AI recommendation vs. user subscription: Which one’s better?
If the goal is to convert ads to sales, companies should strive for high conversion rates. Conversely, if the goal is to drive traffic and generate interest, companies should strive for high click-through rates.
Researchers from Lehigh University, University of Hong Kong, and Wuhan University published a new Journal of Marketing article that examines in-feed advertising’s performance across subscription versus AI recommended news feeds.
How do you get news on a daily basis? Subscribe to topics you are interested in? Or let artificial intelligence (AI) algorithms recommend news to you? Platforms like Google News, Twitter, and TikTok offer two distinct ways of curating organic content: through user subscriptions and via AI algorithms.
If, for example, you log into Twitter (now known as “X”) and open the “Following” tab, you will encounter posts from the sources you have subscribed to. Or if you open the “For You” tab, you will see content recommended by AI algorithms based on what AI predicts you are interested in viewing.
These different methods of delivering content provide distinct contexts for in-feed ads. However, little is known about how the performance of in-feed ads compares between subscription and AI-recommendation channels.
In-feed ads blend into your news feed, matching the format and style of content while clearly indicating their sponsored status. These ads can take various forms, from text-based ads on Apple News to eye-catching images on Instagram and engaging videos on TikTok. In-feed advertising has seen significant growth, with 58.3% of U.S. digital display spending allocated to these ads in 2018.
The authors explain that “in-feed ads ideally fit seamlessly into the organic content stream and their effectiveness is determined by both the ads’ attributes and where they are placed. We examine how the channel affects ad effectiveness and whether the effects also depend on ad attributes.”
They consider two core digital ad attributes:
Ad appeal that describes key content of the ad, which can either be informational (focusing on factual product information) or emotional (emphasizing the product experience through subtle feelings)
An ad link that leads to consumer action, which can be direct (e.g., “buy now”) or indirect (e.g., “click for more information”)
Channel Difference and Consumer Engagement
The manner in which content is delivered (through subscription or recommendation) has a big impact on how customers engage with that content. This, in turn, can determine whether they view in-feed ads as intrusive and if they decide to click on the ads and make purchases.
“We find that subscription and recommendation channels have two key differences: source credibility and content control. Subscription channels have greater source credibility and more content control because consumers can actively choose their sources, motivating them to exert greater cognitive effort in processing content. In contrast, AI-recommended content may be perceived as less credible and reliance on algorithms reduces consumers’ motivation to exert cognitive effort, leading to lower engagement,” the researchers claim.
Ad Intrusiveness and Ad Performance
In the subscription channel, high customer engagement with the organic content makes readers more goal-oriented, and they thus end up perceiving ads as more annoying and interruptive. However, customers who do click on an ad, despite the annoyance, show stronger interest and a higher conversion rate. By contrast, in the recommendation channel, customers are in an exploratory state and thus perceive ads as less intrusive. Consequently, customers are more inclined to click on ads in the recommendation channel.
The study uses two ad performance metrics for analysis: click-through rate (CTR), the ratio of clicks to exposures, and the conversion rate (CR), the ratio of purchases to clicks. In the subscription channel, higher ad intrusiveness leads to lower CTRs but higher CRs, while in the recommendation channel, lower ad intrusiveness may generate higher CTRs, but the proportion of genuine interest and subsequent purchases is smaller. “In addressing which channel has better ad performance, we show that the recommendation channel underperforms the subscription channel in converting sales, but excels at eliciting clicks,” says the research team.
Takeaways for CMOs
The study offers key lessons for Chief Marketing Officers:
If the goal is to convert ads to sales, companies should strive for high conversion rates. Conversely, if the goal is to drive traffic and generate interest, companies should strive for high click-through rates.
If advertisers’ goal is to maximize click-through rates, the optimal strategy is to release emotional ads with indirect links for both the subscription channel and the recommendation channel. Conversely, if advertisers want to maximize conversion rates, informational ads with indirect links work best for the subscription channel while emotional ads with indirect links are the best for the recommendation channel.
For recommendation channels, informational ads with direct links have the largest increase in click-through rates and the largest decrease in conversion rates. By contrast, emotional ads with indirect links have the largest decrease in click-through rates and the largest increase in conversion rates.
