“AI-Personalised Marketing and ESG Claims in Malaysian Online Fashion: How Do They Shape Gen Z’s Trust and Purchase Decisions?”
The topic has three important business issues: artificial intelligence (AI), sustainability and consumer behaviour. It also addresses a practical management question:
How can online fashion businesses use AI to personalise marketing and communicate sustainability claims without losing the trust of young consumers?
The study is particularly relevant to Malaysia, where online shopping, social commerce, influencer marketing and sustainability messaging intersect. We need to establish whether AI marketing or sustainability claims influence consumers, but to understand why they work, when they work, and how businesses should manage them.
1. Break down the research title
A. AI-personalised marketing
This refers to using AI and customer data to tailor marketing messages, product recommendations, advertisements, promotions and shopping experiences to individual consumers.
Example: A Malaysian Gen Z shopper browses modest fashion on an e-commerce platform. The platform learns from their browsing and purchase patterns and recommends similar outfits, personalised discounts or products that match their style.
B. ESG claims
ESG stands for Environmental, Social and Governance. In fashion marketing, these claims communicate a brand's environmental impact, social responsibility or ethical business practices.
Examples: “Made from recycled fibres,” “ethically sourced materials,” “fair wages for garment workers,” or “transparent supply chain.”
The important issue is whether consumers consider these claims credible, meaningful and supported by evidence.
C. Malaysian Gen Z consumers
Gen Z generally refers to people born approximately between 1997 and 2012, although definitions vary. In October 2026, this spans approximately ages 14–29.
It is sensible to focus on adult Gen Z consumers aged 18–29 who shop for fashion online in Malaysia. This makes recruitment and consent procedures more manageable while keeping the study focused.
D. Trust and purchase decisions
Trust concerns whether consumers believe a fashion brand is honest, competent and reliable. Purchase decisions concern their intention to buy, willingness to choose a brand, or actual purchasing behaviour.
Example: A shopper may like a personalised recommendation but distrust a brand's sustainability claim. Alternatively, credible environmental information may strengthen trust and increase purchase intention.
The central idea is that AI personalisation and ESG communication may influence consumers differently. Personalisation may make marketing more relevant, but it can also feel intrusive. ESG claims may make a brand more attractive, but unsupported or exaggerated claims can undermine credibility.
2. Why this topic matters to businesses
Consider a hypothetical Malaysian online fashion brand selling everyday clothing, sportswear and modest fashion.
A practical business scenario
MARKETING ACTION 1
AI personalisation
The brand recommends products based on browsing history, purchase patterns and stated preferences.
MARKETING ACTION 2
ESG communication
Product pages describe recycled materials, ethical manufacturing and reduced packaging, with varying levels of supporting evidence.
CONSUMER RESPONSE
Trust, perceived value and purchase intention
Consumers assess whether the recommendations are useful, whether the brand respects their privacy and whether its sustainability claims are believable.
The business dilemma is straightforward:
More personalisation may improve product discovery and conversion.
More data collection may increase privacy concerns.
Strong sustainability claims may differentiate a brand.
Weak evidence may expose the brand to accusations of greenwashing.
Consumers may express support for sustainability but still prioritise price, quality, convenience or style when purchasing.
The topic can produce actionable recommendations for marketing managers, e-commerce leaders, brand managers and sustainability professionals.
3. The central research problem
In this case, the research problem is that online fashion brands are increasingly able to personalise consumer experiences and communicate sustainability, yet it may be difficult to know whether these strategies build genuine trust or merely attract short-term attention.
There are four related problems.
Problem 1: AI personalisation has both benefits and risks
AI can improve relevance, convenience and product discovery. However, consumers may become uncomfortable when recommendations seem to reveal information they did not knowingly share.
For example:
A shopper receives recommendations that match their preferred style. This may improve the shopping experience.
A shopper sees an advertisement for a product shortly after discussing it privately or browsing similar products. They may feel monitored.
