Do We Really Need “Unbiased” Customer Reviews?

Do we need unbiased reviewsSince the advent of digital commerce customer reviews became an integral and important part of marketing. As adoption of online retailing grew, the role of customer reviews, published on e-commerce (retail) sites and third party reputation sites, became irreplaceable. Consumers, product manufacturers, distributors and retailers greatly benefit from the appropriate mining of this customers’ experiences ore.

The fact that most consumers trust customer reviews more than advertising, does not escape attention of some marketing “geniuses” who try from time to time to manipulate consumers by generating fake reviews. Inevitably these attempts backfire resulting in devastating loss to a company’s reputation and crippling legal costs. However, greed and stupidity often produce very slow learners and such attempts will never stop.

Publicity about use of fake generates a lot of attention and gives rise to variety of beliefs, such as:

  • customers reviews are mostly fake and only fools trust them

  • customer reviews are mostly written by people who want to rant about their experience. Therefore most customer reviews are negative and have little value.

  • most customer ratings are inflated and cannot be taken seriously.

As someone who makes a living mining millions of customer reviews every year for a purpose of product marketing, I can testify that all of these beliefs are mostly unsubstantiated. I could provide evidence to this effect, but this post is not about customer reviews authenticity or accuracy. It is about a role of “encouraged” reviews in the marketing value chain.

I came across an excellent article “Is Amazon Doing Anything To Fight Latest Wave of Fake, Paid-For Reviews?” written by Chris Morris and published on Consumerist website recently. This article seemingly touched a nerve as it was shared via various social media channels over 1,000 times during the first day it was published. The article questions the practice of encouraging (compensating?) customer reviews contributors by providing products (to be reviewed) at discount or free of charge. There is nothing illegal or even unethical about this practice, as long as it is transparent and properly executed. The “reference” marketing practice has been a staple in B2B world for decades, even though it is common knowledge that there are financial incentives involved.

However, this question really made me think:

“Shouldn’t the very fact that someone received something for free call into question that person’s ability to provide an unbiased point of view?”

This poses a very valid question, at a first glance. Upon consideration however, I could not recall ever seeing an “unbiased” opinion in a context of customer’s experience. I can recall balanced opinions, but not an impartial ones. This is not an exercise in rhetoric, it is an inquiry into the practical value of customer reviews.

As a shopper I see the value of a product’s average score (sum of the Liekert scores given by customers in their reviews and divided by the number of reviews for the product) only as the first filter on my journey to reduce the uncertainty of the purchase. I eliminate any product with a low average score without further investigation. I completely agree with Chris

” ignore the star rating, not just the average rating for a product but for each individual review. What is a 3-star product for one person might be a 5-star product for another. Read the actual text of the review and decide for yourself if it’s of any relevance to your needs.”

The key to success is in finding reviews of the customers who “hired that product”1 to do a “job” that is similar to the job I have to do. Here I seek partiality, bias or relevance to me and my “job”. The sentiment, price paid, discounts, etc., as long as it’s transparent, is much less relevant.

For the product “owner” (or competitor) the opinions of people who did not pay for the product is much less important or interesting. However, even in this case I would not call them completely valueless as there are opportunities to learn and utilize this content for marketing intelligence and benchmarking.

I think the real question is – can I trust a person’s ability to provide an honest opinion about using a product if this person was “paid” for providing it. We all have opinions about all sort of experiences, whether we were paid for sharing them or not. If “paid” is the only criteria for trust, how do we still believe spokespeople, journalists, doctors and other professionals.

The quality of the actual content within customer review speaks volumes about the author’s authenticity, whether they are paid or not. Mindful people can easily spot the dividing line between Motivation and Manipulation. Even algorithms can be created to discount integrity challenged content.

1 “Customers want to “hire” a product to do a job, or, as legendary Harvard Business School marketing professor Theodore Levitt put it, “People don’t want to buy a quarter-inch drill. They want a quarter-inch hole!” What Customers Want from Your Products




Posted in Customer Reviews Analytics | Tagged , , , , | Leave a comment

Big Data Is Not Just For Big Marketing

Big Data and NO PrivacyMost discussions involving big data focus on the needs of Big Sport, Big Businesses and Big Political parties, even though some of us find it difficult to see the difference between them. We all read about the big data technologies and methods deployed for the benefit of marketing organizations. Sometimes that was done to the detriment of consumers’ privacy and caused calls for additional regulation.

