AB Angelo Baccelloni Versione italiana
Questions & answers

Questions & answers

Frequently asked questions about Angelo Baccelloni's research on AI, consumer behaviour and brand strategy, with concise answers.

Do recommender systems always help people choose better?
Not always. An agent too specialized on past taste can narrow options and reduce discovery: relevance must be balanced with novelty. doi.org/10.1177/10499968251358181
Does the source (human or AI) change consumer choice?
Yes. Trust and reactions differ depending on whether the recommendation comes from a human or an AI, and the effect depends on product type. California Management Review (2024)
What are 'weak ties' in recommendation algorithms?
They are indirect connections between products and people generated by algorithms. They often drive discovery more than obvious links, and are studied with network analysis. doi.org/10.1108/EJM-02-2024-0136
Can a brand crisis be anticipated?
Partly. The framework detects weak signals — emotion and mobilization — in online discourse before public recognition. It's a reference pattern, not a guaranteed forecast. Working paper (PRIN 2022)
Do nutrition labels really work?
Their design affects comprehension, trust and choices, with differences across formats and countries. More informative labels support more conscious choices. doi.org/10.1108/EJM-10-2021-0764
Nutri-Score or NutrInform: which works better?
It depends on consumer understanding and liking. In a 20-country comparison the two formats show different strengths: there's no absolute winner, but effects that vary by context. doi.org/10.3390/nu15132852
Do European labeling policies have an effect?
Policy initiatives shape how consumers read and use front-of-pack labels: studying them helps design more effective rules. doi.org/10.3390/nu14163423
Does blockchain increase consumer trust?
It can, but only when it closes a real credibility gap and is well communicated. Technology alone isn't enough — how it's framed matters. doi.org/10.1002/mar.21855
Does blockchain work in the food supply chain?
Yes, as a signal of traceability and transparency: when perceived as credible, it supports trust and purchase intention, as shown in dairy and agri-food cases. doi.org/10.1007/s43039-022-00061-0
Does the metaverse create value for luxury brands?
Yes, if it extends (rather than replaces) the physical experience. Value comes from phygital design blending digital and real. doi.org/10.1108/BFJ-03-2024-0248
Is the metaverse suited to sustainable fashion?
It can trigger strong emotional responses toward high-end products, but the medium matters: the experience varies a lot across settings. Sinergie-SIMA (2023)
How do you build a brand story in luxury?
With transparency and uniqueness: the right words matter as much as the product. Coherent storytelling strengthens perceived authenticity. Fondazione CUEIM (2025)
What happens when AI writes the advertising?
AI in advertising can affect perceived brand authenticity: used carelessly, it risks weakening the very thing it should strengthen. Fondazione CUEIM (2025)
Are virtual influencers credible?
Their perceived credibility and engagement depend on how they are designed and presented; they can work, but under rules different from human influencers. SIM Conference (2024)
Can ChatGPT describe a brand's personality?
It can surface brand-personality perceptions and consumer reactions (e.g., in automotive), useful as a starting point — to be validated with real data. Sinergie-SIMA (2024)
How do you measure a company's social impact?
With a structured approach: economic value generated, environmental and social effects, and a Social ROI computed by area, as in the McDonald's outlets project in Rome. McDonald's Italia project
What methods do you use in your research?
Experiments (incl. cross-country), network analysis, text mining and topic modeling, structural equation modeling, mediation/moderation, machine learning and bibliometric analysis.
How can companies use these findings?
By monitoring channels they don't control, weighting mobilization signals (not just sentiment), and designing AI recommenders and phygital experiences better.
← Back to the site