A shopper is repeatedly shown products based on past purchases. They may appreciate the convenience or feel that their choices are being restricted.
Recent research on AI-powered fashion retail examines this balance between the benefits of personalisation, trust, perceived risk and privacy concerns.
How can retailers maximise the commercial benefits of personalisation without creating distrust?
Problem 2: ESG claims may not always be credible
Consumers often cannot independently verify claims about the materials, factories or supply chains behind a garment.
Consider these two product descriptions:
Claim A: Vague sustainability marketing
“Our clothing is environmentally friendly and sustainable.”
Problem: The claim does not explain what makes the garment sustainable or provide supporting evidence.
Claim B: Specific, verifiable information
“This garment contains 70% recycled polyester by weight. See the material specification and supplier documentation.”
Potential advantage: The claim is more specific and gives the consumer a way to assess the evidence.
Even specific claims should be accurate, appropriately substantiated and clear about their scope. A recycled-material claim alone does not establish that the entire garment is sustainable.
When brands exaggerate or misrepresent their environmental performance, the practice is known as greenwashing. A 2026 study of Gen Z consumers in the fast-fashion sector reported a strong negative association between perceived digital greenwashing and brand trust.
How do the credibility and clarity of ESG claims influence Gen Z's trust and willingness to purchase?
Positive attitudes do not always translate into purchases
A consumer may genuinely support environmental protection but still buy a cheaper, less sustainable product.
This is commonly called the attitude–behaviour gap, or, when focusing on stated intentions versus actions, the intention–behaviour gap.
For example, a shopper may:
Prefer recycled materials but reject a garment that costs substantially more.
Support ethical labour practices but prioritise delivery speed.
Like sustainable fashion but prefer the style of a conventional alternative.
Intend to buy sustainable clothing but purchase whatever is promoted during a sale.
Research on sustainable fashion continues to examine these barriers and the difference between favourable attitudes and actual buying behaviour.
Does consumer trust translate into purchase intention, and does purchase intention translate into actual purchasing behaviour?
Problem 4: AI and ESG communication may interact
This is potentially the most distinctive part of your proposed research.
Imagine two fashion brands offering the same garment at the same price.
Brand A uses personalised recommendations and makes a broad claim that its clothing is sustainable.
Brand B uses personalised recommendations and provides specific, transparent evidence supporting its environmental claims.
Would consumers trust Brand B more? Would the answer change if the recommendation were highly personalised but the shopper was concerned about privacy?
These questions go beyond examining AI personalisation and sustainability marketing separately. They investigate how consumers evaluate the combined experience.
The research opportunity is to test whether AI personalisation and ESG claim credibility reinforce one another, operate independently, or create tensions that affect consumer trust and purchase decisions.
This combined relationship is a potential research gap to investigate through a systematic literature review. It should not be claimed as an established gap until you have examined the latest studies, including research on Malaysian consumers.
4. What theories could support the topic?
We need a theoretical foundation that explains why the relationships between variables should exist. We need to avoid combining too many theories simply to make the framework look sophisticated.
We are considerang three main theories or frameworks.
Theory 1: Stimulus–Organism–Response (S-O-R)
A useful overarching framework
This framework explains how external stimuli influence a person's internal psychological state, which then influences a response.
Stimulus: AI personalisation and ESG claim credibility.
Organism: Consumer trust, perceived authenticity and perceived privacy risk.
Response: Purchase intention or purchasing behaviour.
It provides a clear structure for connecting marketing practices with consumer decision-making.
Theory 2: Trust theory
Trust theory helps explain why consumers become willing—or unwilling—to rely on a brand.
Consumers may assess a brand's honesty, competence and integrity. In AI-enabled shopping, you should also distinguish trust in the technology from trust in the fashion brand.
For example, a shopper may trust an e-commerce platform's recommendation algorithm but distrust the retailer's environmental claims.
