable of incorporating payment terms, defaults, transactional costs, inventory replenishment, and fixed costs, and a computational simulator for evaluating competitive scenarios. The predictive modeling process compared different machine learning algorithms, with a linear regression with quadratic and interaction terms being selected for presenting competitive predictive performance, good generalization capability, and greater interpretability. The simulation results indicated that the proposed approach outperformed a baseline fixed-price scenario from a liquidity perspective, producing superior cash trajectories in most of the analyzed horizon. It was concluded that pricing can be treated not only as a commercial problem but also as a cash management problem, opening space for practical applications and future studies with real data, multiple products, and more advanced sequential decision-making methods.

Keywords: Machine Learning; Cash flow; Pricing; Computational Simulation; Decision Making.