I have been using Evernote since April 13th 2011. By that point the company had already raised $45.5MM, so I can hardly call myself an early adopter. However, I have now used the product for 34 months, and since October 21st 2013 I have used it as a paying business customer. I was curious about 3 things regarding my use of Evernote:

  1. Just how ‘sticky’ might Evernote be at this point for me? There are obviously many ways to define ‘sticky,’ but the number of notes I have in Evernote would be one factor that would make it very hard for me to move to an alternative application.
  2. How does becoming a business client affect my consumer engagement?
  3. What does my personal ‘digital exhaust’ look like?

To answer these questions I download an archive of all my Evernotes (see end notes for instructions). I then wrote an R script to analyze these files .

Turns out I have created 2,104 notes over 1031 days, for a total of ~1,079,303 words.

Over the course of those 1k days, there have been some weeks when I haven’t created any new notes, and others when I have created up to a maximum of 82 new notes.

Running a linear regression over the period I create on average 1.77 new notes each day. You can see that there have been two notable increases in my rate of note creation. The first was in September 2012 when I started to actively use Evernote in school, a clear testament to the importance of the edu vertical for Evernote. The other notable increase was in late October 2013, when we started using Evernote as an enterprise team (the blue line was our first payment). This data doesn’t actually include business notebooks, as you can only export an archive of personal notes. It is interesting that becoming an enterprise client has actually increased my personal usage of the product.

Looking at the number of words, using Evernote for education generated a substantial increase in the amount of content I was recording in the product. Becoming an enterprise consumer has also increased my production of content in the app, but not as substantially. I have generated an average of 1,014 new words each day.

As a final analysis I looked at a running cumulative number of words vs. cumulative number of notes. Since becoming an enterprise client each notebook has actually generated less words. In a prior script I wrote I counted the strings in the binary representation of files saved in notes. This generated a radically different graph, as the number of ‘words’ had dramatically increased since becoming an enterprise user. So the number of words has decreased per note (evident below), but the amount of content stored in Evernote as files has actually dramatically increased.

My Takeaways

  1. Evernote is likely an incredibly sticky product for me at this point as I have such a substantial amount of content in the product
  2. Becoming an enterprise client actually increases my usage of the product as a consumer, arguably making it even more beneficial for a company like Evernote to get enterprise adoption

End Notes

  • To download your Evernote archive log out of the Mac App and press ‘cmd+ctrl+option+e’. This saves a whole set of .enex files, one for each notebook.
  • You can only download personal notes, not business notes.
  • You can find the script I wrote to generate this data on Github
  • I just extracted the date for when I created each note. Some notes also include the date of when the note was last updated. There are not dates for when each update was made, which would be very interesting to see.

Secret has been blowing up as the ‘hottest’ new app. Just how ‘hot’ is it? Apparently, just as ‘hot’ as Flappy Bird!

Business Insider recently noted that Blockbuster passed on the opportunity to buy Netflix in 2000. Apparently, Blockbuster could have bought Netflix for $50M. It is very easy to Monday morning quarterback and see this is as an obviously poor decision. But if you were CEO of Blockbuster in 2000, would you have bought Netflix for $50M? The business looked very different back then, and fortunately for us some of the data is public.

If we assume Blockbuster only had 1999 data to consider, they would have seen a company with 12% gross margins, a $30M operating loss, and a $110 CAC. If they had 2000 data to analyze, they would have seen a company with annual churn exceeding 300%; Netflix added 515k in gross new subscribers in 2000, but at the end of the year only had 292k. Netflix had a $50 CAC, and a LTV of only ~$24 assuming average monthly revenue per subscriber didn’t change between 2000 and 2001. Netflix was on track to post an operating loss of almost $60M. Back then, it certainly wasn’t clear that $50M was a bargain for a company with terrible customer unit economics and bleeding cash. At the time Blockbuster had $4B in revenues, 70% gross margins, and $75M in operating income


You can download the data here.