Theory 3: Signalling theory
Signalling theory explains how a business communicates information about product quality or otherwise difficult-to-observe characteristics.
In online fashion, shoppers cannot easily inspect the factory, verify the supply chain or test every environmental claim before buying. Product disclosures, credible certifications and traceability information can serve as signals.
A signal is more useful when it is credible and difficult to fake. A vague “eco-friendly” label may provide less assurance than a specific claim backed by verifiable evidence.
How should you select the theories?
My recommendation is to use S-O-R as the main organising framework, with trust theory explaining the central mechanism and signalling theory informing how consumers assess ESG claims.
However, you could instead use signalling theory as the primary lens if your main research emphasis is ESG credibility and greenwashing. Your final selection should follow the literature review and the specific relationships you intend to test.
A useful starting point is the 2025 study on AI-enabled services in online fashion stores, which examines trust, utility, product uniqueness and privacy over time.
5. Build a conceptual framework
I would initially recommend a model with a manageable number of variables. It should be sophisticated enough for doctoral research but not so complicated that data collection and interpretation become unmanageable.
Independent variables
Perceived AI personalisation quality
Perceived ESG claim credibility
Mediating variable
Consumer trust in the fashion brand
Explains how marketing perceptions may influence buying decisions
Dependent variable
Purchase intention
Potential moderator
Consumer privacy concern
May weaken or change the relationship between AI personalisation and trust.
This is a proposed framework, not an established finding. You would test the relationships using empirical data.
What does each variable mean?
Variable | Role | What you would measure |
|---|---|---|
AI personalisation quality | Independent | Relevance, usefulness and perceived fit of recommendations |
ESG claim credibility | Independent | Believability, clarity and perceived evidence supporting sustainability claims |
Brand trust | Mediator | Belief that the brand is honest, dependable and acts with integrity |
Purchase intention | Dependent | Likelihood or willingness to buy the advertised fashion product |
Privacy concern | Moderator | Concern about how personal data are collected and used |
An important distinction: If your title promises to study “purchase decisions,” a survey measuring only purchase intention may not fully deliver on that promise. You could measure recent self-reported purchases as well, or revise the title to say “purchase intention.”
Also, this initial framework treats ESG claims as perceived marketing information. If you want to investigate the entire ESG concept, you would need to address environmental, social and governance claims separately or justify a broader measurement construct.
6. Research objectives and research questions
The objectives should align directly with your research framework and the business problem.
Proposed research objectives
To examine the relationship between perceived AI personalisation quality and brand trust among Gen Z consumers shopping for fashion online in Malaysia.
To examine the relationship between perceived ESG claim credibility and brand trust.
To investigate the relationship between brand trust and purchase intention.
To determine whether brand trust mediates the relationships between AI personalisation quality, ESG claim credibility and purchase intention.
To examine whether privacy concerns moderate the relationship between AI personalisation quality and brand trust.
To develop managerial recommendations for online fashion retailers to strengthen consumer trust and purchase intention through responsible AI personalisation and credible ESG communication.
Proposed research questions
Research question | What it seeks to discover |
|---|---|
RQ1 | Does perceived AI personalisation quality influence brand trust? |
RQ2 | Does ESG claim credibility influence brand trust? |
RQ3 | Does brand trust influence purchase intention? |
RQ4 | Does brand trust explain how AI personalisation and ESG claim credibility relate to purchase intention? |
RQ5 | Do privacy concerns weaken the trust-building effect of AI personalisation? |
RQ6 | What practical marketing strategies can Malaysian online fashion retailers adopt based on the findings? |
RQ6 is important because the study should lead to recommendations that managers can implement, not simply statistical conclusions.
7. Develop hypotheses that can be tested
If a quantitative research design, we can translate our conceptual framework into hypotheses.
H1 — AI personalisation and trust
Perceived AI personalisation quality is positively associated with brand trust.
Rationale: Relevant, useful recommendations may signal that a retailer understands consumer needs.
