Time to doubt the stars. Tourists are swayed more by vibes than value when rating restaurants

Featured faculty: Michael Zhang
Written by Putro Harnowo

Hungry while travelling, but couldn’t decide which restaurant to go to? No worries. Just grab your phone and check the ratings for nearby restaurants. These user-generated reviews appear to be more trustworthy than traditional advertising due to their open submissions, making them handy not only for tourists, but also for locals craving something special.

No wonder platforms like Google Reviews and Yelp are cherished. Local apps like Japan’s Tabelog, South Korea’s Naver and Kakao, Chinese Mainland’s Meituan, and Hong Kong’s OpenRice also thrive on their community-based and more curated ratings. However, these reviews should be taken with a grain of salt. Ratings often vary across different platforms, and a high rating doesn’t always guarantee quality.

“Online ratings can be biased, leading to misleading information for readers,” says Michael Zhang, the Wei Lun Professor of Business AI at the Department of Decisions, Operations and Technology at the Chinese University of Hong Kong (CUHK) Business School.

In his latest study, Why is the grass always greener on the other side? Tourist bias in online restaurant ratings, Professor Zhang finds that one key source is tourist bias in their reviews. “Tourists are about 13.4 per cent more likely to give restaurants a higher rating than locals. If most reviews come from tourists who rate places more positively, locals might feel the restaurant is overrated.”

This tendency is evidently found across all kinds of restaurants, whether chains or independent, cheap or high-end, and among tourists from both large and small cities, across all genders. The study also looks into the very different experiences and expectations about restaurants from local customers and tourists.

Understanding this bias would help businesses better manage customer satisfaction levels and tailor their offerings to meet their customers’ expectations. Meanwhile, customers can leverage the findings to assist their judgment when looking for restaurants.

“Tourists are about 13.4 per cent more likely to give restaurants a higher rating than locals. If most reviews come from tourists who rate places more positively, locals might feel the restaurant is overrated.”

Professor Michael Zhang

Why do tourists tend to give higher ratings?

Along with Xu Da Peng at South China University of Technology, Hong Hong at Tongji University, and Ye Qiang at the University of Science and Technology of China, Professor Zhang collected data from Chinese Mainland’s most popular restaurant review platform, focusing on 10 cities famous for their local cuisines, namely Changchun, Changsha, Guangzhou, Guiyang, Haikou, Suzhou, Tianjin, Xi’an, Xiamen, and Zhengzhou.

The sample consists of 70,950 reviews written by 747 users for nearly 40,000 restaurants with a wide range of prices and popularities. The researchers identify whether a reviewer is a local or traveller by analysing their location information relative to the location of the reviewed restaurants.

Professor Zhang and his team discovered that the positive emotional state while travelling is the reason for tourist bias. Travelling boosts their mood and makes tourists more willing to express their feelings instead of pondering on analytical or detailed reasoning. Reviews from tourists are also shorter but accompanied by more photos, suggesting they care more about sharing their experiences rather than commenting on taste and price. They are also more forgiving of restaurant flaws.

Tourists on the lookout for more than just functional dining are more sensitive to the emotional aspects, such as the restaurant’s vibe and the hospitality they experience, as they mention in their reviews. Meanwhile, locals write more about locations, cooking methods, and prices. The researchers further identify the distinctive words used by each group

Focusing on experience may give the impression that tourists seek more expensive or higher-quality restaurants, but a deeper analysis finds that both locals and tourists who write reviews often avoid fancy establishments. It’s just that reviews from tourists tend to be more positive.

How can restaurants and customers leverage tourist bias?

Given the different interests between locals and tourists, Professor Zhang suggests restaurants apply a dual-focus promotional strategy that addresses the distinct needs of both groups. Restaurants can emphasise service quality to attract tourists, create photo-worthy environments and highlight experiential aspects that tourists can share with others, while also training staff to provide attentive service that generates positive emotional experiences.

To retain local customers, restaurants should focus on fundamentals, such as competitive prices and consistent quality and taste, while offering value propositions that appeal to the rational mind and build a reputation through authentic cooking techniques that locals prefer to mention in reviews.

“To integrate both target markets, restaurants may separate seating areas or dining times to cater to different atmospheres, or differentiate menus by highlighting signature experiences for tourists and the most favourite dishes for local patrons,” he says. “Both groups value quality, so maintaining high standards across all dimensions benefits all target markets.”

For the customers, Professor Zhang suggests taking extra steps to find reviews that better match their preferences and weigh the ratings differently. “Tourists wanting to eat like a local should prioritise longer reviews that highlight the restaurant’s location, cooking methods, value for money, with more analytical wordings. Meanwhile, those seeking a memorable experience should consider reviews highlighting service and ambience, with more photos and emotional expressions.”

As restaurants located in touristy areas may have their overall ratings inflated by travellers, the actual local feedback might be closer to lower-rated reviews.

Taking bias into account for better reviews

To ensure the users receive more accurate reviews, Professor Zhang suggests platforms identify which reviews are submitted by tourists and locals, then calculate separate average ratings for each group and provide functionality that allows users to view ratings from either locals or tourists.

With ever-evolving technologies, machine learning can be deployed to train algorithms to scan and detect reviewer types. Algorithmic adjustments can also be carried out to moderate ratings from identified tourists, considering their tendency to inflate their rating submissions.

For transparency, platforms can add labels mentioning locals or tourists to their reviews or display notes or disclaimers about potential bias in highly rated restaurants with predominantly tourist reviews. Platforms can also show trend lines of ratings separated by reviewer type over time.

Furthermore, platforms can suggest personalised recommendations. For example, platforms can recommend restaurants emphasising service and environment for travellers, while suggesting restaurants highlighting location, cooking quality, and value for local users.

“Overall, we find that the tourist bias is consistent across restaurant types, whether chains or independent, as well as across price levels and cities, making the findings broadly applicable,” Professor Zhang adds.

On this page...

A new study reveals how reputation, competition, and regulation shape the way businesses treat your personal data

Featured faculty: Jesse Yao
Written by Sally Ho

It’s fair to say that no one has been spared from the universal nuisance of scam calls and messages. Bad actors often find the targets from data brokers that collect personal data illicitly or purchase it from platforms to which users have entrusted their information.

Companies collect vast information about their customers to personalise advertisements and product offerings, yet many users are unaware of how valuable their personal data is. While the exact market value of personal data remains obscure, the global data broker industry is estimated to be worth US$277 billion in 2024.

Although countries around the world have issued stringent data protection laws, data misuse and breaches are increasingly common. Many companies mention their commitment to keep users’ data private in their privacy policies, but end up sharing or selling it to third parties without consent.

“Companies have a strong motivation to sell users’ data for immediate profits, and may not realise that they can actually gain consumer trust and loyalty by honouring their commitments to user privacy,” says Jesse Yao, Assistant Professor at the Chinese University of Hong Kong (CUHK) Business School. “Data sales are often not transparent or directly observable. Consumers can’t easily verify if a company is truly keeping its privacy promises.”

Ultimately, the nature of data collection creates a conundrum, where consumers are unsure about sharing their information as companies juggle between keeping and selling data. In his latest paper, Reputation for privacy, Professor Yao explains that such a dilemma is called the “holdup problem,” where two parties hesitate to cooperate due to concern about the opportunistic behaviour of the other party.

Can reputation protect data privacy?

Consumers who believe that a firm values its reputation over immediate profits are more confident in entrusting their personal data with the said firm. Building on this idea, Professor Yao models various scenarios of firm and consumer data interactions and then compares the outcomes in different market structures.

“Companies have a strong motivation to sell users’ data for immediate profits, and may not realise that they can actually gain consumer trust and loyalty by honouring their commitments to user privacy.”

Professor Jesse Yao

The analyses find that privacy protection thrives in a monopoly setting, where one firm dominates. In this setting, customers quickly find out which firm to blame for any data sales. Considering the risks of permanent reputational damage by selling or recklessly handling user data, the firm has a strong incentive to protect data to keep consumer trust and loyalty.

