Word-of-machine effect with recommendation engines

Recommendation engines have been around for years, it is one of the early drivers of Amazon.com’s success (first written about in 2003), but recently they have evolved from a competitive advantage to a cost of doing business. A recent article in the Harvard Business Review, When Do We Trust AI’s Recommendations More Than People’s? by Chiara Longoni and Luca Cian, provides useful advice for leveraging further AI-driven recommendation engines.

Amazon rec engine

The central finding by Longoni and Cian is that accuracy is not the only element important in recommendation engines, but also context. They write, “simply offering AI assistance won’t necessarily lead to more successful transactions. In fact, there are cases when AI’s suggestions and recommendations are helpful and cases when they might be detrimental. When do consumers trust the word of a machine, and when do they resist it?”

Word-of-machine effect drives acceptance of AI recommendations

For recommendations to impact positively the customer experience, a key factor is what is being recommended. Longoni and Cian found that when customers focus on the functional and practical aspects of a product, its utilitarian value, they trust a machine. If they focus on the experiential and sensory aspects of a product, its hedonic value, they prefer and trust a human recommendation. If a customer is focused on utilitarian and functional qualities, then, from a marketer’s perspective, the word of a machine is more effective than the word of human recommenders. For someone focused on experiential and sensory qualities, human recommenders are more influential.

This phenomenon is referred to as word-of-machine effect. The authors write, “the word-of-machine effect stems from a widespread belief that AI systems are more competent than humans in dispensing advice when utilitarian qualities are desired and are less competent when the hedonic qualities are desired. Importantly, the word-of-machine effect is based on a lay belief that does not necessarily correspond to the reality.… The fact of the matter is humans are not necessarily less competent than AI at assessing and evaluating utilitarian attributes. Vice versa, AI is not necessarily less competent than humans at assessing and evaluating hedonic attributes.”

The word-of-machine effect extends to how people evaluate a product. If a product, like food (or a slot machine), that is evaluated based on hedonistic traits, is recommended by AI rather than a person, the consumer is more likely to discount its quality after trying it. In a study cited by Longoni and Cian, “[they] recruited 144 participants from the University of Virginia campus and informed them that we were testing chocolate-cake recipes for a local bakery. The participants were given two options: one cake created with ingredients selected by an AI chocolatier and one created with ingredients selected by a human chocolatier. Participants were then asked to eat one of the two cakes, which were identical in appearance and ingredients, and rate the cake for two experiential/sensory features (indulgent taste and aromas, pleasantness to the senses) and two utilitarian/functional attributes (beneficial chemical properties and healthiness). Participants rated the AI-recommended cake as less tasty but healthier than the cake recommended by the human chocolatier.”

Customers, however, trust AI-generated recommendations more when utilitarian features are most important. In another study, when consumers who were looking to purchase a winter coat were given recommendations based on practical features (warmth, breathability, etc), they preferred an AI shopping assistant over a human, and the more they cared about hedonic features, the more they preferred a human shopping assistant over an AI.

Augmented intelligence

Another interesting finding by Longoni and Cian is that consumers will embrace AI recommendations if they believe a human was part of the recommendation process. In one experiment, the authors framed AI as augmented intelligence that enhances and supports human recommenders rather than replacing them. The AI-human hybrid recommender fared as well as the human-only recommender even when experiential and sensory considerations were important. Going back to Amazon, they are quite adept at combining human and AI recommendations. As part of their recommendation process, they will show customers what other customers like them have bought, thus it is effectively the other customers, not AI, making the recommendation.

Human recommendations

Balancing word-of-machine effect

One way to mitigate word-of-machine effect is telling customers about the phenomenon (though wording it simpler than an academic would). Longoni and Cian found that, by prompting people to consider a different viewpoint about the recommender’s ability, it reduced the word-of-machine impact.

Another option that the authors found useful is using a chat bot. In one study, “[they] invited 299 online respondents to read about an app called ‘Cucina’ that would rely on AI to give recipe recommendations. Within the app, the participants were able to interact with a chat bot – an AI chef – that was programmed to assist them. The chat bot greeted each participant and introduced itself (‘Hi, Mark! I am here to suggest a recipe for you to try!’). The AI chef then delivered the consider-the-opposite protocol using a fun, interactive nudge: ‘Some people might think that an artificial intelligence chef is not competent to give food suggestions, but this is a misjudgment. For a moment, set aside your expectations about me. When it comes to making food suggestions, could you consider the idea that I could be good at things you do not expect me to be good at?’ That resulted in more favorable perceptions of the AI recommendation even when people were considering the experiential and sensory qualities of a recipe like taste and aromas.”