Structure of online reviews shapes their helpfulness
Reviews that grow increasingly positive are most helpful to readers, while those that turn negative are least helpful. For average-rated products, progressively negative trajectories enhance helpfulness, whereas reviews that start negative and grow positive are least effective.
A study of nearly 200,000 Amazon reviews shows that the usefulness of online product reviews depends not only on what is said, but on how the information is structured.
The researchers, from the Universities of Cambridge and Queensland, studied Amazon reviews for products ranging from clothing to food to electronics. They found that how the information is organised matters as much as what is said, and that different review structures are more or less helpful, depending on how highly the reviewer has rated the product.
Their results, published in the journal Scientific Reports, could help companies and third-party review platforms design their review pages to prompt the sort of reviews that will be most helpful to potential customers.
For example, a reviewer assessing a laptop might praise its performance and design while criticising its battery life, so how should such information be structured to be most useful to the reader? Should the review begin with criticism and end on a positive note, or start positively before turning to drawbacks?
“Any target of evaluation typically has both positive and negative aspects, which makes crafting evaluative messages challenging,” said co-author Dr Yeun Joon Kim from Cambridge Judge Business School. “The key question is how to structure these elements within a single message. For example, one might present criticism upfront and then move to praise, or instead integrate negative points within an otherwise positive evaluation. Yet research has paid little attention to this structural dimension.
“We wanted to understand whether certain structures are consistently more effective, or whether their effectiveness depends on the performance of the target being evaluated.”
The study was based on 195,675 reviews of 5,487 distinct products, and assessed performance and related factors, and a helpfulness score as measured by reader votes.
The researchers identified nine possible structures of online reviews ranging from Type A reviews that start positive and become more positive as they go along, to Type I reviews that start negatively and become even more negative – with lots of variance in between.
For highly-rated products, reviews that grow increasingly positive are most helpful to readers, while those that turn negative are least helpful. For average-rated products, progressively negative trajectories enhance helpfulness, whereas reviews that start negative and grow positive are least effective. For low-rated products, reviews are judged most helpful when they open constructively before introducing criticism.
“The results are nuanced but very clear,” said co-author Dr Luna Luan from the University of Queensland, who carried out the research while earning her PhD at Cambridge Judge Business School. “Looking at the overall sentiment of reviews does not fully translate into message effectiveness. It is the broader structure of sentiment – how positivity and negativity evolve throughout the review – that shapes how readers interpret online reviews.”
“Our findings have practical implications for how platforms and companies can design review pages in order to elicit the sort of reviews that will be most helpful to readers based on how highly products are rated,” said Kim. “For example, instead of simply asking ‘Write your review here’, the online review form could instead include micro-prompts that guide how reviewers structure feedback in a way recipients find most helpful.”
The researchers found the most commonly used review styles are not necessarily the most helpful to readers. In particular, for average- and low-rated products, the structures that reviewers tend to adopt often differ from those that readers find most useful.
This mismatch likely reflects different underlying motivations. Reviewers are not always writing to maximise usefulness for others, but may instead be expressing their own experiences, frustrations or emotions – especially when evaluating products of moderate or poor quality. As a result, review writing often serves both as information sharing and as a form of self-expression. This helps explain why widely used review styles do not always align with what readers perceive as most informative or helpful.
In an era when most TikTok videos are prerecorded, can a band with a new single create a tighter bond with fans by debuting via livestream instead? Can a business do the same when promoting a new product?
New research from the McCombs School of Business at The University of Texas at Austin suggests they could.
Since the pandemic, the livestreaming industry has been booming. The global market is expected to reach $345 billion by 2030, up from $100 billion in 2024. Nearly 30% of internet users watch livestreams at least once a week on social media.
Adrian Ward, associate professor of marketing, is one of them. A few years ago, he was viewing a livestream of a town hall meeting and found himself gripped by a speaker’s comments, feeling as if he were actually in the room. On reflection, he suspected it was the liveness of the event, as much as the speaker, that kept him glued to the screen.
“As we spend more of our time online and on social media, it’s worth asking how we can feel as complete and connected as possible in these spaces,” Ward says.
Live and Let Stream
With Alixandra Barasch of the University of Colorado Boulder and Nofar Duani of the University of Southern California, Ward began to investigate what he calls the “mere liveness effect”: the idea that simply knowing an event is streaming in real time makes a viewer feel more connected to the performer.
