I use their product and haven't visited the site in months. I just look at the weekly summaries to get a decent picture of how productive my week was (hours in work opposed to waste of time).

Traffic to their site is not nearly as useful as their signups growth is for statistics.

Fair point. Might be worth having the application show something from the website, so that website stats are more representative, or at least look more impressive :)

You are starting to sound like a lot of the big web portals out there.

'Let's break an article into 10 pages so we get 10 times the hits!'

'Let's turn that AJAX gallery that loads really quick into separate pages so we get way better traffic scores!'

In jest to you - but they really have those types of meetings.

Again, fair point :) As long as the subscriber numbers are good, that's cool. I'm used to looking at websites rather than desktop software so it's a whole different game.

Also depends on their strategy. I'd expect the summary to display some recommendations from a website, maybe some recommended software (adverts), etc Maybe some social networking - "Here's some people who have similar usage patterns to you. Go chat."

QuantCast is horribly unreliable, far worse than even their competitors who are themselves terribly unreliable.

Unless you're quantified of course. I have to say I find Alexa to be pretty accurate these days though. It's got lag, but it seems to get sites in the right sort of ballpark.

do you seriously think any of the public stats sites are able to understand traffic of small sites like ours? Not so much, as it happens. And this fact is pretty well documented (seomoz has some good posts documenting how comscore, alexa, and quantcast don't do so well with smaller sites/services). I'm in NYC on an iPhone- I'll find the posts next time I'm in front of a real screen.

The 7 percent number (we were up to 9 for a while) is user growth-- not traffic. Both are still growing nicely.