H2 — ESG credibility and trust
Perceived ESG claim credibility is positively associated with brand trust.
Rationale: Clear and credible sustainability information may increase confidence in the brand.
H3 — Trust and purchase intention
Brand trust is positively associated with purchase intention.
Rationale: Consumers may be more willing to purchase from brands they consider dependable.
H4 — Mediation by brand trust
Brand trust mediates the relationships of AI personalisation quality and ESG claim credibility with purchase intention.
Rationale: Marketing practices may influence purchase intention partly by shaping trust.
H5 — Moderation by privacy concern
Privacy concern weakens the positive relationship between AI personalisation quality and brand trust.
Rationale: Highly privacy-conscious consumers may respond less favourably to personalisation they perceive as intrusive.
These are proposed hypotheses, not findings. Existing research suggests that privacy and transparency matter, but their effects can depend on the context and the precise variables measured.
We should also consider whether ESG credibility and AI personalisation interact. For example, does personalisation make credible ESG information more persuasive, or does it make the message seem manipulative? That is a potentially valuable extension, but include it only if the literature and research design justify it.
8. How should you conduct the research?
My initial recommendation is a quantitative explanatory study, potentially supplemented by an experiment if wewant stronger evidence about cause and effect.
Option A: Quantitative survey
Step 1 — Define your population
Malaysian residents aged 18–29 who have purchased fashion online within a specified recent period and have encountered personalised recommendations or marketing. Establish eligibility criteria carefully, especially if participants cannot distinguish algorithmic personalisation from ordinary targeted advertising.
Step 2 — Develop a questionnaire
Use established measurement scales where appropriate, adapting them to the Malaysian online fashion context. Include questions on personalisation quality, ESG claim credibility, brand trust, privacy concern and purchase intention.
Step 3 — Conduct a pilot study
Test whether respondents understand terms such as AI personalisation, ESG and sustainability claims. Check questionnaire length, reliability, clarity and the quality of responses.
Step 4 — Collect and analyse data
Use an appropriate sampling strategy, assess measurement quality and test the hypothesised relationships. Consider structural equation modelling (SEM), including PLS-SEM if justified by your model, objectives and data characteristics.
How many respondents would you need?
Do not automatically choose 384 respondents simply because it is a commonly quoted survey sample size.
Your sample size should be determined by the statistical analysis, expected effect sizes, desired power, number of predictors, model complexity and the feasibility of recruiting eligible respondents.
For an initial planning discussion, you might explore a target of approximately 400–500 usable responses. However, this is only an illustrative planning range, not a statistically validated requirement. A formal power analysis should determine your final target, including allowances for incomplete or invalid responses.
Option B: Experimental research
An experiment may be particularly valuable because your title asks how marketing approaches shape trust and purchase decisions.
You could develop a simulated online fashion product page and manipulate two factors:
AI personalisation: high versus low.
ESG information: credible and specific versus vague or unsupported.
This produces a 2 × 2 factorial design.
Illustrative experimental design
Group 1
Low personalisation
Vague ESG claims
Group 2
High personalisation
Vague ESG claims
Group 3
Low personalisation
Credible ESG claims
Group 4
High personalisation
Credible ESG claims
Participants are randomly assigned to one condition. The simulated product, price, visual quality and other information should be held constant wherever possible.
After exposure, participants could rate:
Brand trust.
Perceived authenticity of ESG claims.
Perceived intrusiveness of personalisation.
Purchase intention.
Willingness to choose the product over an alternative.
The analysis would examine the main effect of each factor and whether the combination produces an interaction effect.
My advice: A survey is generally simpler for a DBA project with limited time and resources. An experiment is stronger for testing the effects of deliberately manipulated marketing cues. A carefully designed combination of an experiment and a follow-up survey could provide richer evidence, but only if the additional time and recruitment demands are manageable.
Neither design automatically proves actual purchasing behaviour. If actual decisions are central to your thesis, consider an incentivised choice task, a behavioural outcome or follow-up purchase data where feasible.