In a competitive market with multiple firms relentlessly trying to outmanoeuvre each other, reputation fails to serve as a privacy gatekeeper. Since consumers cannot easily identify which company leaks or sells their data, and individual firms do not bear the full consequences of harming consumer trust, the repercussions for the offenders are weak and temporary.

Take Apple’s operating system iOS, for instance. The company has quasi-monopoly power over its closed ecosystem and tight controls on both hardware and software. Consumers cannot install software from other manufacturers and will easily point fingers for any data breaches. Since Apple cares so much about reputation, it has strong incentives not to misuse users’ data.

On the other hand, the open Android ecosystem has numerous hardware manufacturers and software developers sharing the market. Consumers often cannot tell which apps were responsible for privacy violations due to noisy data flows and shared infrastructure, making the environment more prone to privacy risks, as evident in many news reports.

“We even find situations where two competing companies at first wanted to stick to their privacy commitment and were very willing to keep their reputations, but the temptation was so strong that they eventually failed to do so,” says Professor Yao.

Why privacy policy and regulation matter

Unfortunately, there is no purely monopolistic market in the real world, and people nowadays use multiple apps on various devices. Companies can easily gain short-term profits by selling data, and customers will find it hard to track down the responsible party. This lack of accountability reduces firms’ incentives to protect privacy. In this case, regulation plays a vital role in ensuring consumers’ rights.

Sound regulation can extend benefits for both consumers and companies. Protecting data privacy will make consumers more confident to share their information and to get better product offerings. At the same time, companies benefit from enhanced reputations and sustainable profits from loyal customers.

“Consumers and companies actually share the same interest, but the companies might be tempted to act against consumers’ interest by selling their data for short-term gains,” says Professor Yao. “Therefore, well-designed regulations are important to encourage businesses to behave responsibly, which benefits both consumers and companies in the long run.”

On top of that, regulators shall impose fines to deter firms from violating privacy commitments. “Liability fines by themselves are not enough, but liability fines with regulatory monitoring could do the trick,” he says.

The study offers a general framework for future exploration on the effectiveness of different policy tools and business practices, especially on how emerging technologies like AI will shape privacy dynamics.

As technology advancements tend to increase the benefits and complexity of data use, enforcing privacy in ever competitive markets becomes challenging. Regulators may use AI-based monitoring and enforcement, alongside liability fines, to encourage firms to maintain privacy commitments. Without strong and enhanced regulatory interventions, enforcing data privacy may become more and more difficult as technology progresses.

On this page...

Want to fight food waste? General eco-friendliness is just the starting point, and it’s not enough

Featured faculty: Elisa Chan
Written by Pan Jingyi

You might think that being eco-friendly, such as choosing sustainable products, conserving energy and water, and using public transport instead of driving, means you’re doing your part for the planet. But if you’re still tossing out wilted lettuce or forgotten leftovers, you’re missing a crucial piece of the sustainability puzzle.

Food production, transport, and disposal emit carbon dioxide, and when wasted food ends up in landfills, it releases an even more potent greenhouse gas: methane. According to the UN’s 2024 report, food waste accounts for 8 to 10 per cent of annual global greenhouse gas emissions, nearly five times the total emissions generated by the aviation sector each year.

“Enhancing individuals’ awareness of the risks and consequences associated with food waste should be a primary goal in addressing the food waste challenge.”

Professor Elisa Chan

Nevertheless, food waste often slips under the radar. The UN Sustainable Development Goal has aimed for a 50 per cent reduction in food waste at the retail and consumer levels by 2030, but an estimated one-third of the food produced globally for human consumption is still either wasted or lost in the supply chains.

A new study by Elisa Chan, Assistant Professor at the School of Hotel and Tourism Management at the Chinese University of Hong Kong (CUHK) Business School, reveals that even environmentally conscious people often overlook the serious consequences of food waste. People’s eco-friendly habits often stop at the kitchen door, and the solution isn’t just simply more education.

Being environmentally friendly only leads to real action against food waste if individuals also understand the actual harm it causes. In other words, general awareness is not enough to bridge the gap between pro-environmental beliefs and practical behaviours. “Enhancing individuals’ awareness of the risks and consequences associated with food waste should be a primary goal in addressing the food waste challenge,” says Professor Chan.

The culinary, restaurant, and broader food industry should tailor their sustainability campaigns to emphasise the serious consequences of food waste, thereby strengthening consumers’ negative attitudes towards wasting food. This will urge consumers to take more environmentally conscious actions, such as careful food planning, proper food storage, and mindful food consumption.

Why doesn’t a green mindset translate to action?

In collaboration with Carlos Martin-Rios of the EHL Hospitality Business School (Switzerland) and his PhD student Julneth Rogenhofer, as well as Maria-Eugenia Ruiz-Molina of the University of Valencia (Spain), Professor Chan explores Young professionals’ attitudes towards food waste: A global study on awareness, behaviour, and regional differences.

The team conducted an online survey of more than 600 young professionals across Asia, Europe, and North America, particularly those transitioning from academic studies into professional careers. “These young professionals are the future decision-makers and are encouraged to take on more social responsibilities,” says Professor Chan.

Participants were asked how strongly they try to avoid wasting food, how often they engage in various environmental actions, and how well they understand the negative impacts of food waste.

The researchers presumed young people’s general pro-environmental behaviours would be positively associated with negative attitudes towards food waste. However, the analyses show that there is no significant correlation between environmentally friendly actions and anti-food waste attitudes.

Professor Chan points to a psychological phenomenon called “intention-behaviour gap,” where people have good intentions but fail to follow through. “This is why most pro-environmental campaigns fall short,” she says.

The analysis of the data also indicates that the gap lies in the lack of awareness about the consequences. When people realise that throwing away food means wasting resources and harming the environment, they are much less likely to do it. “The risks and consequences awareness acts as a critical connector and turns a general concern for the planet into a tangible action about the food on our plate,” Professor Chan adds.

Why culture matters

The study also reveals cultural influences when it comes to food waste. Young professionals from European countries have stronger negative attitudes towards leftovers compared to their counterparts in Asia and the US.

Professor Chan notes that deep-rooted cultural values play a role and may still be ingrained in young professionals. For example, in Chinese culture, showing hospitality often means offering voluminous and variety of food at social events, shifting the focus away from overconsumption and leftovers concerns.

This doesn’t mean one culture cares about food waste more than another. Rather, it highlights that anti-waste campaigns need to be culturally tailored. As Professor Chan notes, “Food waste reduction messages and educational initiatives that neglect these culturally embedded values may be less effective.” She also points to traditional practices, like Shanghai’s “xiao die” or small plates, as ways to provide variety without excess in meal planning for both restaurateurs and individual households.

The education paradox

Education is often assumed to be a powerful tool to fill the gap. Therefore, the researchers specifically examined whether students in hospitality or food management programmes had stronger anti-food waste attitudes.

Surprisingly, the results show that students enrolled in such programmes do not exhibit stronger negative attitudes towards food waste. This finding may suggest that academic institutions have not effectively incorporated sustainable and ethical approaches to food waste management into their curricula.

“The current food waste discussion at school may be mostly focused on the supply chain aspect. For example, hotels and restaurants invest in technology to track kitchen waste, but such tools are not available at the consumer level,” Professor Chan says. “Supply side aspect is essential in food waste management and reduction from industrial perspective, but it seems we didn’t put enough focus on consumers’ individual consumption.”

Practical strategies for a greener future

Making food waste more visible can be a strategy to increase awareness. When leftovers and scraps are hidden away, it’s easy for people to ignore the problem. Professor Chan suggests restaurants or canteens could install transparent trash bags and bins, or use waste trackers, to make people aware of how much they’re throwing away.