How to benefit from word-of-machine effect

As I mentioned earlier, effective AI recommendation engines have gone from being a competitive advantage to a must have feature of any online product. To provide the best customer experience, you need to understand the word-of-machine effect and biases it creates. It is important to remember these are biases, the AI can often give a better recommendation even if customers believe otherwise. Thus, you need to find a way to lead your customer to their best possible experience, without misleading them or creating the perception that it is a sub-optimal experience. The key is framing the recommendation in a way the customer will trust it, given what is being recommended.

Key takeaways

  • AI driven recommendation engines, popularized on Amazon and Netflix, have gone from being a competitive advantage in online and mobile apps to part of the cost of doing business.
  • While AI recommendations are often closer to what customers want, people trust human recommendations more when it is based on hedonic or experiential factors, and trust machines more when it is based on utilitarian factors; this is referred to as word-of-machine effect.
  • When presenting recommendations, you must be cognizant of word-of-machine effect; for recommendations around hedonistic properties try to bring in a human element (augmented intelligence) or explain that these are better recommendations.

What really went wrong at Quibi

Back in April, I wrote that Quibi had already lost and was not competitive and last week this “prediction” was confirmed by the announcement that it was closing down. Quibi, which raised nearly $2 billion in investment, was doomed from the start.

While I would love to dedicate a blog post congratulating my predictive capability, the lessons from Quibi are that there are clear drivers to success in the entertainment space. Rather than focus on the Monday morning quarterbacks (including the Quibi executives) who hypothesize why it failed (if they are so all knowing, why let it happen in the first place), let’s understand what drives successful entertainment (particularly games and streaming).

Not being feature driven

The first lesson is that a cool technology is not critical to consumers of entertainment. Quibi raised almost $2 billion by convincing investors that its unique technology would separate it from other streaming services (Netflix, Hulu, Amazon Prime, etc.). It touted the ability to switch in real time between horizontal and vertical viewing as its key differentiator as well as design around short-form videos. The founders believed this technology would satisfy consumers as it would make the content easier to consume while commuting and traveling.

The ability to switch between portrait and landscape mode is the latest example of a unique technology that ultimately consumers did not care about. Foldable phones are another case of something that looks amazing but the bulk of customers see virtually no value. In the entertainment and gaming space, the industry (particularly naïve investors) is littered with other great technology that customers never valued:

GonzoVR

  • Blockchain and crypto-currency Blockchain technology has been touted for years as the next big thing for gaming (both social and real money). I still see many people on LinkedIn talking about the missed opportunities in gaming with Crypto, but it is hard to fathom how everyone is missing an opportunity that has been top of mind for years. The people who tout crypto for gaming have never been able to connect the potential with any real problem consumers experience.
  • 3DO There have been many examples of game consoles that had better technology or more features but could not survive. One of the most salient is 3DO, a gaming console named Product of the Year by Time magazine in 1993. Despite being “Product of the Year,” it failed versus “inferior” products from Nintendo and Sega better delivered what customers actually wanted.

Quantity is critical

The second lesson from Quibi’s failure is the importance of having an (over) abundance of content. The biggest driver of revenue in the gaming and entertainment space is quantity of content, despite many people hoping, often wishfully, to find other drivers. When managing and growing games that are already live, the most certain way to increase revenue is to release more content (and the most likely way to see a decline is not to freshen up the content).

When building and growing a successful game, it is critical to create twice as much content as you would expect your most voracious VIP to consume. I have seen repeatedly companies build a six-month content pipeline that they were confident would satisfy all players only to have their top players run through it in a matter of a couple months (or even weeks). Once they run out of content, they churn to other options and are quite difficult to re-engage. Given that these are your most valuable customers, the cost can be enormous.

Companies frequently try to get off of this “content treadmill,” by innovating on features, events or putting in blockers to slow down consumption. While new features and great in-product events can add significantly to the value, they are not a replacement for more content. You still get a huge bump from new content. Conversely, putting in artificial blockers to slow content consumption almost always frustrates players and makes them more likely to churn, so while it does limit them to your available content, it does not negate the opportunity of delivering an abundance of content (you are actually covering up the underlying issue).

This effect is not limited to gaming, having a wealth of content is critical to consumers of entertainment across all media. Netflix has continued its huge growth by offering a surplus of new, original content weekly. Same driver for Amazon Prime’s success. Other companies solve the content treadmill by relying on user-generated content, which is almost endless. User-generated content is what has driven TikTok, YouTube and even social networks like Facebook to great success. Quibi tried to compete for the same users with a fraction of what customers could find elsewhere.

