Cell Phone Usage Trends

Be prepared to feast your eyes on a bunch of data!

The idea of this project came from two events. First, my Mom and sister recently joined my phone plan. As such, our costs are different and I'm curious to see how they actually use their phones. Second, for another project it was suggested that I learn how to use R - an open source statistical tool. So, I thought this would be a good occasion to take it out for a spin.

Well.. I didn't get too far on the R side of things. I was able to generate the chart at the top in about 10 minutes. It's pretty cool except for one not-so-small problem. Looking at the x-axis: what does 200 mean? Well, basically, it's the 200th call I during this billing cycle. I wanted to change the x-axis to reflect dates, but after 40 minutes of reading tutorials I gave up. So, not even a full hour in and I decided to move on.

With my brief time with R, it looks pretty cool and potentially very powerful. I'm not convinced it's the best tool for doing ad hoc data displays, but generating repeat reports or doing more advanced stats might be right up it's alley. I'll learn more over time, but for the remainder of my time I switched back to trusty ol' Excel.

Anyways, on with the findings!

I've broken this up into 3 parts: Phone minutes, text messages and data used. It's pretty cool to see the differences in how each of us use our phones. For context, here's what we're all using:

Dawn (Mom) -> iPhone
James (Me) -> iPhone
Jessi (Wife) -> iPhone
Matthew (Brother) -> Feature Phone
Lisa (Sister) ->  Feature Phone


Phone Minutes
First we'll look at minutes used. I don't differentiate on whether the minutes are free (off peak, A-list, or mobile-to-mobile), but just focus on how much the device is used.

Well, I'm clearly the winner when it comes to talking on the phone. This isn't a surprise since I use my phone for work. If you take a quick look a the graph up top you'll notice I either have short conversations or talk for about an hour. It's also not a surprise that my mom and Lisa have similar total minutes because a majority of their calls are to each other.

Next I decided to break it out by the day of the week.

Apparently Monday is a great day to talk. I'll have to ask Matthew who he chats with on Mondays. I'm also surprised with how consistent Lisa is. Since she has classes 2 days I week I would have expected to see those days lower. Interesting.

Then I got a little "stats fancy". I wanted to see the average length of a phone call. However, just looking at the averages wasn't good enough. I also wanted to see a relative spread. So, the dot is the average. The blue lines are a single standard deviation; it gives a feel for the range in which most of the calls fall.

So when my mom and Lisa talk, they're super fast conversations. This follows what I've observed: Lisa will call my mom multiple times a day just to say hi. My distribution is so spread out because of those office calls. The data actually looks bi-variate to me, which means there are really two patters merging. It would take a lot of effort (too much to actually do it), but it would be cool to separate my work vs. personal calls. I'd probably find that my patter falls closer to Jessi's.


Text Messages
Next we'll take a look at text messages. I bundled MMS and SMS together.

And I thought Matthew sent a lot of text messages! If Matthew and Lisa are any sign of where the world is heading..  Wow! You can't read it because of Matthew and Lisa, but my mom is right under 300 and Jessi & I are just above 200. I guess it's a good thing we have the unlimited plan. By the way, Lisa is averaging 80 text messages a day. Matthew only sends 51 a day. Who are they sending all of these to?

I thought it would be interesting to again break up the days of the week.

I'm trying to figure out what it is Jessi and I do on Tuesdays which results in a bunch of texts being sent. Matthew also drops way off on the weekends. That's probably because he's actually hanging out with his girlfriend and therefore doesn't need to text her as often.


Data Used
This next group was a total surprise to me. This does not include accessing the internet via wi-fi. I'm sure if it did, I would give my mom a run for it.

This is the second month my mom has had her phone and she's clearly figured out the power of the iPhone.

Next I looked a the number of times we access the internet using the data network. This includes anytime we check our email, use the browser, or use an app which connects online.

OK - weird. I use the phone for internet related activities just as much as my mom, but have the lowest overall usages. What's up with that? Furthermore, when Jessi gets online, she means business. What is she doing? I'll explain what I think it is in just a little bit.

First, here's another look at our average usage measured in Mega Bytes.

So, I had a little look into the usage pattern of my mom. It turns out she likes to use her phone for directions. When she does, it uses a ton of data. You can see these huge spikes (which makes the rest of the graph worthless) that go from 1MB to 15MB.

For me, it's just a bunch of super quick look-ups. I'll turn my email app on to let the messages load. Turn the phone off. Look at my news feed. Phone off. Look at email that's loaded. Phone off. Add something to my to-do list. Phone off. Add something else to my to-do list that I just remembered. Phone off. And so on. Without multi-tasking all of those are independent tasks.

Jessi can't stand the constant bouncing. When she decides to use her phone, she sticks with one app at a time. She also likes to use Pandora which uses a lot of data.


So there you have it. Lots of data as promised. No real conclusions, just interesting to look at.