From the consumer perspective, Professor Chan notes, “Another useful strategy is to leverage rather than change people’s habits in encouraging pro-environmental behaviours.” For instance, as consumers are habitually concerned about pricing, restaurants can charge meals based on the weight of the amount of food taken to discourage food waste. A buffet restaurant in Switzerland called Tibits is known for this pricing scheme.

Restaurants can also lead by example. Showing the details about their source of ingredients will highlight transparent waste-reducing practices throughout their supply chain. “Seeing how a restaurant expends effort in ensuring pro-environmental and pro-social operations would legitimise the notion of sustainable dining, which would motivate the guests to minimise food waste,” says Professor Chan.

Additionally, Professor Chan notes that creating interactive dining experiences that emphasise aroma, taste, and visual presentation can shift the focus from quantity to quality, making meals memorable for the experience. In doing so, restaurants not only help lower the environmental impact but also meet the growing consumer demand for sustainability.

On this page...

Pressure creates diamonds. Some companies have embraced this mindset to spark inventions

Featured faculty: Kevin Tseng
Written by Putro Harnowo

Public companies around the world are increasingly buying back their own stock. American firms are expected to set a record of US$1.1 trillion in share repurchases this year, led by Apple and Nvidia, while Asian companies have bought back US$266 billion in shares by August 2025, already 1.7 times more than last year.

Despite the growing trend, reactions surrounding this practice, simply called buybacks, have been mixed. American venture capitalist Nick Hanauer once said stock buybacks kill the economy, Senator Elizabeth Warren called them a tool to inflate executive pay, and the US Secretary of State Marco Rubio has aimed to limit buybacks so companies invest in capital spending instead.

Companies buy their shares back for many reasons, such as returning excess cash to shareholders or signalling confidence in their prospects. In some cases, buybacks can also help lift earnings per share (EPS), a key metric to show how much profit is attributed to each share, by reducing the number of shares outstanding.

Public firms typically set quarterly targets for their EPS, which also serves as a benchmark for evaluating firm performance. When firms miss this goal, their stock price and investors’ confidence may drop.

Treading on the verge of missing the EPS target put companies under “earnings pressure,” leading them to buy back some of their own shares. The caveat is stock buyback requires substantial cash, so companies often cut back spending on capital expenditures, employment, and research and development afterwards.

Critics argue that such a short-term cash flow strategy can compromise long-term innovation. However, Kevin Tseng, Professor at the Chinese University of Hong Kong (CUHK) Business School, proposes an alternative view. His recent collaborative study, Innovation under pressure, shows some firms increase their innovation output following EPS-motivated buybacks, even expanding their product lines beyond existing offerings.

“Buybacks spend a considerable amount of money, forcing firms to prioritise higher-value projects rather than indiscriminately trimming research and development,” he says. “Earnings pressure from buybacks does not appear harmful to innovation outputs because it can spur innovation efficiency.”

“Pressure does not necessarily suppress innovation but can sharpen resource allocation, encouraging focus on the most promising projects.”

Professor Kevin Tseng

How earnings pressure can drive innovations?

Along with the University of Illinois at Urbana-Champaign’s Heitor Almeida and Mathias Kronlund, as well as Vyacheslav Fos of Boston College and Hsu Po-Hsuan of National Tsing Hua University, Professor Tseng obtains the US firms’ financial information from the Centre for Research in Security Prices and Compustat, and collects EPS data from the Institutional Brokers’ Estimate System.

They focus on firms that have filed patents with the US Patent and Trademark Office from 1988 to 2015, within two years before experiencing earnings pressure. After excluding financial and utility firms, the final sample consists of 2,312 unique firms.

The researchers categorise the firms based on their innovative efficiency by assessing how effectively they have converted research and development spending into new patents. Those good at producing patents relative to their spending are classified as innovation-efficient.

Firms that are more innovation-efficient tend to cut their less productive projects and are more agile in reallocating their research and development towards higher-impact and the most profitable projects. Earnings pressure pushes these firms to produce more new ideas and inventions, particularly in the first year after buybacks, and persists for about three years before diminishing.

“Thus, the pressure can repeatedly shape innovation outcomes, but the benefits are not permanent, as they taper off unless renewed,” says Professor Tseng.

Innovation-efficient firms facing earnings pressure are also more likely to increase the number of researchers and scientists. By doing so, these firms are better positioned to access more new knowledge and explore more innovative technologies.

Further analyses show innovation-efficient firms increasingly use new knowledge by 2.5 per cent and boost their patent ratio in previously unknown fields by 3 per cent following earnings pressure. Using trademark data from the patent office, the researchers find such reallocation results in 4 per cent more product categories than before.

Meanwhile, less innovation-efficient firms are prone to cut research and development spending when facing earnings pressure. “Less innovation-efficient firms’ projects usually don’t lead to many patents, so when these companies face pressure to meet earnings goals, cutting back on research is the easiest and least harmful way to save money,” says Professor Tseng.

How do some firms thrive under pressure?

It may sound puzzling to see that some firms could increase the number of researchers and patents amid earnings pressure. Further analyses find innovation-efficient firms reallocate their expenditure by selling or closing their existing plants, then invest in new technologies.

Professor Tseng attributes such agility to short innovation life cycles and patent portfolio diversity. A short innovation life cycle means new products are developed, improved, and replaced quickly, such as those in consumer electronics, software, and fast fashion industries. In contrast, a long innovation life cycle means the products remain relevant for many years before being replaced, such as pharmaceuticals, chemicals, and industrial equipment.

Firms in industries with short innovation life cycles are used to a fast pace in technological development and have greater flexibility in reallocating resources across different projects. Additionally, firms with a diversified patent portfolio can easily switch focus and adapt to changes.

Finally, to be more innovation-efficient, Professor Tseng suggests firms examine how well their research projects perform, explore different technology areas, and keep an eye on how new and different their patents are. “By identifying and emphasising their most productive or novel lines of research, firms can weather earnings pressure without sacrificing long-term innovation,” he says.

Although the study focuses on US public firms, he believes the findings can apply to a wider context. Even without public earnings targets, startups and private firms often face external milestones, such as investor expectations or financing constraints, which generate similar short-term pressure.

“Pressure does not necessarily suppress innovation but can sharpen resource allocation, encouraging focus on the most promising projects,” he adds. “The key is to avoid across-the-board cuts and instead redirect resources towards higher-value opportunities.”

On this page...

Technology can empower individuals with disabilities to thrive in the workplace across industries, including hospitality, in which personal interaction is key to success

Featured faculty: Choi Sungwoo
Written by Pan Jingyi

It’s no longer shocking to see robots working in restaurants and hotels. To create more inclusive opportunities, some businesses employ people with disabilities to control these service robots. The operator’s face appears on the robot’s screen, maintaining a human touch while promoting inclusion.

These telepresence robots, also known as avatar robots, allow individuals with mobility disabilities to navigate around physical barriers and engage in work interactions remotely. “Robotic technology has significant potential to enable hospitality and tourism companies to hire people with disabilities despite their mobility restrictions, creating a meaningful social impact,” says Choi Sungwoo, Assistant Professor of the School of Hotel and Tourism Management at the Chinese University of Hong Kong (CUHK) Business School.

“Consumers generally prefer in-person interactions, either when employees use advanced technologies like wearable robots or traditional aids such as wheelchairs, compared to employees working through telepresence robots.”

Professor Choi Sungwoo

In addition to telepresence robots, wearable robotic technologies, such as exoskeletons and bionic arms, have also made it possible for people with physical disabilities to stand, walk and perform daily tasks that may be impossible with traditional prosthetics and mobility aids. While not widely adopted in workplaces, these technologies promise to further empower individuals with disabilities to participate in the job market.

Many companies and universities are now developing technologies to help individuals with limited mobility. However, little is known about how consumers actually feel, especially when different kinds of robotic technology are now available. To address this gap, Professor Choi, together with Sara Kim of the University of Hong Kong, recently conducted a study titled Consumer perception of employees with disabilities using robots.

