Amazon.com Bought Me for $100/Year

     It occurred to me a while back that every purchase I've ever made from Amazon.com (account opened in 2004) is stored in my user profile, and I could go back and read all of it. A dataset that extensive has to have something interesting in it. There are several big changes that affected my purchasing habits from start to finish, such as moving between continents multiple times, changing jobs multiple times, and purchasing Amazon Prime. The biggest question I had was whether or not purchasing Amazon Prime was a good idea. I put every purchase I've made in one big spreadsheet and plotted it as many ways as, usefully, I could imagine (Without going full ML).



     The data is pretty chaotic overall, and from 2007-2012 it makes sense only if you know where I was and what I was doing during that time, but three general trends are visible, as far as I can tell:
  1. Average purchase price was exceptionally high from 2006 to 2008
  2. Number of purchases and total $ spent increases
  3. Rate of purchase and dollars-spent increases dramatically after 2013.
     These are all pretty easily explainable, from my perspective. First, in 2007/08 I lived in Hawaii, where you're economically isolated from the rest of the world. If you want to buy a camera, you can always find A camera, but you can't necessarily find THE camera you want. If you want a specific high-end product, chances are you're out of luck. So even though I didn't use Amazon all that much at the time, I made several large purchases for specific items.
     Second, the number of purchases increases from each year to the next, with the exception of 2006->2007 (went from decent amount of free time to zero free time), 2009->2010 (went from lots of expendable income to no expendable income), and 2011->2012 (moved to Europe).
     Third, number of purchases and total money spent explodes after 2013. This is pretty straight forward. I bought Amazon Prime. By 2015, I'm buying most of my food at grocery stores and almost everything else from Amazon.



     I'm not big on holidays, but I was also curious as to whether or not my spending was seasonal. It appears no, not really. Spending is pretty linear, with the exception of years where there aren't enough points to identify a significant trend. No consistent big jumps at Christmas or any other time.



     I'm not really sure what this graph shows. I just came up with the idea of combining a histogram and a scatter plot and I thought it looked it cool. I call it a skyline plot, because it kind of looks like skyscrapers, and that's as creative as I feel today.



     This is the big one. After I bought Amazon Prime in November 2013, my purchase rate more than quadrupled. And it's only going up. Jesus Christ. Now for some non-graphical analysis.

     I pay $100/year for Amazon Prime. Assuming a constant spending rate (which obviously isn't the case), I was spending $2.50/day on Amazon before buying Prime, and $11.10/day after. Since Nov 2013, Amazon has gotten an additional $6,800 out of me due to Prime. I spent 9.24% of purchase price on shipping before buying Prime, which means I've saved an estimated $593 on shipping costs by buying Prime ($8,750 spent * 9.24% - $100/year * 2.16 years). So this answers the question "Did I save money on shipping by buying Amazon Prime". Yes, I saved almost $300/year.


     Did I save money overall? This is more or less impossible to answer, since I don't have a time machine and can't do a controlled experiment. However, I think I'm probably about even on dollars spent. When I do price comparisons between Amazon and Target (my main things-other-than-food store prior to Amazon), Amazon is generally a few percent more expensive, but I don't graze and I don't buy anything I don't specifically search for on Amazon. The main reasons I use Amazon are greater selection and greater convenience. I drove across town to go to Target a few days ago, for the first time in several months, because I needed one thing that was significantly cheaper there than on Amazon. They were out of stock. I had ordered it from Amazon on my phone before I made it back to the parking lot and it was sitting on my front porch 36 hours later. I could have gotten it that same day from Prime Now, but I didn't need it that badly.


     So did I save money overall? Probably not. But I like it a lot more.

Comments

  1. Could you post the code or a quick guideline to how you extracted and plotted the data? Interested to see my own purchasing patterns! :)

    ReplyDelete
    Replies
    1. Past orders are listed under "Your Account" > "Your Orders". I went through there and copied everything, order by order. You might be able to write a script to download all your old invoices and extract the info from those, but unless you have thousands of orders, it's probably easier to just copy by manually.

      Delete
    2. Actually, if you go to "Your Account" > "Download Order Reports", you can download CVS files containing all of your items, or orders, including shipping info. Much faster.

      Delete
    3. Oh nice!! That's really convenient Amazon provides that. I use Amazon...a lot.... So that would have been a lot to manually copy haha

      Delete
  2. There is actually a way to download a cvs with all your purchases. It can take a minute to generate but it goes all the way back to you initial purchase. Go to the 'your account' page and look for order reports

    ReplyDelete
    Replies
    1. That would have saved me several hours. Thanks.

      Delete

Post a Comment