You also need exceptional quality

While quantity is critical, it does not negate the need for tent pole content. Successful entertainment companies build and maintain their position by creating must-have content that is strong enough to drive consumers to switch over to their channel or product. This content needs to be exceptional and draw in customers that previously thought they were satisfied with the options they are already enjoying.

This phenomenon is consistent in entertainment, from broadcast television to over the top media to gaming. The US television network NBC established decades long dominance driven by one or two situation comedies (Cheers, Friends, Frasier). HBO took the step from niche cable network to must-have content with The Sopranos and Sex and the City. Netflix moved from profitable movie rental company to an integral element of people’s lives with shows like House of Cards and Orange is the New Black. In gaming, Epic went from technology provider to gaming dominance with Unreal.

Quibi failed to deliver tent pole content that people felt they had to see. It invested in original content with A-list stars serving as actors and producers, much of which was entertaining but nothing that was unmissable. While engaging shows from the likes of Kevin Hart, Jennifer Lopez and Steven Spielberg may have entertained viewers, they did not have any flagship content that drove or even retained customers.

Need to plan for Complex Environments

I recently wrote about General Stanley McChrystal’s leadership lessons, with one of the keys being you need to plan for complex, rather than complicated environments. Being complex is different from being complicated. Things that are complicated may have many parts, but those parts are joined, one to the next, in straightforward and simple ways. A complicated machine like an internal combustion engine might be confusing to many people but it can be broken down into a series of neat and tidy deterministic relationships. Conversely, things that are complex, such as insurgencies or the streaming entertainment industry, have a diverse range of connected parts that interact regularly. A small disturbance in one place could trigger a series of responses that build into unexpected and severe outcomes in another place.

With Quibi, that disturbance was Coronavirus. In June, Quibi Founder Jeffrey Katzenberg said he attributed “everything that has gone wrong to coronavirus.” The reality is that to be successful in the entertainment space now, you need to build a company that is resilient and can adapt as the environment changes.

The Halo Effect confirmed

A final lesson from Quibi’s failure is that the Halo Effect is alive and well in the entertainment space. Quibi largely secured its financing on the reputation of its two leaders, founder Jeffrey Katzenberg and CEO Meg Whitman. I am a big fan of both. I love the content Katzenberg created as the Chairman of Disney for ten years and co-Founder of DreamWorks. I admire Whitman, who I credit with growing eBay into the company it is today and I hoped would run for the US Presidency several years ago. Success in these efforts, however, only highlights that you need a different skill set to grow a great streaming company.

The Halo Effect is attributing success or failure to an individual or specific action, which is often misleading. Success and failure are driven by multiple factors and there are no shortcuts to achieving great results. In the situation of Quibi, the leadership skills needed to succeed where very different to what was needed at DreamWorks or eBay:

  1. Start-ups are different. Scaling a company from zero to significant market share is very different than growing a large company. At a start-up, you need to find product/market fit and then grow from there. At a company like eBay, you already have the fit and need to focus on driving out the competition, marketing and margin.
  2. Streaming is a different world for entertainment. At a company like DreamWorks, you have a complicated problem to solve. You need to deploy hundreds of millions of dollars to create one piece of ultra-compelling content. In streaming media, you need a flow of compelling content while dealing with a complex environment (see previous point).
  3. Technology has evolved When Whitman led eBay, recommendations and personalization were a unique add-on. Now machine learning is part of the cost of doing business and you need to ensure every customer is getting the right experience for that individual. Companies like Netflix run thousands of tests a week and have terabytes of data on their customers, Whitman and Katzenberg did not have experience leveraging this data to meet customer expectations.

The predictability of the entertainment industry

Rather than chalking up Quibi’s failure to bad luck or the inability to create a hit in a hit-driven industry, it is the perfect example of the fundamentals to succeeding in the entertainment (including gaming) space:

  • Do not expect to rely on great technology or unique features to succeed, your customers need great content
  • Successful entertainment companies provide a glut of content so their customers never get satiated
  • You need to expand beyond great content to truly compelling and unique content that forces customers to go out of their way for your offering
  • Successful companies must be resilient and adaptive so they can adapt to a very complex environment where there are many unpredictable and uncontrollable events
  • The leadership team has to have the right skillset for this effort, not simply a track record of success in other ventures.

Key takeaways

  1. Quibi’s failure was very predictable and these predictors provide a framework of what companies need to do to succeed in the entertainment space.
  2. Quibi relied on a unique technology, the ability to watch content seamlessly in portrait or landscape mode, rather than relying on creating content people wanted. You cannot succeed in entertainment by relying on technology.
  3. Other key lessons are that great entertainment companies need to deliver a overabundance of content, much more than you expect even your heaviest users to consume, and some flagship products that forces people to try your offering.