“We found that not all technologies are perceived equally,” says Professor Choi. “Consumers generally prefer in-person interactions, either when employees use advanced technologies like wearable robots or traditional aids such as wheelchairs, compared to employees working through telepresence robots.”

The findings suggest that companies should feel confident about hiring individuals with disability, as consumers usually react positively as long as they can provide in-person service. As employing people with disabilities onsite not only promotes diversity and fairness but also enhances the overall consumer experience, companies aiming to create an inclusive environment may find assistive technology a worthy investment.

Mixed reactions to telepresence

In their first experiment, Professor Choi and his collaborator recruited around 200 participants from an online data platform, Prolific, and examined consumer reactions towards services provided by employees with disabilities in different scenarios.

Participants were asked to imagine visiting an art museum and rate their likelihood of choosing a guide from three options: an in-person using a wheelchair, a guide operating remotely via telepresence robot, and an in-person guide without disability.

The result shows that participants are less likely to choose the guide using a telepresence robot than either of the in-person guides. Furthermore, the guide in a wheelchair is sometimes chosen even more often than the guide with no disability. A second experiment conducted at a public university with 85 participants in a real-world setting confirmed these findings.

Psychological and social disconnection

Why are telepresence robots less accepted? The researchers attribute it to a psychological phenomenon where people subconsciously perceive them as lacking warmth and emotions. This happens because consumers struggle to look past the machine’s physical form to connect with the human operator, effectively overlooking the employee’s feelings and personal qualities.

When consumers speak with employees through a screen, they may feel as if they’re engaging with machines rather than a real person, due to the physical separation and the robotic nature of the interaction, explains Professor Choi.

“Service employees with disabilities working through telepresence robots may inadvertently create a sense of psychological distance from consumers, resulting in feelings of social disconnection,” he adds.

Physical presence is key in hospitality

To explore how to ease potential concerns towards employees with disabilities who provide service via telepresence robots, the researchers conducted a third experiment, recruiting nearly 200 people from Prolific.

This time, participants were presented with scenarios involving a restaurant server. Similar to the previous experiments, the server was either working remotely through a robot, working onsite using a bionic arm, or working onsite without disabilities. “We found that consumer perceptions of telepresence interactions were generally less positive,” Professor Choi says.

Across three experiments, consumers were less receptive to frontline staff working remotely. Professor Choi notes that the findings suggest physical presence is paramount in frontline service roles. No matter whether the staff is using a traditional wheelchair or a cutting-edge wearable robot, consumers’ responses are as positive as they are towards employees without disabilities.

Using technology more thoughtfully

In the wake of the pandemic and rising sustainability consciousness, hospitality and tourism companies face labour shortages and growing pressure to demonstrate social responsibility, making the hiring of people with disabilities both an ethical and strategic imperative. While employing remote-controlled robots sounds like a win-win solution, companies must recognise that some technologies foster acceptance and inclusion, whereas others may backfire.

Given consumers’ reactions to telepresence robots, Professor Choi recommends that companies committed to social inclusion prioritise assistive technologies like mobility aids or wearable robotics for frontline employees. He also encourages governments and NGOs to provide financial assistance to individuals and businesses that cannot afford these technologies.

Nevertheless, telepresence robots can still play a role. “They can offer valuable opportunities for individuals with disabilities to engage in the workforce, especially in behind-the-scenes roles like accounting and admin support, or when in-person work isn’t possible, such as during natural disasters, severe weather conditions, or personal injury,” says Professor Choi. In these cases, he emphasises the need for additional efforts to maintain personal connection through warmth and empathy, bridging the psychological gap that technology may create.

Finally, it is crucial to take the perspective of people with disabilities into prime consideration. Professor Choi notes that many people with disabilities are hesitant to enter the workforce, often due to a lack of confidence rather than employer reluctance. Based on the current research, he plans to explore how robotic technologies influence employees’ job performance and identity perceptions in his future studies.

“With the advancement of robotic technologies, we hope more people with disabilities can access these tools, navigate around physical barriers, and become more active in the workplace and society at large,” Professor Choi says.

On this page...

When facing sustainability backlash, strategic communication and transparency are key, a new study finds

Featured faculty: Zhang Qi
Written by Putro Harnowo

The global community is increasingly taking a stand on climate change. In July this year, the World Court called out countries failing to meet goals outlined in the Paris Agreement as violating international law, supporting the decision made by the Inter-American Court on Human Rights that a healthy climate is a human right.

While these advisory opinions are not legally binding, a growing number of climate change-related lawsuits are emerging worldwide. Businesses are held accountable for their contributions to or inaction on greenhouse gas emissions and for making false claims about their net zero commitments, a practice known as “greenwashing.”

Recent cases include a US$1.16 million fine imposed on online fashion retailer Shein by an Italian regulator for misleading environmental claims, and legal actions by the Australian consumer watchdog against the producer of popular sunscreen brands, Hawaiian Tropic and Banana Boat, over the claims of reef-friendly products.

sustainability, climate change — Reputational damage unfolds quickly as negative climate news sparks public scrutiny, spreading rapidly across the media.

These trends signal businesses to double down on or, at the very least, remain mindful of their environmental, social and governance (ESG) efforts. “Reputational damage unfolds quickly as negative climate news sparks public scrutiny, spreading rapidly across the media,” says Zhang Qi, Assistant Professor of Accounting at the Chinese University of Hong Kong (CUHK) Business School.

When facing an ESG crisis, Professor Zhang notes that issuing public statements in a timely manner is critical for businesses to mitigate the damage. However, companies often face challenges in weighing their options. Some prefer to disclose detailed information, and others choose vague messages, corporate lingo or even remain silent. “Negative news amplifies criticism and forces firms to respond, but the responses can either benefit or backfire,” he adds.

In a paper titled Responding to climate change crises: firms’ trade-offs, Professor Zhang and his co-authors underline that sharing information early is crucial for maintaining public trust, and disclosing veriﬁed information, such as audited reports, can enhance credibility and lower the risk of further damage. “Without credible, substantive information directly addressing the environmental incident, firms risk losing stakeholder trust, facing boycotts, or experiencing sustained negative sentiment.”

Although giving out detailed information is effective, firms may strategically balance it with a softer approach. For instance, when the consequences of giving less detailed information are negligible, the firm’s established reputation can substantiate allusive statements, but this only works if such reputation exists through the history of publishing ESG reports. So, how to tailor a media strategy that effectively addresses environmental crises threatening a company’s reputation?

Machine learning and social media can help

Along with Felix Fritsch of the University of Mannheim and Xiang Zheng of Nanyang Business School, Professor Zhang leverages machine learning to look at how companies balance different approaches when deciding what to say about climate issues after negative media coverage. For this purpose, social media provides an ideal setting for examining corporate climate-related disclosures, due to its speed, reach and ﬂexibility.

Specifically, the researchers turned to Twitter, which later rebranded to X in July 2023, and developed Climate-TwitterBERT that combines a keyword discovery algorithm with bidirectional encoder representations from transformers (BERT), a language model developed by Google. The researchers then analysed tweets posted by various firms from 2007 to 2020 and were able to pick 136,899 tweets relevant to climate change crises.

“We cover all US public companies with active corporate Twitter accounts, which naturally includes many well-known brands experiencing environmental incidents related to greenhouse gas emissions, global and local pollution, biodiversity or landscape impacts, and waste management,” says Professor Zhang.

“Reputational damage unfolds quickly as negative climate news sparks public scrutiny, spreading rapidly across the media.”

Professor Zhang Qi

The language model was further fine-tuned to categorise tweets into three types: hard responses, soft responses and promotional tweets. Hard responses provide verifiable and measurable information, such as emission reports detailing specific emissions data and quantitative progress toward climate targets. Soft responses reflect vague messages, including aspirational remarks about a commitment to sustainability without specific data and statements on moral positioning on climate issues. Promotional tweets focus on marketing green products and are excluded from analyses.