However, the advances in big data methods and technologies can directly benefit multiple constituencies, including consumers, and without infringing on their rights to privacy.  Here are some examples:

Deep analysis of online customer reviews in conjunction with pricing information can help consumers to find an outstanding value as illustrated in the example below. You can find the details by clicking on the link to the article. All the relevant data came from publicly available sources and was supplied voluntarily by the guests of the concerned hotels.

SB Best Value hotel

Deeper analysis and mining of the same data can help hotel managers to get a better understanding of what differentiates their property, in the guests’ perception, from  other properties in that area. Such mining reveals root causes behind NPS numbers. The traditional, not involving big data,  methods of uncovering such intelligence require series of relationship and transactional surveys. Correlation of such findings with occupancy rate patterns will likely to produce actions that result in gaining competitive advantage. Below is an example of intelligence that can be produced by such mining of opinions.

SB Hotels scorecard

The attributes in the table above are listed in order of frequency with which the guests have mentioned them in their reviews.

Similar analysis can be performed

In summary, big data is not about a size and number of data sets, it is not about mega technology investments or invasion of consumer privacy. It is about using more than one source of internal or external data to answer legitimate, real and practical questions that benefit your customers and your business.



Posted in Customer Experience CX, Customer Reviews Analytics | Tagged | Leave a comment

End of Product

Better Mousetrap catThe phrase “Build a better mousetrap, and the world will beat a path to your door” is often used as a metaphor about the power of innovation. If it is so simple, why has building a better mousetrap been so unrewarding? Why have most new product development efforts, that follow this wisdom, resulted in market failure?

Let’s start to peel this onion by asking what “better mousetrap” really means:

  • less expensive?
  • more effective, i.e. kills more mice?
  • more attractive in appearance?

the list can go on, and incorporation of these improvements may motivate a marginal increase in units sold, but the world rarely beats a path to a door proudly labeled “new and improved mousetraps”.

Those, who subscribe to the school of thought pioneered by Clayton Christensen, would argue convincingly that the “world” would rather not beat a path to a mousetrap, preferring the choice not to deal with mice at all. It is easy to understand that customers would prefer to pay for vermin being repelled from their dwellings, rather than to buy, set the traps, and deal with the dead mice disposal. In other words, a mousetrap is not a product at all – a vermin free dwelling is what customers want.

“The customers rarely buy what the business thinks it sells them. One reason for this is, of course, that nobody pays for a product.”  Peter Drucker.

Indeed, customers pay for obtaining a desirable outcome, they “hire” a product to do a “job”. From that perspective, they would likely pay more to obtain that outcome faster and with fewer complexities. In fact, minimizing the steps in the process of obtaining a desired outcome is at the roots of true innovation, the innovation that brings commercial success. In order to be successful an innovator has to be an expert in a “job” that customers would “hire” the new product to do. Yet, most want-to-be innovators are experts in products and/or technologies their companies sell. The critical pieces of the innovation puzzle are missing:

  1. clear understanding of the actual outcomes their customers desire,
  2. intimate knowledge of how customers experience the processes they use to obtain these outcomes, and
  3. empathy to motivate an innovator to find ways to simplify their experience.

Without these three ingredients, the “magic” of innovation is not likely to happen. For all intents and purposes, you are no longer set to design a “product”, but to design a “service” to deliver a simple and consistent experience for your customers, who are trying to obtain the outcome they actually desire. The degree of simplicity and consistency, and nothing else, will truly differentiate your offering from the others and will afford you to extract premium margins.

Posted in Customer Experience CX, Product Management, Product Marketing | Tagged , , | 3 Comments

Airbandb vs Hotels – Which Provides Better Guest Experience

AirB&B vs HotelI recently returned from visiting Italy and for the first time used to find and book accommodations for a considerable number of nights. Now, whenever I share my experiences of this wonderful country with my friends, they eagerly ask for my opinion about the differences between and hotels. It is not surprising if you consider that relatively few people had direct and multiple experiences staying with the hosts of this well publicized, but quite new platform. Most people, who ever traveled, have experienced hotel services.

My personal experience of was relatively positive as the process of room selection, booking and payment is substantially simpler than most hotel booking services. In most cases (81%) the description of properties and their reviews by the guests were specific, authentic and accurate. The price, and ability to pre-pay that eliminates effects of currency fluctuation, offer substantial value advantage over the comparable hotel room rates.