The researchers ran five experiments with 3,500 total participants. By manipulating various factors, they compared how, when, and why viewers reacted to watching livestreams versus prerecorded videos online.
In one experiment, participants watched live or recorded videos of their choosing on the platform Twitch. In another, they viewed a performance by the R&B cover band Sunny and the Black Pack, either live on YouTube Live or its recording the next day on YouTube.
In a third, the researchers created their own streaming platform to show participants identical videos, manipulating whether the content appeared to be live or prerecorded.
The experiments provide evidence that watching an online performance in real time boosts several aspects of the viewing experience:
Connection. Viewers in one experiment felt 7 percentage points more connected to the performers in the live video. Another experiment showed the effect was even stronger when viewers believed no one else was watching.
Enjoyment. In another experiment, viewers enjoyed the live video 5 percentage points more than the prerecorded one.
Engagement. Real-time streams carried a “liveness lift.” Viewers chose to continue watching longer, and they were more willing to follow and subscribe to the live streamer’s channels.
A common factor underlying those effects was a heightened sense of presence, Ward says. “When we watch something live, we are psychologically transported there.
“It’s not that there’s actually something different about the video itself. It’s that we know that it’s live right now, and that breaks down barriers between our world and the world on the other side of the screen.”
Lessons for Liveness
One quality weakened the liveness effect: not being able to see a performer’s face. When viewers saw only a musician’s hands, they felt less connected, even though they were watching the same performance.
The findings have implications for marketers, platform developers, and content creators, Ward says. In an age when people increasingly meet their social needs online, going live can benefit streamers by motivating audience engagement.
As a follow-up, he’s working with a graduate student to study whether the liveness effect translates into greater brand trust or sales.
“From influencers to businesses, it’s about the experience of real people seeing other real people live and in the moment,” Ward says. “It makes you feel like you’re sharing something.”
Renting out your place? Human connection key to a successful holiday rental
Warmth, friendliness and a sense of belonging, or the “homely” side of the experience, strengthen guest loyalty, making them more likely to return to the same host. However, these feelings alone didn’t necessarily make guests more likely to recommend the property to others.
Striking up a connection with the property host is the factor that drives repeat bookings on holiday accommodation platforms such as Airbnb.
This is according to a new study, carried out by universities in the UK and Iran and published in the February 2026 edition of International Journal of Hospitality Management, that suggested that quality and value of accommodation also play a part in guest satisfaction, but personal connection is key to people deciding to stay again.
The research analyzed hundreds of online guest reviews and conducted in-depth interviews to understand what shapes guests’ evaluations of their stays in what is known as “peer-to-peer accommodation”.
Conducted over six years, the study shows that guests assess their stays using emotional cues such as warmth, atmosphere, and aesthetics; and cognitive cues such as cleanliness, safety, and convenience.
The study found that warmth, friendliness and a sense of belonging, or the “homely” side of the experience, strengthen guest loyalty, making them more likely to return to the same host. However, these feelings alone didn’t necessarily make guests more likely to recommend the property to others.
In contrast, affective and intellectual experiences – the enjoyment and perceived value of the stay – were stronger predictors of recommendations and positive reviews.
The research also examined how the quality of booking websites, such as Airbnb’s platform, influences guest behaviour. Although the website didn’t change how guests felt about the property itself, a well-designed and trustworthy site directly boosted guest loyalty and word-of-mouth.
Co-author Nektarios Tzempelikos, Professor of Marketing at Anglia Ruskin University (ARU), said: “Guests think carefully about both emotional and practical aspects before booking. Hosts who focus only on one side – either charm or functionality – may be missing the bigger picture.
“Platforms like Airbnb thrive when they’re designed for trust. Guests return to sites that are clear, reliable and easy to use. But it’s not just about tech, it’s about people. The most memorable stays come from warmth, authenticity and genuine local connection.
“By encouraging friendly, personal communication between hosts and guests, and balancing smart technology with a human touch, platforms can create experiences that feel less transactional and more meaningful.”
The study was carried out by researchers from Brunel University, University of Bradford, Newcastle University, Anglia Ruskin University and the University of Tehran.