Their analyses found hard responses generate greater engagement, as measured by number of likes, retweets and replies, suggesting that the public appreciates empirical evidence the most. “Firms in the utilities, consumer cyclicals and basic materials sectors are most likely to issue hard climate disclosures,” he adds. “These industries face higher climate exposure and stakeholder scrutiny, increasing the incentive to provide verifiable information.”

Different problems require bespoke solutions

To understand the distinct approaches employed by the firms, the researchers classify environmental incidents based on media attention. An incident is of low inﬂuence if small NGOs, local media and local government report it. Medium-inﬂuence are covered by national media and government, as well as international NGOs. Highly inﬂuential ones are those covered by international media.

Soft responses typically follow low-influence incidents, but medium-influence crises prompt hard responses. High-inﬂuence disasters trigger both responses, supporting the idea that high-profile cases pressure ﬁrms to act across multiple fronts. Firms are 37 per cent more likely to disclose hard responses within three days of the incidents, particularly driven by medium and high inﬂuences.

Furthermore, having previously published ESG reports gives firms a higher degree of ﬂexibility in deploying hard and soft responses. In contrast, firms that never issued ESG reports tend to avoid providing soft responses, but are also very unlikely to provide hard responses, making the best use of silence.

Firms with ESG reports dealing with low-influence problems tend to use soft responses to provide a quick and straightforward approach. When medium-influence incidents arise, these firms prefer to elicit a hard response as the public expects a substantive explanation. Sustainability reports act as a buffer, allowing firms to respond with initial messages that may be irrelevant and provide more detailed information later.

It is worth noting that not all ESG reports are assured by a credible third party, such as an accounting or audit firm. Firms with assured reports have higher chances of issuing hard responses compared to those with unassured reports. Third-party verification strengthens the substantive messages to deal with medium and high-influence crises.

Interestingly, firms with unassured ESG reports are more likely to respond to non-environmental issues with hard responses on climate change to showcase their commitment to sustainability. However, firms with verified ESG reports tend to respond to non-environmental issues simply with soft responses.

Strengthening ESG disclosures with AI

While the study was conducted using social media data, Professor Zhang believes that the findings can be applied to broader platforms, although the magnitude may differ. “The underlying mechanism should apply across media channels,” he says. “Incidents covered by national or international media outlets tend to elicit hard responses, but whether traditional channels like broadcast and print media exert a stronger effect than social media remains uncertain.”.

ESG issues comprise three core elements. While climate action is powered largely by the environmental piece, ignoring the social and governance sides would be a costly blind spot. As the machine learning model used in the study now focuses on environmental issues, further refinement will be extended to analyse disclosures across wider aspects. “Our model is a general BERT-based framework that can be adapted to other ESG topics. While our current study focuses on Twitter and US firms, future work will extend the approach to different ESG dimensions such as labour practices and diversity.”

Amid the increasing scrutiny on sustainability issues and the advancement of technology, the study opens a higher possibility of applying artificial intelligence (AI) to improve ESG disclosure practices and identify greenwashing. Having said that, achieving greenwashing detection with AI requires clear terminologies and categories, robust fact-claim training datasets and refined models, which means much more needs to be done to make automated greenwashing detection feasible.

As recent large language models can further link the company’s ESG claims to external evidence, such as news articles, financial reports and regulatory filings, enabling scalable fact-checking, Professor Zhang anticipates that these technological breakthroughs will eventually provide easier access for stakeholders to monitor firms’ disclosure practices, forcing companies to provide truthful and high-quality ESG information.

On this page...

The relationship between success and self-development runs deeper than previously thought

Featured faculty: Li Wendong
Written by Gigi Wong

Those who have worked long enough may still recall the feelings of receiving their first paycheques. After years passed, many would have realised how far they had come, and more importantly, how their personalities or the people they know had evolved over the years. At some quiet hour, a random thought may pop up: Do we shape our income, or does our income shape who we are?

A new study suggests our personalities and our paycheques may be more intertwined than we realise, each influencing the other over time in a reciprocal dance. “When income increases, it signals that the current career progress is significant and the status in society is improving,” says Li Wendong, Associate Professor in the Department of Management at the Chinese University of Hong Kong (CUHK) Business School.

Professor Li’s latest study challenges the assumption that our personalities are relatively fixed and set in stone, that a person is always either an introvert or an extrovert, agreeable or argumentative, cautious or reckless. Furthermore, his paper titled A continuous time meta-analyses of reciprocal relationships between personality traits and income suggests that some personality traits can also predict future income changes.

His study also found an overlooked drawback of pay rises. Positive reinforcement that comes with income increases may lead employees to stick to routines and conventional ways of doing work, eventually limiting them from exploring new approaches to improve their performance.

career, personality — Rising income reinforces career success and social status, encouraging effective behaviors that may gradually shape personality.

“Such a positive reinforcement likely prompts employees to exploit the current effective work methods and strategies further. Over time, changes in such behaviours, thoughts and feelings likely foster personality change.”

“Thus, after a pay rise, organisations may also consider using incentives to motivate employees to come up with novel strategies to complete their work responsibilities and think out of the box to spur creativity and innovation,” he adds.

Which came first: pay rise or personality change?

Along with his PhD students, as well as Zhang Zhen of Southern Methodist University, Zhang Xin of Shanghai University of Finance and Economics, and Christian Dormann of Johannes Gutenberg University Mainz, Professor Li analysed data from 11 large national and regional longitudinal studies in the world involving 134,551 participants in total.

Studies were included if they measured at least one of the Big Five personality traits and individual income at three or more time points. These personality traits refer to five broad dimensions of personality that make each person unique, including emotional stability, conscientiousness (e.g., being dependable and hardworking), agreeableness (e.g., friendliness), extraversion (e.g., being outgoing), and openness (e.g., being curious and open to new experiences).

The team integrated findings across all the studies, which followed participants over periods ranging from four to nineteen years. The researchers also examined patterns of personality change and income shifts within the same individuals over time, painting a much richer picture than a simple snapshot study.

They found that conscientiousness and emotional stability tended to increase following a boost in income. The results also demonstrated significant effects of income changes on subsequent changes in both traits.

“While an income increase can make employees more stable and reliable, it may also reduce their tendency to seek novel strategies and goals — that is, their exploratory, creative and risk-taking behaviours may decline.”

Professor Li Wendong

In other words, a fatter paycheque seems to make employees more disciplined, dependable, poised, and unflappable over time. On the other side, as expected, people who became more conscientious and emotionally stable went on to see larger income increases over the next several years.

On the contrary, there was a negative reciprocal relationship between extraversion and income: being more extroverted seems to make people less likely to earn more money later on and earning more salary appears to make people more introverted at a later point in time. The effects of conscientiousness, emotional stability, and extraversion on income fluctuate over time, with the strongest impact occurring around one year after personality traits were assessed and tend to weaken afterwards.

Other personality traits have a negative relationship with the pay rise. The effect of income on changes in agreeableness is negative and the relationship between agreeableness and change in income is unclear.

“Increases in conscientiousness and emotional stability are positively related to subsequent increases in the employee’s income over time,” says Professor Li. “Our analyses also revealed that the reciprocal effect between extraversion and income was negative and significant across different time points.”

Success rewires our inner traits

Income-personality link cuts both ways, but what explains these intriguing connections between who we are and what ends up in our bank accounts? The researchers propose a few rationales, drawing on the cybernetic Big Five theory.

As mentioned above, pay rises reinforce our current strategies and behaviours at work, and in turn, motivate us to hone and further exploit those strategies. This results in deepening the motivational stability or conscientiousness quality, as well as emotional stability and social stability.

However, the cybernetic Big Five theory suggests that income increases are less likely to provide incentives for employees’ exploration and risk-taking behaviours, which may further enhance extraversion and openness traits. Based on empirical findings, the researchers posit that all five traits above correspond to one of the two broad dimensions of the Big Five that are associated with career strategies: stability and plasticity traits.

