AcademiaAgriculture & FarmingHumanitiesSocial SciencesSTEM

Amazon Price and Quality

Updated on September 13, 2016

Joined: 5 months agoFollowers: 0Articles: 1

You Get What You Pay for

We've all heard the old saying or variations of it. How true is this adage in reality, though? In this article we take a look at the price and average rating of more than five hundred items from Amazon. From hammers to hammocks all the way to cutting boards, we find that there is a slight (albeit statistically insignificant at this sample size) correlation between higher prices and higher product rating.

While this article does not entirely explain how price and quality are related, it does provide a glimpse into the price elasticity of various popular consumer products with respect to the quality of the product (as measured by the average five star rating on Amazon).

The Analysis

To measure the price elasticity of demand for given products, we collected data from Amazon between August and September of 2016. We collected two key data points: the average rating out of five stars as well as the lowest listed price of the product. Note that "quality" is difficult to quantify, thus the ratings are used as a proxy—we acknowledge that a users' opinion of rating could factor in price which would distort this analysis to a degree.

The products we evaluated are separated into five categories:

  1. Electronics
  2. Home
  3. Kitchen
  4. Outdoors
  5. Tools.

Below is a table of the items evaluated, the number of observations, the average star rating, the average price, the multiplier and the r2 value.

(click column header to sort results)
Item  
Number of Observations  
Average Rating  
Average Price  
Multiplier  
R-Squared  
Blender
25
4.02
$58.27
.385099
.126271
Bluetooth Speaker
25
4.46
$41.25
.00567965
Negligible
Hair Dryer
23
4.52
$32.67
.289241
.0274119
Laptop Bag
24
4.38
$34.98
.189537
.0098443
Mouse
30
4.23
$14.04
.367864
.000529
8x10 Frame
25
4.26
$13.89
.646534
.554922
Desk Chair
30
3.93
$73.92
.381206
.11602
Table Lamp
33
4.26
$23.51
.161089
.0386089
Towels
28
3.96
$41.59
.245156
.0750896
Twin Sheets
22
4.25
$21.37
.29064
.134088
Cutting Board
28
4.52
$18.16
.00488241
Negligible
Dish Set
27
4.35
$43.69
-.475564
.176554
Skillet
25
4.42
$32.73
.688566
.15374
Bike Helmet
29
4.43
$26.26
.41474
.0675681
Hammock
24
4.44
$34.37
.500268
.262739
Spinning Fishing Reel
30
4.42
$35.32
.206665
.0323875
20 Oz. Hammer
22
4.48
$22.32
.296853
.143347
Power Drill
30
4.42
$55.91
.689415
.173239
Tape Measure
28
4.29
$15.73
-.217309
.03817

To compare how price and rating are related for these items, we found the average star rating of the sample as well as the average price. Now, if a particular item is rated above the average rating for other competitive items then we would expect to see a price above the average price of these items. To compare price across groups we use the percent of the average price.

A couple of things appear interesting on first glance. Both dish sets and tape measures have a negative multiplier ,which means that for this sample, if items received a higher rating they cost less. Also, for the given sample cutting boards and dish sets do not show any correlation between price and rating, because of the near-zero r2 value.

When all items are viewed together, the linear equation (y=m(x)+b) that describes the relationship between the difference from the average rating and the percent of the average price is:


Percent of Average Price = 1.00861 + (Distance from Average Rating * .267374)


The r2 value for the sample is .0529325 and the p-value is less than .0001. From this we can see that rating alone does not explain a significant amount of pricing, but with other variables included likely (opinion) explains a percentage of it.

Across all 506 items, the average rating is 4.31. If we assume this is true across all products, the the predicted pricing structure within a given item category is as follows.


Rating
Percent of Average Price
1
12.4%
1.5
25.7%
2
39.1%
2.5
52.5%
3
65.8%
3.5
79.2%
4
92.6%
4.5
105.9%
5
119.3%

So, if the average price of widgets is $100, and the average rating of widgets is 4.31 then we would predict that:

  • a 1-star rated item would cost $12,
  • a 2-star rated item would cost $39,
  • a 3-star rated item would cost $66,
  • a 4-star rated item would cost $93,
  • and a 5-star rated item would cost $119.

Discussion

While we acknowledge that online reviews are indeed a proxy for quality due to the lack of reliable, quantifiable data, past research has found that "[P]roduct quality has a positive impact on generating positive online reviews."

Most of the sample of items chosen in this article were from the first one to five pages of search results on Amazon. This may create a bias towards the products that are purchased most frequently (which is not expected to be an issue), but it also created a sample more heavily weighted towards higher-rated items. This, again, is not expected to be an issue, as that is the group of items that consumers see, purchase and review most often and is thus most relevant.

In the end, it appears that consumers do still pay a premium for higher quality goods - except when it comes to tape measures and dish sets.

And the next time you're shopping, remember that the difference between a 4-star product and a 5-star product is about 13%, so don't get ripped off!

Click to Rate This Article