On the presentation of online shopping customer reviews

Reviewing one of the reports submitted to the organizing committee of the conference CHI2007 , experienced a sense of discovery, which arose as a constructive criticism of the approach described in the report.
I have no right to retell the text of the report, because I can not talk about the authors and their report before they themselves make it public. I will say that they considered the question of how best to present feedback from online store customers. The situation is well known to all: you are on a page offering to buy, for example, an mp3 player. But you don't have enough information to make a decision. You want to know what those who bought it say about this model.
Most online stores provide you with customer reviews. When there are many reviews, the problem arises, how to present them in the most convenient way. You hardly want to read all the reviews just to find out people's opinion about one of the player’s functions, for example, about the actual duration of his work without recharging, or about how well this model records sound from a microphone.
The authors of the report choose a very technological solution. Using modern means of automatic text summarization, they make a summary of the functions of the model. From statements, those statements that mention a function of the presented product are automatically selected.

Of course, the automatic referencing tools are still not accurate enough. Such a technique can snatch out irrelevant phrases, only because they contain, for example, the word "record" or "battery". And at the same time, to pass more significant information, if in its description there were no keywords, but there were words closely connected, for example, “electricity” or “nothing is heard”.
But my criticism is not connected with this, but by the very method of obtaining customer feedback, which was open, of course, not by the authors of the report, but widely distributed on the Internet.
Before the Christmas holidays, I wanted to buy a digital camera, because was going on vacation. Under the influence of a friend, I decided to buy a camera with a hard drive. Having never been interested in such cameras, I was forced to do some marketing. I visited a large number of domestic online stores, collecting bit by bit an idea of ​​what kind of camera I want to buy. Customer reviews would be very useful to me. But almost everywhere I was met by the inscription " No reviews! You can leave your review first " and a link to the form for writing a review.
Interesting, more precisely, strange! Is there really nobody else interested in these cameras except me? Maybe you should not buy something that no one needs? Purchase motivation is clearly reduced.
And most importantly, how can I leave feedback about the camera, which I never held in my hands? It looks like the pages I was on are not visited by those who have already acquired such a camera. Why should he come here again? After all, these pages are intended for those who only choose, but they still can not say anything about the product.
As is often the case, there is confusion in the processes. With such processes, it is necessary to engage in manual referencing, and to send forces to correct the process.
Now, in fact, best practice, which I want to offer for the described case.
First, remove the words "No Reviews", where there are none. The phrase is alarming, raises unnecessary doubts. The form for writing a review can be left, although no one will use it. What for? I will explain later.

Secondly, all Internet shops register their customers, which means that there is an opportunity to contact the customer after they have made a purchase. Suppose the store has a new product for which there is not a single review. After someone bought this product through the store, the system after a couple of weeks (to let the buyer try out the product), writes him a letter asking for the lowest price to answer a few questions.
Questions related to the functions of the goods. For example, “Did you use this player as a voice recorder, if yes, what were your impressions of using this function?” The system collects answers to questions that, with the described approach, have a structured appearance, and therefore, without complex automation, they can be presented in a convenient form for the reader.
Third, as soon as the system has collected enough reviews on the product or one of its functions, it stops asking the customers opinion, first, not to force people to answer questions that are no longer needed, and second, not to create a problem with the presentation of volumetric data.
An additional advantage of the described method is that the online store communicates with its customer, and this is important for creating an image of a company that cares about its customers and is interested in their opinion.
To stimulate the writing of reviews, customers may receive discounts on subsequent purchases, which gives a logical basis for integrating the described method with the loyalty program.
It remains to answer the question of why you need to leave a form for entering a new review on the product page, although no one will use it.
From experience, I know that users are nervous when they cannot understand the source of certain data. In our case, it may seem to them that customer reviews are fabricated, because the process described is not visible to them. The form is a self-explanatory explanation of where the reviews come from. She, of course, will not convince those who are configured to see only deception around, but she will allow others not to ask unnecessary questions.