4 ms·
The great thing about user numbers is that you have so many to choose from! There are straightforward, well-used definitions of "user" for which I have 60, 1.5
by patio11 17y ago
The great thing about user numbers is that you have so many to choose from!
There are straightforward, well-used definitions of "user" for which I have 60, 1.5k, 2.9k, 50.0k, and "well over a hundred thousand" users. They're all equally true and none of them singly or in combination tell you a thing about the health of my business unless you're intimately acquainted with it.
- mahmud 17y agomy favorite metric is: `wc /var/log/httpd/access_log` :-)
- wildmXranat 17y agoHey man, how did you know ?? My patented uniques metric: cat access.log | awk -F " " '{ print $1 }' | sort | uniq | wc -l
- lincolnq 17y agoDo you know about cut? You can replace awk here. cut -d ' ' -f 1
- inferno0069 17y agoAlong the lines of lincolnq's comment: |sort |uniq can be shrunk to |sort -u
- jarin 17y agoWell you know what they say: there are user counts, damned user counts, and Alexa rankings.
- staunch 17y agoThe problem is that 37 signals qualified the statement. They're claiming there are 3 million "people who use our web-based apps". That pretty explicitly means active and current users of their products. If they said they had 3 million registered users that'd be an entirely different thing, and, ya'know: true.
- run4yourlives 17y agoThat pretty explicitly means active and current users of their products It does? Says who? This is your definition of a user. That's fine, but it doesn't extend to being a common one. Unless of course you delete all the accounts of people that have registered for your service but aren't using it as often as you'd like.
- staunch 17y agoThe question is not about what the _lone word_ "user" means. That's obviously not specific enough to have a single definition. The question is what does "3 million people who use our web-based apps" mean. A simple truth test: Are there "3 million people who use [their] web-based apps"? Absolutely not. Not under even the most remotely reasonable definition.
- jarin 17y agoOf course you are assuming that there are not 3 million active and current users of their products. If you consider that many people have multiple user accounts on 37signals products (and that they count free but active users too) it is not unreasonable.