Your browser (Internet Explorer 7 or lower) is out of date. It has known security flaws and may not display all features of this and other websites. Learn how to update your browser.
If you’ve ever struggled to convince a client on the benefits of allowing negative comments about their products, here’s one for you. A piece in Harvard Business Review (thanks @euanwilcox) reveals that getting bad reviews early on can actually be good in the long run.
The findings come from a study of 51,854 reviews contributed to Amazon, covering 858 books from 2000 to early 2004. Researchers Ye Hu and Xinxin Li found that recent user comments often focus more on addressing the previous reviews than evaluating the book itself. Meaning that books that initially received negative reviews often ended up with broadly positive reviews at the top, and vice versa.
Why? The authors cite two main reasons: expectations and self-selection:
Your friend’s assertion that “The Tree of Life” is the greatest movie you’ll ever see is pretty much a guarantee that you’ll be scratching your head afterward…An expectation disconnect is one of the factors that prompt readers to post reviews…
Self-selection refers to readers’ motives for taking the time to write. Some contribute reviews in order to correct what they see as misperceptions; others do so just because they like to be contrary in order to stand out. Either way, the result tends to be disagreement with the initial wave of reviews.
This follows previous research into Amazon reviews suggesting that the specificity of a review can be as telling as whether the overall sentiment is positive or negative.
In digital agencies, we love content personalisation. It makes us look smarter in front of clients and hopefully keeps things a bit more interesting for the people we’re trying to communicate with. But there’s also a downside to excessive personalisation – or at least that’s the message Eli Pariser made in his fascinating TED talk on the ‘filter bubble’ (thanks to Eduardoand Mervyn for sharing this).
Pariser describes how content providers like Google are increasingly personalising the information we see based on how likely we’ve been to click on similar topics in the past. The downside being that over time we see just the lazy stories that we like to read (eg What Justin Bieber thinks about the Royal Wedding), rather than the important events we should be aware of (eg the uprising in Egypt).
Pariser focuses on the impact of personalisation on news, but there’s also an important cultural implication. If what we consume becomes purely based on past behavior, over time we run the risk of creating a positive feedback loop where our tastes and attitudes become increasingly narrow and self-reinforcing. The fact that you’ve never visited a Vietnamese restaurant doesn’t mean you won’t enjoy the experience when you stumble on one.
The flipside of this argument is that we’ve always gravitated to information sources that resonate with our own world view – think Fox Sports or the UK’s Daily Mail. Furthermore, where once we only had a handful of TV channels, newspapers and local radio stations to form our opinions from, the internet now offers us access to attitudes we’d never of previously seen or heard about.
But it all comes down to our willingness to look for them. Personalisation has come about because most of us need help in making sense of the paralyzing volume of information available online. And in this way, it often helps us discover things we’d perhaps otherwise not have considered, for example Amazon’s recommendations.
If Google becomes our only frame of reference, there’s no doubt that Pariser’s vision is a chilling one. But ultimately it’s up to us to look beyond a single source of information and keep exploring.
Attempting to clear some of my Christmas shopping on Amazon, I noticed a rather curious new section under my personal recommendations – English Cuisine Bestsellers.
Currently sitting at number one is a family size consignment of Scampi Fries. It’s great that an artificially flavoured wheat-based snack has been recognised as a leader in the world of English cuisine. Other top sellers include Cheese Moments, catering packs of PG Tips and Buxton Still Water. Who’s buying this stuff?
In The Long Tail, Chris Anderson persuasively argued that the internet is facilitating a shift from mainstream hits towards increasingly niche interests, however, the continuing popularity of blockbusters such as Avatar suggests that the hits are as popular as ever. More interestingly, the experience of ‘long tail’ companies such as Netflix indicates that contrary to conventional wisdom, the engaged customers aren’t always the most satisfied.
It is logical to assume that, given the breadth of movies offered by a service like Netflix, customers with a specialist interest in films would get more pleasure from the DVDs they choose than your average punter. However, according to The Economist, the reverse is true.
A Wharton Business School study analysing reviews on Netflix found that blockbusters get better ratings from the people who have watched them than more obscure titles – even Transformers: Revenge of the Fallen, which was consistently slammed by the critics, is awarded four stars out of five. Another study by Harvard Business School apparently found a similar story on Quickflix, the Australian equivalent.
So why do hits continue to be so popular?
One of the classic arguments is that people still crave shared experiences. Another key factor is that consumer decision-making doesn’t seem to be keeping pace with technological change. Back in 1963, William McPhee’s 1963 study Formal Theories of Mass Behaviour explained the dynamics of a hit: “a disproportionate share of the audience for a hit was made up of people who consumed few products of that type”. The Economist argues that this phenomenon still rings true today:
“A lot of the people who read a bestselling novel, for example, do not read much other fiction. By contrast, the audience for an obscure novel is largely composed of people who read a lot. That means the least popular books are judged by people who have the highest standards, while the most popular are judged by people who literally do not know any better.”
My current Quickflix recommendations
Indeed, my initial interest in the dynamics of DVD ratings was sparked by my own slightly frustrating experience with Quickflix, which seems in line with the findings of the above studies.
As someone who watches quite a broad range of movies, I started noticing that most of the fairly obscure titles I am recommended by Quickflix seem to be rated no higher than two to three stars, while blockbusters typically receive three or four stars minimum.
The result: rather than encouraging users further down the long tail, the current user-powered recommendations may in fact be reinforcing the tyranny of the hit.
Shunning greater choice
With the amount of content on the internet predicted to double roughly every five years, we’re clearly not short of information to help us make these decisions. However, as Barry Schwartz’s The Paradox of Choice suggested, greater choice has been shown to make decision-making more complex and ultimately less satisfying.
Blockbusters receive consistently higher ratings on Quickflix
Arguably many of us gravitate to the familiar in order to avoid being overwhelmed by the amount of information at our disposal, subconsciously screening out options that require further investigation. According to Rory Sutherland, this is because, except in a few areas of specialist interest, minimising risk has a far greater influence on our decision-making than optimising choice. And if minimising risk is driving our decision-making, it is likely that we are applying similar criteria to rating what we bought.
The success of retailers such as Amazon demonstrates that serving niche interests can be an incredibly lucrative growth strategy. However, as marketers continue to evangelise the benefits of deeper customer engagement, it’s important to remember that these highly engaged customers can also be the hardest to please.
UPDATE 14/04/10: An article in FastCompany reveals that Zappos’ best customers return around 50% of the goods they buy, compared to an average return rate of roughly 35%. However, while free return shipping may make these customers more expensive to service, by reducing the risk of a bad decision, it encourages them to spend significantly more overall.