Extraversion and openness are included in plasticity traits, which encourage exploration by generating novel strategies when old plans prove insufficient, or when environmental changes occur, allowing for adaptation in uncertain environments. Stability traits comprise conscientiousness, emotional stability, and agreeableness, which primarily serve to maintain stability and prevent disruption to ongoing strategies for achieving long-term goals.

When experiencing a career setback, employees are likely to adjust their strategies and explore novel approaches or acquire new strategies to advance their careers —a characteristic of openness. They may also become more extroverted and enhance their social capital for career development or for exploring other alternatives. These changes over time may get habitualised across various life domains.

“The idea that income increases confidence and exploration remains a theoretical possibility, but our empirical evidence suggests that, over time, higher income may instead lead to more stability and less exploratory behaviour or lower extraversion, likely because employees settle into effective, established work strategies rather than seeking novelty,” Professor Li says.

Career development and pay incentives

The upshot is that our personalities are not entirely fixed to our job descriptions. As we progress through our careers, our personality inevitably gets imprinted by our professional successes and stumbles, but those experiences also get funnelled back into further shaping our future fortunes, for better or worse.

According to Professor Li, the findings have important practical implications for employees and organisations. Earning a higher income may make us more conscientious and emotionally stable over time, but also diminish our extroverted and explorative tendencies to find novel strategies.

Employees should be mindful of these potential shifts, he advises, to better navigate their career development. Organisations, too, should take note that the common workplace practice of using pay rises to motivate staff may have some unintended consequences on employees’ personality development.

“Given that many organisations tend to employ pay rises as an important incentive scheme to motivate their employees, the findings suggest that organisations should also be aware of positive and perhaps, unintended consequences of pay rises,” he adds.

“While an income increase can make employees more stable and reliable, it may also reduce their tendency to seek novel strategies and goals — that is, their exploratory, creative and risk-taking behaviours may decline. This may lead to adverse consequences for organisations in adapting to new and uncertain environments.”

To mitigate the adverse effects of pay rises, Professor Li emphasises the importance of acknowledging the benefits and risks, and find a way to balance stability and creativity. Companies need to adopt practices that encourage employees to explore new ideas, such as supportive management and more job autonomy, to foster intrinsic motivation and creativity.

On this page...

Recent research reveals why some executives negotiate easier goals while others embrace ambitious challenges

Featured faculty: Wu Fan
Written by Pan Jingyi

When Elon Musk declared that Tesla aims to have “hundreds of thousands, if not a million” self-driving taxis on US roads by the end of 2026 in a CNBC interview, it was another example of a CEO setting a bold target publicly. Leaders like Musk often position themselves as visionaries, unafraid to shoot for the moon and literally, Mars.

Not all CEOs share the same appetite. Since their compensation is often tied to company performance, many have chosen to set more realistic and sometimes conservative targets to minimise the risk of missing goals while maximising the likelihood of securing bonuses. So, what factors make a CEO set high goals or play it safe?

“CEOs typically have their own vision and expectations, and sometimes possess unique information that others don’t have, which means they can meaningfully influence how targets are shaped in ways that reflect their preferences, priorities and insights,” says Wu Fan, Assistant Professor of Accounting at the Chinese University of Hong Kong (CUHK) Business School.

The study found that CEOs with high prestige power are more willing to go the extra mile.

While company performance targets are formally approved by the board of directors, they are often the outcome of negotiations with the CEOs. Professor Wu’s recent research identifies two types of CEO power, structural and prestige, which influence how the executives set and negotiate these targets.

Structural power arises from formal authority within the company, such as when a CEO also chairs the board or holds a dominant position that protects him or her from being ousted by shareholders. Prestige power, on the other hand, stems from external force, such as the CEO’s social standing, reputation and elite networks. The study found that CEOs with high structural power tend to negotiate for easier targets while those with high prestige power are more willing to go the extra mile.

How past performance affects future goals

Professor Wu and his collaborators, Aishwarrya Deore of Georgetown University and Matthias Mahlendorf of Frankfurt School of Finance and Management, analysed CEO annual performance targets from various firms on the S&P index. The US Securities and Exchange Commission requires all US-listed companies to report the performance measures on which CEOs’ bonus payments are based.

In their study titled CEOs’ structural power, prestige power, and target ratcheting, the researchers examined how CEO’s past performance and their type of power affect their target settings in different ways. Companies normally use previous performance as a benchmark for establishing future goals, commonly known as target ratcheting.

The researchers observed an asymmetric target ratcheting trend in their overall sample, where the degree of target adjustments was noticeably unfair, as good performance brings tougher goals but bad performance doesn’t lead to much relief. “If the CEOs do better than the expected target, the future target tends to go up a lot. But if the CEOs underperform, the future target doesn’t fall by nearly as much, or might not go down at all,” Professor Wu says.

CEO’s structural power arises from formal authority within the company, while prestige power stems from external force.

CEOs with low power experience the most asymmetric ratcheting, making targets harder to achieve over time after satisfactory performance. The study found that having high structural and prestige powers helps shield CEOs from such demands and enables them to negotiate more effectively. However, there are fundamental discrepancies between the two types of power.

To bargain or to impress?

High structural power gives CEOs more favourable ratcheting as they can negotiate smaller target increases after achieving goals or larger decreases if they miss targets. “The position of CEO with high structural power in the organisational hierarchy grants them greater negotiation power in target adjustment decisions,” Professor Wu says.

CEOs with high prestige power can also settle for a more achievable target, but they tend to choose the opposite for several reasons. Pursuing ambitious goals signals their capability and confidence to their external network, including investors, analysts and potential future employers. “Accepting a more challenging target signals a CEO’s value to the market, which is essential to their career prospects,” Professor Wu adds.

“Accepting a more challenging target signals a CEO’s value to the market, which is essential to their career prospects.”

Professor Wu Fan

Moreover, prestigious CEOs often have greater privilege to explore other options at different companies and are less affected by performance concerns. “CEOs with prestige power likely have elite ties with other influential institutional players, such as executives at other firms, media and regulators,” he explains, adding that these connections can hinder the board’s oversight of CEOs and lessen the adverse consequences if CEOs miss their goals.

After controlling other variables, the results indicate that CEOs with structural power saw their target adjustments eight per cent lower compared to their counterparts with prestige power. CEOs may possess both types of power simultaneously, but they may exhibit different target adjustment behaviours than those with only one dominant power. When CEOs possess both high prestige and structural power, the two effects do not simply add up to weaken ratcheting further. Instead, prestige power tends to moderate structural power by accepting higher targets than CEOs with only structural power.

Target adjustment not only affects the CEO’s earnings, but also broader implications on the corporate culture.

Credible targets vs. unrealistic pledges

Target adjustment not only affects the CEO’s earnings, but also broader implications on the corporate culture. Setting relatively easier targets may risk weakening the CEO’s motivations to push for operational or strategic improvements if lower goals persist. “This could lead to a culture of complacency at the top, where performance plateaus and innovation slows,” Professor Wu says.

However, setting higher performance targets is not without its flaws. “While prestige-powered CEOs could foster a more aspirational performance culture within the firm, this could also come with increased risk-taking or pressure on subordinates to meet lofty goals,” he adds.

Pursuing higher goals demands greater effort, and Professor Wu notes that many CEOs with prestige power are aware of the trade-offs. “Ambitious targets can attract market attention, bolster a CEO’s reputation and unlocks benefits like prospective board appointments. So, even if there is a risk of missing targets and losing compensation, some see ambitious target-setting as a net gain in other ways, like elevating their profile as bold leaders or enhancing public visibility.”

