Medium Link: Balancing Personalisation and Ethics in AI-Enabled Marketing Communications | by Manasa R | Jul, 2026 | Medium
Course Relevance: MBA, PGDM, Executive MBA, and undergraduate courses in Marketing, Digital Marketing, Business Ethics, Artificial Intelligence, and Data Analytics. This Caselet is relevant for courses in:
- Business Communication and Professional Presentation
- Decision-Making and Strategic Management
- Business Analytics and Data-Driven Decision-Making
- IT Project Management and Product Strategy
- Leadership and Organizational Behaviour
Academic Concepts
- Data-Driven Decision-Making (DDD)
- Strategic Storytelling and Narrative Framing
- Object oriented Programming Language-Java
- Cognitive and Emotional Engagement in Leadership
- Analytics Interpretation vs Analytics Communication
- Stakeholder Management and Executive Influence
- User-Centric Product Management
Learning Objectives
After analysing this case, students should be able to:
- Examine how algorithmic bias can influence marketing communication outcomes.
- Evaluate ethical issues associated with consumer data collection and AI-driven personalisation.
- Understand the relationship between data privacy, informed consent, and customer trust.
- Recommend responsible AI governance practices for marketing organisations.
Discussion Questions
- What forms of algorithmic bias are evident in NovaCart’s AI marketing system?
- How can excessive personalisation affect consumer perceptions of privacy?
- Is legal compliance sufficient to ensure ethical AI-enabled marketing? Explain.
- What measures can NovaCart adopt to improve transparency and accountability?
- How should organisations balance commercial objectives with ethical responsibilities when using AI for marketing communications?
Teaching Note (Brief)
This case highlights three major dimensions of responsible AI in marketing:
- Algorithmic Bias: AI models trained on historical data may unintentionally discriminate against certain demographic groups.
Data Privacy: Ethical marketing requires informed consent, transparency, and responsible handling of personal information beyond minimum legal requirements.
Case Title: The AI Campaign Dilemma at NovaCart
NovaCart, a rapidly expanding online retail company, adopted an artificial intelligence (AI)-powered marketing platform to strengthen customer engagement and improve conversion rates. The platform analysed browsing patterns, purchase history, demographic information, and social media interactions to deliver highly personalized advertisements through email, mobile applications, and digital channels.
During the first six months, the company experienced a 28% increase in click-through rates and a 19% improvement in online sales. Senior management considered the initiative a major success and decided to invest further in AI-driven customer relationship management.
However, concerns emerged when the marketing analytics team noticed that premium product advertisements were shown primarily to customers from higher-income neighbourhood, while users from economically weaker regions mostly received promotions for discounted products. Similarly, certain job-related educational advertisements rarely reached female users, despite their demonstrated interest in professional development courses. These patterns raised questions about whether the AI model was unintentionally reinforcing existing social and economic inequalities.
At the same time, several customers expressed discomfort after receiving advertisements that appeared to predict personal interests they had never explicitly shared with NovaCart. Investigations revealed that the AI platform collected behavioural information from third-party websites through tracking technologies. Although this practice was legally disclosed within lengthy privacy policies, many users were unaware of the extent of data collection.
The company’s ethics committee argued that legal compliance alone was insufficient to maintain customer trust. Consumer advocacy groups criticised NovaCart for prioritising business performance over transparency and fairness. Social media discussions intensified, with many users demanding greater control over how their personal information was collected and used.
In response, NovaCart’s leadership initiated an internal review. The company considered introducing explainable AI tools, conducting periodic algorithmic bias audits, simplifying privacy notices, strengthening customer consent mechanisms, and establishing ethical guidelines for AI-based marketing decisions. However, executives were concerned that stricter privacy controls and reduced data collection could weaken the effectiveness of personalised campaigns and affect revenue growth.
NovaCart now faces a strategic decision: should it continue maximising AI-driven personalisation for competitive advantage, or redesign its marketing communications to prioritise fairness, transparency, privacy, and long-term consumer trust?
The Strategic Dilemma
While the proposed reforms aligned with ethical best practices, the Chief Marketing Officer expressed concern that restricting data collection could reduce personalisation accuracy and lower campaign effectiveness.
The Chief Financial Officer estimated that implementing fairness audits, explainable AI systems, and stronger privacy safeguards would increase annual operating costs by approximately ₹18 crore.
Board members were divided.
One group argued that maintaining aggressive AI-driven marketing would preserve NovaCart’s competitive advantage.
Another believed that prioritising responsible AI practices would strengthen long-term customer trust, protect the company’s reputation, and reduce future regulatory risks.
NovaCart must now determine whether success should be measured primarily by immediate financial performance or by creating an ethical, transparent, and trustworthy AI-enabled marketing ecosystem.