9. What makes this a doctoral-level study rather than an ordinary marketing project?
A DBA is expected to contribute to professional practice while demonstrating rigorous academic reasoning.
Your contribution should ideally have three dimensions.
Theoretical contribution
Test whether brand trust explains how AI personalisation quality and ESG claim credibility relate to purchase intention in the Malaysian online fashion context. Examine whether privacy concern changes these relationships.
The contribution is not simply applying a familiar model in a new country. It is demonstrating what the combined framework explains, what it does not explain, and whether its proposed relationships are supported.
Managerial contribution
Develop evidence-based recommendations for online fashion businesses about the design of personalised recommendations, the presentation of sustainability evidence, consumer data controls and brand-trust management.
Contextual contribution
Investigate whether these relationships hold for adult Gen Z consumers in Malaysia, a specific market and consumer group. Do not assume that results from China, Europe or other countries necessarily generalise to Malaysia.
For the managerial output, you could develop a practical framework such as a Trust-First AI and ESG Marketing Framework. It might guide retailers to assess personalisation relevance, transparency, evidence quality, privacy safeguards and consumer responses before launching campaigns.
This framework would be your proposed practical output; you would need to derive and validate it through your research rather than claim its effectiveness in advance.
10. Is there a genuine research gap in this topic?
This is a crucial question before registering your DBA proposal.
A preliminary literature search reveals that several components of your topic have already been studied:
AI-enabled personalisation and online fashion purchasing, including trust and privacy.
AI personalisation in sustainable advertising, including privacy concerns and transparency.
Digital greenwashing, Gen Z and brand trust in fast fashion.
Malaysian Gen Z and environmentally friendly purchase intention.
This means you should not claim that AI personalisation, consumer trust, or ESG marketing has never been researched before.
The more defensible research opportunity is to investigate the specific combination of:
AI personalisation in online fashion.
Credibility of ESG claims.
Brand trust as a possible explanatory mechanism.
Privacy concerns as a boundary condition.
Adult Gen Z consumers in Malaysia.
You should verify whether this exact combination has already been examined in published studies, theses and conference papers. The search should include papers published in 2025 and 2026, not just older foundational research.
A possible gap statement
Although research has examined AI-enabled personalisation in online fashion retail and the influence of sustainability communication on consumer trust, these streams provide an incomplete basis for understanding how consumers evaluate both practices together. In particular, further investigation is warranted into how perceived AI personalisation quality and ESG claim credibility jointly relate to brand trust and purchase intention among adult Gen Z consumers in Malaysia, and whether privacy concerns alter these relationships. Addressing this issue could extend understanding of responsible digital marketing while providing online fashion retailers with practical guidance on balancing personalisation effectiveness, credible sustainability communication and consumer trust.
Notice the cautious wording: further investigation is warranted. That is more academically defensible than declaring a complete absence of research before conducting the review.
11. Suggested literature review and reading list
Start with the following recent studies. They are useful for mapping the field, but you should read the full articles and verify their methods, measures, limitations and theoretical contributions before using them in your thesis.
1. Understanding customers’ intentions to use AI-enabled services in online fashion stores – a longitudinal study
International Marketing Review, 2025
Useful for understanding trust, utility and privacy across the fashion shopping journey.
2. When and how AI personalization drives sustainable purchases
Journal of Retailing and Consumer Services, 2026
Particularly relevant to AI personalisation, message relevance, privacy and transparency in eco-friendly advertising.
3. Decoding Digital Greenwashing: A Quantitative Analysis of Signaling Theory, Brand Trust, and Purchase Intention among Generation Z Consumers in the Fast-Fashion Sector
Social Reports, 2026
Useful for developing your ESG credibility, greenwashing and trust constructs. Evaluate its methods and journal quality critically, as you should with all sources.