Given that structural power may lead to stability and stagnation, whereas prestige power can result in growth and volatility, Professor Wu suggests that boards should also refer to external benchmarks, such as analyst forecasts and target adjustment trends among peer firms, to assess whether a proposed target is reasonable. “These comparisons can help distinguish between credible ambition rooted in performance potential and excessive promises that may not be grounded in reality.”

Industry and culture factors

Finally, while the study offers valuable insights on the influence of different types of CEOs in setting their target performances, Professor Wu expects the effects would be more pronounced in innovative industries, including artificial intelligence, technology, electronics, robotics and such, where leadership vision plays a larger role in shaping firm direction and investor expectations.

Individualism-oriented markets like the US may also create stronger incentives for CEOs to demonstrate ambition, as opposed to more hierarchical or risk-averse cultures like those in many parts of Asia, where the CEOs may be less inclined to overpromise.

However, he notes that CEOs are a highly self-selected group, and “individuals who reach these roles often share certain traits like confidence, risk tolerance and a desire for recognition, which may transcend national or industry boundaries.”

On this page...

The new normal leaves commercial property markets in many major cities with vacant premises. A new study offers an answer

Featured faculty: Desmond Tsang
Written by Putro Harnowo

A couple of years have passed since the pandemic subsided and borders reopened. Commercial real estate markets have shown signs of recovery globally, but the progress appears to be much slower than expected. Oddly, in some parts of the world, the market seems to be losing ground.

Commercial property values in European cities such as Berlin, Frankfurt and Paris have been slashed over the past few years. In the US, the office market in business hubs like New York, San Francisco, Los Angeles and Chicago has seen better days. Hong Kong’s commercial real estate market has also been feeling the pinch, with high vacancy rates in both the office and retail segments.

Desmond Tsang, Associate Professor of Real Estate at the School of Hotel and Tourism Management at the Chinese University of Hong Kong (CUHK) Business School, points out that one of the key reasons is reduced human mobility, referring to fewer people physically visiting or moving around, in commercial real estate. Remote working and digital commerce have become the new normal, but many real estate companies are slow to adapt.

“Reduced mobility is here to stay, and hence, commercial real estate demand would be under pressure, although some sectors or regions would be more affected than others,” he says. “For instance, offices in the US would be more deeply affected since remote working is more prominent there than in some other regions like Hong Kong or Singapore.”

His latest study, Human mobility and commercial real estate: Evidence from REIT operating performance, documents how changes in human mobility have become a permanent fixture. Along with Erkan Yönder of Concordia University and his PhD student, Halil Özgür, Professor Tsang found that for office properties, a one per cent decline in human mobility leads to a 1.9 per cent drop in rental revenue, driven by a sudden drop in demand for office space. As the supply cannot adjust quickly, higher vacancies occur.

For retail properties, a one per cent decrease in human mobility increases operating expenses by 2.1 per cent. This is because shopping centres and malls feature common areas like hallways, restrooms, seating areas, courtyards, and the like that require continuous maintenance and enhancement, where operating expenses are fixed regardless of the number of visitors. The shift in shopper behaviour drives retail owners to invest in additional features, promotional events and other campaigns to draw customers back.

The study urges the public and private sectors to monitor and consider human mobility factors when evaluating property performance. At the same time, property owners should develop strategies to respond to the structural shifts or consider repurposing spaces to accommodate such changes.

“The outlook is not all bad, as we may see some conversions from office to residential and some flight-to-quality phenomenon, in which high-end, amenity-rich buildings will fare much better,” Professor Tsang says. “Many brick-and-mortar department stores may face declining sales and the same goes for lower-tier shopping centres and districts, but there could still be value for some new retail concepts, such as experiential retail and food delivery hubs.”

“Reduced mobility is here to stay, and hence, commercial real estate demand would be under pressure, although some sectors or regions would be more affected than others.”

Professor Desmond Tsang

How mobility affects real estate performance

In the study, the researchers utilised Google mobility data that tracks the movement of US citizens to help mitigate the pandemic’s impact from 2020 until the tech giant terminated it in October 2022. They then matched data from Google with local points of interest from an open-source platform, OpenStreetMap, during the same timeline to observe actual changes in human mobility.

Mobility was observed to decline significantly from the second quarter of 2020, when authorities issued stay-at-home orders and closed public facilities, and then further eased by 2022, following the widespread availability of vaccines. Interestingly, despite the social restriction no longer in place, people were still 4.6 per cent less mobile in 2022 compared to before the pandemic. The analyses revealed a persistent decline in mobility due to structural shifts in how people live, work and shop.

To identify local factors behind such a change, the team collected demographic data from the US Census Bureau and then applied a machine learning approach, extending analyses through 2023. Locations with a higher proportion of residents holding a bachelor’s degree or higher were found to experience a greater reduction in mobility. The researchers argue that individuals with higher level of education are more likely to be able to work remotely during and after the pandemic.

To measure the impact of human mobility on commercial real estate, the researchers examined US real estate investment trusts (REITs) data from S&P Global Market Intelligence from 2019 to 2023. REITs own and operate numerous commercial real estate properties and are also publicly traded on stock exchanges, requiring their financial reports to be detailed and transparent.

Their analyses demonstrated that human mobility has a significant impact on REITs’ earnings, with considerable effects on office and retail properties both during and after the pandemic. For both types of properties, a one per cent drop in mobility lowers REITs’ net operating income by 2.8 per cent during the pandemic and by 1.8 per cent post-pandemic.

Hong Kong commercial real estate landscape

Although the study was conducted using US data, Professor Tsang believes that other developed markets are seeing similar trends of decline in human mobility, including Hong Kong. “There are many different reasons for mobility changes in different regions, making reduced mobility remain in many parts of the world in 2025,” he says. “For the US, it goes back to work-from-home and flexible work models. For Hong Kong, decreased mobility in some central areas stems from the travel flows between Hong Kong and mainland China after the pandemic.”

Mainland Chinese visitors have always been the largest group of tourists in Hong Kong, but their numbers have yet to return to pre-pandemic levels, according to the city’s tourism board. On the other hand, an increasing number of Hong Kong residents prefer to spend weekends and holidays in mainland China due to more affordable options, which affects the city’s commercial real estate rent.

As Hong Kong landlords and retailers battle against shifting consumer habits and a wave of store closures, Professor Tsang suggests businesses must devise plans to reduce operating costs while sustaining revenue and enhancing their brand reputation. “Focusing on new and innovative ways of shopping and retail is a possible strategy,” he adds.

For instance, he suggests encouraging more mixed-use developments by integrating living, working and leisure spaces with a 24/7 urban ecology rather than traditional retail, or adopting holistic district renewal instead of revitalising isolated building projects.

The study with US data found that mobility dropped significantly near commercial areas, restaurants, banks and schools due to pandemic restrictions and lasting changes, such as remote work. Areas near warehouses, industrial buildings and farms saw smaller declines, as these places often depend on physical presence and were less impacted by remote work trends. To extend the findings in the Hong Kong market, more data are needed to measure the magnitude of mobility’s impact accurately.

“Some companies track cell phone data for the Hong Kong market and this can be put to good use to understand travel patterns and office utilisation in different areas,” he adds. “It would be interesting to see whether mobility patterns in commercial hubs that are more remote would be different from those in downtown areas. This relates to the importance of location in human mobility and commercial real estate performance, which, in essence, we examine as a novel mechanism for mobility change.”

On this page...

artificial intelligence, video platform, social media

Generative artificial intelligence (AI) can improve content consumption by filling missing metadata, but a study finds it can’t beat humans, for now

Featured faculty: Philip Zhang Renyu
Written by Putro Harnowo

Social media has become more than just a means to connect, but also a way to build communities and markets beyond borders. When TikTok was on the verge of a US ban due to national security concerns earlier this year, a mass exodus of American users to another Chinese social media platform, RedNote, ensued as a form of protest.