One of the most important attributes of customer experience is consistency of its delivery. Customers want to know that their expectations, created by company marketing messages, are going to be met. Every time. Most of us know what to expect when we book a hotel room. Judging by the analysis of 45K customer reviews, published on sites like and, on an average 87% of guests have their expectations met or exceeded. I could not extract a similarly sized data set of customer reviews, but the one I got (~3K) has a substantially lower average satisfaction rating of 69%.

It appears the customer experience delivered by is more like Forrest Gump’s “box of chocolates – you never know what you gonna get”.  While not explicitly promising to take care of any potential problems in interactions with hosts, the company does create the expectation of personalized service and local knowledge. I do appreciate the time difference complexity, but 12-15 hour lag in a e-mail response time to a critical customer problem, as we experienced with one of the bookings, is not acceptable. In terms of local knowledge and advice, only 60% of the hosts were willing or capable to provide anything of value.

airbandb guests growth started as a “shared economy” platform for low cost accommodation rental, “air mattress bed and breakfast” and was a better alternative to a hostel. Now, many “room” hosts are small hotels, real b&b and real estate investors. That transformation was not clearly understood by me when I made my reservations and my experience did not meet my expectations of authenticity in 3 out of 5 hosts.

In summary, I would recommend to experiment with only if you have more time than money as the cost advantage is indisputable at this time.


Posted in Customer Experience CX, Customer Reviews Analytics, Social Media Research | Tagged , , | 2 Comments

Customer Reviews – Trust, but Verify

Customer Reviews-Trust, But VerifyDuring the last two decades customer reviews and ratings became an ubiquitous part of e-commerce. Over 88% of consumers gratefully use them to select goods and services, whether they purchase them online or off. There are numerous studies that tie in availability of customer reviews to increases in sales, occupancy rates and other business benefits. Nevertheless, there are some people who question the authenticity or usefulness of customer reviews and ratings at large. There are certainly examples of attempts to manipulate consumers with fake reviews or inflated scores, but overall the usefulness of publicly available product/service reputation information is undeniable.

While there is a law to protect consumers from predatory marketing practices, it is important for consumers to learn how to assess the meaningfulness of information they use for purchasing selections. I would like to illustrate this point with personal experiences I had during my recent travel to Italy.

After researching the reputation of multiple airlines, that fly between San Francisco and Italian destinations, I booked the flight with Turkish Airlines. The company seem to combine high levels of satisfaction (according to online customer reviews) and very attractive rates.

Turkish Air sNPSWhile in air service and schedule reliability where quite good, the experience of a layover in Istanbul was a complete buzz killer. The airline did not live up to their online reputation by failing to assist us in any way. Upon return, I took a closer look at the online reviews and noticed that nearly 92% of them was published by passengers who traveled within very close distance and did not experience layovers. In retrospect their experiences were not all that relevant in supporting my selection. Iberia, that has similar reputation scores and not prohibitively more expensive, would have been a better choice as much higher percentage of their reviews were published by the inter-continental travelers.

Since I could not get any promised assistance from Turkish Airlines, I used to find the hotel with the airport shuttle service. I booked the site’s highest ranked hotel since my phone’s screen was too small for reading multiple reviews. Upon arrival to the Istanbul’s airport I discovered that the Coresh Suites Hotel shuttle service did not exist, the taxi costs equaled our room rate, the hotel could not possibly be reviewed as high as ratings implied. The bed was tiny and uncomfortable, and the bathroom was not clean. Overall, it was a miserable experience.

A day later I received an email from with request to review my experience. Looking at their request form I understood what caused discrepancy between the hotel high score rating and my actual experience – asks guests to rate a limited number of customer experience attributes, and then algorithmically generates an overall score that they call a customer review. In my opinion such an approach makes their “customer reviews” too easily manipulable and not trustworthy. Therefore from now on I will book my hotels from

The customer reviews work well when consumers can read actual descriptions of customer experiences and apply them to their own expectations of experience. The details of the described experience, the language and phraseology help consumers to recognize authenticity and ascertain meaningfulness of the content to their own situation. The statistical averages offer very little utility and the algorithmically produced ones are almost as good as “faked”, regardless of the intent.

Posted in Customer Experience CX, Customer Reviews Analytics | Tagged , , , , | 2 Comments