Rating: Like vs Stars

2 Mar 2010 - 7:07pm
3 years ago
8 replies
2293 reads
bmeunier
2007

Here an example for two similar products: one received 100 «like» by buyers vs the other product received an average rating of 2 stars out of 3 by 200 buyers. Which one do you buy?

The concept of "like" vs "stars" do not have the same power, I'm pretty sure of it. And that, even for a product or for a blog post.  But wich one is better? So I decided to do severals smalls but efficient user researches and I've found that users 52% users preferred "stars" over "like" buy a product and 51% preferred "like" over "stars" to choose which blog post they will give their attention.

I'm going nowhere. What do you think?

P.S.: Don't let me start with 5 stars rating or notes on 10 points. That's bad.

Comments

2 Mar 2010 - 8:05pm
futurerichperson
2010

Context matters.

For example, on amazon.com people tend to conscientiously rate (stars) and comment on products to help others, to stand up for products they like, or to get others to listen to their advice. This tends to involve products that people purchase, have, and hold.

For youtube.com though, while they have the same 5-star rating capability, people treat it more as a like/dislike interface. If people like something they give it high marks. If people don’t like something, they tend to move-on by just leaving without rating items. There is no real vested interest on the part of the user.

http://youtube-global.blogspot.com/2009/09/five-stars-dominate-ratings.html

Don’t ask users what they prefer, look at their motivations and actual behavior in use, not comparisons.

As for which one I buy, it is the one with the best written reviews.

26 Mar 2010 - 3:11pm
Dave Edelhart
2009

On Mar 3, 2010, at 1:18 PM, futurerichperson wrote:

> Context matters. > > For example, on amazon.com people tend to conscientiously rate (stars) and comment on products to help others, to stand up for products they like, or to get others to listen to their advice. This tends to involve products that people purchase, have, and hold. > > For youtube.com though, while they have the same 5-star rating capability, people treat it more as a like/dislike interface. If people like something they give it high marks. If people don’t like something, they tend to move-on by just leaving without rating items. There is no real vested interest on the part of the user. > > http://youtube-global.blogspot.com/2009/09/five-stars-dominate-ratings.html > > Don’t ask users what they prefer, look at their motivations and actual behavior in use, not comparisons. > > As for which one I buy, it is the one with the best written reviews. > > (((P

3 Mar 2010 - 6:13am
giaq
2008

I think Stars work great when u have to make an idea of something you are going to buy. As an example: apple store. 

like/dislike work great for fast and not-so-critical decisions, as a content that is funny. 

you have to think of the degree of perceived importance of the action you are using the rating for. this action is buy something? the importante is high, so stars in my opinion are a better choice for a shop

3 Mar 2010 - 3:24pm
M Larsen
2009

You want to think of the value of the product in relationship to the value of the rating. Agreeing with earlier comments, the value of a video on YouTube is free, so a 2-option "Like/Dislike" makes sense. Once you add cost to a value judgement, your discernment changes. For instance, movie reviews most clearly show a typical value judgement such as "Watch as a matinee," indicating that the product is valuable, but not valuable enough to warrant a perceived standard cost. 50 million people may have watched Armageddon, but I live in a $10.50 movie town, and many of those blockbuster-watchers may be going to their town's $6 movie house. Is it worth $10.50? Is it worth $6? This is where I like to compare star ratings, half-star ratings...

Another thing about the "Yes/No" option vs. the more critical 5+ option scale is which scale appeals to which kind of personality. Some people are judgers, and are happy to lay down a Yes/No verdict on any topic. Others are more likely to weigh options, pros and cons, cost and experience... especially in a marketplace environment, these folks will be prosumers. Their thought process will take slightly longer. They will weigh multiple variables. If I was going to base a purchase solely on a rating, I would be more likely to buy a product whose rating had more thought put into it than a quick Yes or No.

Let's put it this way: I might buy a cup of coffee from a new shop if 5 people on the street tell me they love it. There could be 5 others on the street who hate it, and who may or may not tell me. But I'll try it. However, I won't buy a Nexus One because of 5 people who love it, and 5 more who may or may not tell me they hate it.

3 Mar 2010 - 4:20pm
Bryce Glass
2007

We cover different input mechanisms for ratings in our (being printed this week!) new book, Building Web Reputation Systems. Most of the chapters are available in draft form. In particular, you might find some of the following to be helpful:

What Makes for a Good Reputable Entity?

Determining Inputs

Common Explicit Inputs

And we have a pretty good series of Ratings-related posts on our blog as well (some of them are repurposed from the book chapters, others are original content.)

http://buildingreputation.com/writings/ratings/

Finally, Christopher Allen & Shannen Appelcline have a fantastic series on Systems for Collective Choice:

http://www.lifewithalacrity.com/2005/12/systems_for_col.html

That last link is well worth some quality read-time. Follow every link at the bottom.

26 Mar 2010 - 3:11pm
chris Cocca
2010

Both have utility. As usual, I think it depends on context. 
One thing you can't do with "liking" is provide anything that indicates a negative "score". This is great in social spaces, if you want to constrain social interactions to the positive. 
For products/e-commerce, as a user I don't just want to know the positive. I'd much prefer to see that a product is rated 1 out 5  than liked by 1 one person. It allows for the entire range of feedback, from positive to negative.
My two cents. 
Chris


2010/3/3 <"Benoît Meunier <contact@ixda.org>"@host.ixda.org>

Here an example for two similar products:* one received 100 «like» by buyers* /vs/* the other product received an average rating of 2 stars out of 3 by 200 buyers. Which one do you buy?*

The concept of "like" vs "stars" do not have the same power, I'm pretty sure of it. And that, even for a product or for a blog post.  But wich one is better? So I decided to do severals smalls but efficient user researches and I've found that users 52% users preferred "stars" over "like" buy a product and 51% preferred "like" over "stars" to choose which blog post they will give their attention.

I'm going nowhere. What do you think?

P.S.: Don't let me start with 5 stars rating or notes on 10 points. That's bad.

(((Please l
13 Apr 2010 - 11:49am
bmeunier
2007

Good thoughts.

Again, rating 1 to 5 is not very useful for anyone. What it mean 2 vs 4? Why did the user choose 3? What kind of indication that tell me or tell the user?

After reading all the materials you've all sent to me, my conclusion if, I don't see real studies to help me decide which one is better for the user and to make better business decisions. It's to new to settle on one concept.

17 May 2010 - 2:55pm
socialamigo
2010

Even with the kind of survey info you received, and it does sound like interesting data, you may have a situation where you may need to drive something into service - stars or likes - and track it to see what visitors/users are saying/doing.

Even with exhaustive testing, other decision elements may come into play that don't appear in surveying like graphical presentation might swing visitors to "stars" while brand messaging and "voice" might swing visitors to using "likes".

I'd be interested to hear back what your practical experience has discovered and writing about it on http://www.designforseo.org

 

Syndicate content Get the feed