A few months after the ban was postponed, ByteDance, the company behind TikTok, aims for 20 per cent annual revenue growth to US$186 billion, just a smidge below Meta’s projection of US$187 billion, this year. Platforms are not the only ones living the good life. Social media users can earn money from sponsors as content creators, and if they have amassed a large following, they can earn celebrity-level income as influencers.

artificial intelligence — Titles with hashtags and descriptions are the metadata used by the recommender system to improve content relevance.

The secret to keeping users engaged is matching content with viewers accurately through organic recommendations, where algorithms scan a vast array of video titles to suggest content relevant to users based on their viewing history and preferences. Users can also find content through a search query, but this accounts for only a fraction of total viewership.

“That’s the reason why content creators are encouraged to add titles with hashtags or descriptions, which are the only metadata used by the recommender system to improve content relevance,” says Philip Zhang Renyu, Associate Professor of the Department of Decisions, Operations and Technology at the Chinese University of Hong Kong (CUHK) Business School. “This metadata provides structured and concise details about the video, enabling the system to better understand the content crucial for the platform’s recommender system.”

Unfortunately, most user-generated content platforms suffer from metadata sparsity, where much of the content lacks descriptive titles or hashtags. Many of them then leverage generative AI to create and fill in missing metadata, but the outcome leaves much to be desired, as Professor Zhang and his co-authors found in their latest work, The value of AI-generated metadata for UGC platforms: Evidence from a large-scale field experiment.

“Although human-generated titles tend to outperform AI-generated ones, AI-generated titles remain valuable as they reduce the effort of title creation and at least provide a usable starting point for humans,” he says. “Even if AI performs worse overall, its output can still offer guidance, particularly for low-skilled producers who may struggle to craft effective titles on their own.”

When AI becomes a game changer

In the study, Professor Zhang, along with Goh Khim Yong of the National University of Singapore and his PhD student Zhang Xinyi, as well as researcher Sun Chenshuo, partnered with a leading short-video platform in Asia to conduct a series of experiments. While the platform boasts more than 300 million daily active users, only 60.7 per cent of the videos have titles, similar to other user-generated content platforms.

The team analysed more than two million users who produced more than 10 million videos. The videos were categorised into utilitarian videos, whose primary purpose is practical, informational, and educational, such as news, reviews and tutorials; and hedonic videos like entertainment or lifestyle content, such as vlogs, comedy sketches and travel videos. Before the experiment, 61.3 per cent of utilitarian and 57.3 per cent of hedonic videos had titles.

“AI-generated metadata can enrich the input to recommender systems, thereby improving the accuracy of user-content matching and enhancing overall personalisation within the system.”

Professor Philip Zhang Renyu

The platform introduced generative AI tools in July 2023 to create metadata by capturing multiple frames from the video, then extracting visual elements and text to write metadata that reflects the content. The researchers assigned creators randomly to either the treatment group, where they were given access to AI-generated titles, or the control group, which could only make titles by themselves.

The AI tool was found to increase the likelihood of videos having titles by 41.4 per cent and hashtags by 72.4 per cent. Overall, videos with AI-generated titles had an average 1.6 per cent increase in viewership and a 0.9 per cent increase in watch duration.

For low-skilled creators or those with a total number of followers and videos below the median among all creators, AI-generated titles helped increase viewership by 1.6 per cent and watch duration by 1.3 per cent. Looking beyond viewership, a deeper analysis of an additional dataset of almost 94 million videos found that the number of likes, shares, and follows was significantly higher among the group with AI tools.

“AI-generated metadata can enrich the input to recommender systems, thereby improving the accuracy of user-content matching and enhancing overall personalisation within the system,” Professor Zhang says. “Videos with AI-generated titles are associated with higher viewership diversity, suggesting that improved metadata enables the system to surface content to a broader and more varied audience.”

Human brain vs. machine intelligence

While the results seem positive, AI may not be suitable for everyone. Utilitarian content saw a 3.1 per cent decrease in viewership and a 3.0 per cent decrease in watch duration with AI-generated titles. For content that already had titles written by humans, changing them with AI decreased viewership by 37.9 per cent and 32.6 per cent in watch duration.

However, when creators edited the AI-generated titles, each 10 per cent change would bring 9.8 per cent more views and 8.2 per cent more watch duration. When the similarity between the AI-generated and the revised titles drops to 20 per cent, the videos perform better than those without AI-generated titles. This suggests that content creators should significantly edit AI-generated titles to maximise their performance.

Further analyses revealed that AI produces a basic thematic summary but often misses context-specific descriptions helpful for clarity and engagement. For instance, AI would create a title that says, “Enjoy the beauty of nature #ScenicNature,” and the content creator could enhance it to “Lush mountains and flowing streams: embrace nature’s serenity”. More detailed descriptors improve lexical richness and enable the recommender system to match videos with users more accurately.

Collaborative forces across platforms

Generative AI has been praised for its ability to create content, but it also raises concerns about originality and copyright issues. The study offers another perspective on how AI can enhance metadata to boost recommendation systems and content discovery.

Given that AI-generated metadata is beneficial for low-skilled and hedonic content creators, platforms should consider focusing on these segments when deciding to scale up AI tools. Prioritising new content creators to access AI would have the most immediate and noticeable impacts. Additionally, rather than automatically integrating AI titles, platforms should consider providing an option to edit to enhance the generated metadata. This strategy is not only beneficial for user-generated content platforms but also for other platforms where recommendations drive content consumption, including e-commerce.

“On e-commerce platforms, products often require metadata such as material, colour, and size, which can be time-consuming for sellers to input due to the large number of items. AI-generated metadata can help streamline this process and is likely to produce similar improvements in recommendation accuracy and efficiency,” says Professor Zhang.

With advances in large language models and multimodal AI that can process and integrate information from multiple data types, Professor Zhang believes that AI-generated metadata is likely to evolve far beyond simple titles to include richer and more structured descriptions. Future AI would automatically generate summaries, hashtags, emotional tones, scene-level annotations, and even inferred narratives that integrate visual, audio, and textual cues.

“Such metadata would offer deeper semantic insights into content, enabling recommender systems to make more nuanced and context-aware matching decisions, especially valuable for addressing cold-start problems and enhancing personalisation across diverse content types.”

Time to doubt the stars. Tourists are swayed more by vibes than value when rating restaurants

Why do tourists tend to give higher ratings?

How can restaurants and customers leverage tourist bias?

Taking bias into account for better reviews

A new study reveals how reputation, competition, and regulation shape the way businesses treat your personal data

Can reputation protect data privacy?

Why privacy policy and regulation matter

Want to fight food waste? General eco-friendliness is just the starting point, and it’s not enough

Why doesn’t a green mindset translate to action?

Why culture matters

The education paradox

Practical strategies for a greener future

Pressure creates diamonds. Some companies have embraced this mindset to spark inventions

How earnings pressure can drive innovations?

How do some firms thrive under pressure?

Technology can empower individuals with disabilities to thrive in the workplace across industries, including hospitality, in which personal interaction is key to success

Mixed reactions to telepresence

Psychological and social disconnection

Physical presence is key in hospitality

Using technology more thoughtfully

When facing sustainability backlash, strategic communication and transparency are key, a new study finds

Machine learning and social media can help

Different problems require bespoke solutions

Strengthening ESG disclosures with AI

The relationship between success and self-development runs deeper than previously thought

Which came first: pay rise or personality change?

Success rewires our inner traits

Career development and pay incentives

Recent research reveals why some executives negotiate easier goals while others embrace ambitious challenges

How past performance affects future goals

To bargain or to impress?

Credible targets vs. unrealistic pledges

Industry and culture factors

The new normal leaves commercial property markets in many major cities with vacant premises. A new study offers an answer

How mobility affects real estate performance

Hong Kong commercial real estate landscape

Generative artificial intelligence (AI) can improve content consumption by filling missing metadata, but a study finds it can’t beat humans, for now

When AI becomes a game changer

Human brain vs. machine intelligence

Collaborative forces across platforms

Want even more insights?

Subscribe to our digest