4. Sustainable Choices in Fashion: Insights into Gen Z’s Purchasing Behaviour
Journal of Logistics, Informatics and Service Science, 2025
A relevant starting point for Malaysian academic research on sustainable apparel and Gen Z purchasing behaviour.
5. AI-powered personalization in online fashion stores: Exploring the impact on shoppers’ perceived risk and intention to buy and recommend
Information & Management, 2026
Useful for examining perceived risk, trust deficits, shopping experience and purchase intention.
For foundational theory, add these classic works to your reading plan:
Ajzen (1991), The Theory of Planned Behavior — useful if you investigate the relationship between intention and behaviour.
Mehrabian and Russell (1974), An Approach to Environmental Psychology — foundational work behind the stimulus–organism–response perspective.
Spence (1973), Job Market Signaling — the foundational signalling theory paper.
Morgan and Hunt (1994), The Commitment-Trust Theory of Relationship Marketing — useful for understanding trust in marketing relationships.
For the literature review, build a matrix recording each paper's country, population, theory, variables, methodology, findings, limitations and the research opportunity it leaves open.
12. Important decisions before you finalise the title
Your current title is relevant, but it contains a few terms that need to be clarified.
Issue | Recommendation |
|---|---|
“AI-personalised marketing” | Specify whether this means AI-driven product recommendations, personalised advertisements, AI styling tools or all of these. |
“ESG claims” | Decide whether you will measure environmental claims alone or environmental, social and governance claims. |
“Gen Z” | Define your age range, preferably adult consumers aged 18–29 for a manageable study. |
“Purchase decisions” | Decide whether the outcome is purchase intention, self-reported past purchases or observed behaviour. |
“How do they shape” | If you use a cross-sectional survey, interpret results as relationships or associations rather than definitive causal effects. |
Recommended working title
“AI-Personalised Marketing, ESG Claim Credibility and Purchase Intention among Gen Z Online Fashion Consumers in Malaysia: The Mediating Role of Brand Trust and the Moderating Role of Privacy Concern.”
Copy titleWhy I prefer this version:
It identifies the two marketing-related predictors.
It specifies brand trust as the proposed mediator.
It identifies privacy concern as the proposed moderator.
It makes purchase intention the clearly defined outcome.
It specifies the Malaysian online fashion context.
If you intend to measure actual purchasing behaviour, retain “purchase decisions” or “purchasing behaviour” and design your data collection accordingly.
13. Practical roadmap
Phase 1 — Define the business problem. Identify the precise managerial challenge faced by online fashion retailers and the decisions your findings could improve.
Phase 2 — Conduct a systematic literature review. Search academic databases such as Scopus, Web of Science and Google Scholar. Map existing research on AI personalisation, ESG credibility, trust and Gen Z purchasing.
Phase 3 — Establish the research gap. Demonstrate what is already known, what remains unclear and why Malaysia provides a meaningful context for further investigation.
Phase 4 — Develop the theoretical model. Finalise your constructs, conceptual framework, research questions and hypotheses. Avoid including variables without a clear theoretical purpose.
Phase 5 — Design and validate the methodology. Choose the survey or experimental approach, conduct a power analysis, pilot the instruments and obtain the required ethics approval.
Phase 6 — Analyse the evidence. Evaluate measurement quality, test the hypotheses and assess the practical significance of the findings, not just statistical significance.
Phase 7 — Translate findings into a management framework. Recommend specific actions concerning personalisation transparency, ESG substantiation, privacy controls and trust-building practices.
Final perspective
I believe this topic has good potential for a DBA, provided you narrow its scope and establish a defensible research gap. The most valuable outcome would not simply be proof that AI marketing or sustainability communication influences young consumers. It would be an evidence-based explanation of how online fashion retailers can use personalisation and credible ESG communication together while protecting the consumer trust on which long-term business performance depends.
The next step is to decide the scope of your study, because that choice will determine the literature review, framework, questionnaire and methodology.
No comments:
Post a Comment