The Emerging Problem
Despite the positive financial performance, several unexpected issues began to surface.
Issue 1: Algorithmic Bias
A data scientist conducting routine performance testing observed that the recommendation engine consistently treated customer groups differently.
Examples included:
- Luxury fashion advertisements were shown predominantly to customers living in metropolitan areas.
- Rural customers rarely received premium product promotions regardless of their purchase history.
- Female users interested in technical books frequently received advertisements related to home décor rather than professional certification courses.
- Older consumers were seldom shown advertisements for newly launched electronic gadgets.
- Students from smaller cities were rarely offered financing options for expensive laptops.
Although no employee intentionally programmed these outcomes, historical purchasing patterns used to train the AI had influenced the recommendations.
The AI had learned existing market behaviour rather than recognising potential customer interests.
This resulted in unequal marketing opportunities across different demographic groups.
Issue 2: Privacy Concerns
Customer complaints also increased.
Many users were surprised by the highly personalised advertisements they received.
Examples included:
- A customer searched for diabetes-related information on a health website and later received advertisements for sugar-monitoring devices on NovaCart.
- Another customer discussed holiday plans on social media and immediately began receiving travel package promotions.
- A recently married couple received advertisements for baby products despite never purchasing such items from NovaCart.
Customers questioned how the company had obtained such personal information.
An internal investigation revealed that NovaCart’s AI platform purchased behavioural datasets from advertising networks and combined them with its own customer database.
Although these practices were disclosed within the company’s privacy policy, the information appeared in lengthy legal documents that very few customers actually read.
Several consumer organisations argued that legal disclosure alone did not constitute meaningful informed consent.
Issue 3: Lack of Transparency
Customer service representatives struggled to answer a common question:
“Why am I receiving this advertisement?”
Because the recommendation engine operated as a “black-box” machine learning model, even NovaCart’s marketing managers could not fully explain why certain advertisements were shown to particular users.
This lack of explainability reduced customer confidence.
Some users believed that the company was secretly monitoring their private conversations.
Although these assumptions were incorrect, NovaCart found it difficult to reassure customers due to limited transparency in the AI system.
Issue 4: Ethical Boundaries
The ethics committee identified several practices that raised ethical concerns.
These included:
- Targeting financially vulnerable customers with easy-credit offers.
- Encouraging impulse purchases using emotionally personalised messages.
- Sending frequent promotional notifications late at night.
- Predicting sensitive customer attributes based on indirect behavioural signals.
- Creating psychological profiles without explicit customer awareness.
Although these practices increased sales, committee members questioned whether they respected customer autonomy.
The Chief Ethics Officer argued:
“The fact that technology allows us to influence consumer behaviour does not necessarily mean we should.”
Business Performance
Within eight months of implementation, NovaCart reported impressive improvements.
| Performance Indicator | Before AI | After AI |
| Email Open Rate | 19% | 34% |
| Click-Through Rate | 8% | 29% |
| Conversion Rate | 3.8% | 7.5% |
| Average Order Value | ₹2,950 | ₹3,640 |
| Marketing Cost per Acquisition | ₹1,150 | ₹820 |
| Customer Retention Rate | 62% | 74% |
Encouraged by these outcomes, NovaCart planned to automate nearly 80% of its digital marketing activities using AI.
- Ethical Boundaries: Organisations should implement AI governance frameworks, fairness audits, explainable AI, and privacy-by-design principles to balance business performance with consumer trust.
References:
- Marketing 5.0: Technology for Humanity
Kotler, P., Kartajaya, H., & Setiawan, I. (2021). Marketing 5.0: Technology for Humanity. Wiley.
Why use it? Explains AI, data-driven personalization, customer experience, and responsible marketing. - Sandra Matz, Michal Kosinski, Gideon Nave, & David Stillwell (2017).
Matz, S. C., Kosinski, M., Nave, G., & Stillwell, D. J. (2017). Psychological targeting as an effective approach to digital mass persuasion. Proceedings of the National Academy of Sciences, 114(48), 12714–12719.
Why use it? A landmark study showing how AI-driven personalization influences consumer behavior. - McKinsey & Company. (2021). The Value of Getting Personalization Right—or Wrong—Is Multiplying.
Why use it? Provides industry evidence on the benefits of personalization and the importance of maintaining customer trust. - European Union. (2016). Regulation (EU) 2016/679 (General Data Protection Regulation – GDPR).
Why use it? The foundational regulation for privacy, consent, and ethical handling of customer data in AI-driven marketing. - Artificial Intelligence for Marketing and Product Innovation
Pradeep, A. K., Appel, A., & Sthanunathan, S. (2018). Artificial Intelligence for Marketing and Product Innovation. Wiley.
