If you are a gadget author, you’re probably interested in knowing how many users have your gadgets installed. There are a few ways to deduce this information, but none are perfect. I’m going to show you one way to generate useful information that is probably something you’re looking for.

Gadget Directory Details

The first approach is to just look at the iGoogle Gadget Directory entry for your gadget. It should show how many users you have. In fact, my Gadget Monitor gadget retrieves its information from here for user counts. Unfortunately, if you look at different internationalized versions of the directory, you’ll see different numbers based on how many of those users have it installed. There is no place to get a single consolidated number.

Analytics

A better approach is to use Google Analytics. In the gadget developer documentation there are instructions on how to use analytics with gadgets. But using it in the default way – recording one hit every time your gadget is loaded – is not all that useful because:

• Users may load your gadget many times in one day, and this fact isn’t really relevant like page views on a web site are
• If you record multiple hits per person using your gadget, you still don’t know how many distinct users are using your gadget
• Even looking at the “unique pageviews” data in Analytics can be misleading, because if the user has multiple sessions in the same day, it still gets counted more than once

New Goals

The approach I use takes advantage of Analytics, but in a different way. The goals are:

• Get as accurate a count as possible of how many unique users interacted with your gadget in a day
• Get a count of how many times your gadget was added by new users in a given day

This approach, unfortunately, will not get you any closer to knowing how many people have your gadget on their iGoogle page and didn’t hit it one day, or maybe have it on a tab that they didn’t look at one day.

A Different Approach

I’ll explain this new approach by example.

First, in the ModulePrefs declaration of your gadget, you need to require two features:

<Require feature="analytics"/>
<Require feature="setprefs" />

Second, you need a hidden user preference to store some data:

<UserPref name="today" datatype="hidden" defaultvalue="" />

Finally, some javascript code in your Content section:

// Get a reference to user preferences
var prefs = new _IG_Prefs();
// Get the user's current date/time
var now = new Date();
// Create a date string of the form yyyy-m-d
var today = (now.getFullYear?now.getFullYear():now.getYear())+'-'+(now.getMonth()+1)+'-'+now.getDate();
// Get the stored value of the "today" user preference.
var prefs_today = prefs.getString('today');

// If the "today" user preference is blank, then the user has never loaded
// this gadget. They must have just added it!
if (prefs_today=='') {
    _IG_Analytics("UA-0000000-0", "/addgadget/mygadget");
}
// If the generated date string doesn't match the one stored in the
// preferences, then this is the first time the user has loaded the
// gadget today. Record a hit and change the preferernce to be
// today's date string so it won't get recorded again.
if (prefs_today!=today) {
    prefs.set('today',today);
    _IG_Analytics("UA-0000000-0", "/gadget/mygadget");
}

(Note: in the above code, replace UA-0000000-0 with your Analytic account’s unique ID, and ‘mygadget’ with a unique identifier for your specific gadget)

The concept here is simple. Every time the gadget loads, it will get a string representing today’s date. If the date doesn’t match the value stores in the user’s preferences (for example, if the gadget was last loaded yesterday), then record a hit for today and update the preference.

This way, a single user will only record one hit a day to the Analytics counter. You get a pretty accurate count of how many unique users are using your gadget in a single day.

Since the preference gets updated every day the user loads the gadget, if it is blank that means this must be the first time it’s being loaded. That effectively means they have just added it. In that case, record a hit to a separate url starting with /addgadget to separate that from the other url.

An advantage of putting each hit in a “subdirectory” in analytics is so you can effectively use the drill-down report. By looking at the numbers for the entire /addgadget directory you can get a count of gadgets added in a day across all your different gadgets. Looking at the numbers for /gadget as a whole gives you a user count of all your gadgets combined.

I hope this approach proves to be as useful to you as it has been for me!