Widget Analytics – Measuring the widgets in the wild

Helping web analysts navigate the measurement and tracking of widgets.

Archive for March, 2008

Widgets versus Applications – what’s the difference?

Posted by widgetgirl on March 24, 2008

Clearspring_Facebook

What is the difference between a widget versus a Facebook application – and now a MySpace application? This question comes up a lot when working with our customer base and other peers in the industry. Widgets (and I am going to focus on web widgets here) are stand alone applications that are sharable and can run on any web page as long as you have an internet connection and a browser (I am sure someone will debate me on this – and in the Social Media Standards committee we are debating these same topics). The main thing to note here is that a widget can stand on its own inside or outside of Facebook or MySpace, but there are benefits to also tapping into what these platforms can offer as they can make your widget much more viral in how they are shared and spread in those environments. For this post, I am going to focus on Facebook as the newly launched MySpace apps are still maturing (and I spend a lot more time on Facebook than MySpace).

A Facebook application is an application that is registered on Facebook and can take advantage of all of the FB features that their platform supports (and keep it mind that it does not have to be a widget). For example, FB Apps have the following features:

  1. Left navigation inclusion – a persistent link on the user’s navigation bar (this is on a user’s profile page) .
  2. Canvas page – allows the user to see your content on a separate page from their profile in a larger canvas area.
  3. Profile – the application is added directly on to the profile page for display and usage.
  4. Friend Requests – at the end of a successful application add, most applications (if configured to do so), will redirect you to an “invite your friends” page. Note: be wary of those that pre-select all of your friends – or those applications where you have to share with your friends to see the results of some quiz or poll. My friend Rich has a great story (I was a personal witness) of accidentally hitting “see results” for a quiz (which was actually a Trojan horse version of an “Invite Friends” button).
  5. News Feed and Mini Feed – notifications are sent to your friends news feed and mini feed upon adding and removing an application. What the heck is the news or mini feed you may ask? These are both “feeds” into your profile page or “Facebook Home page” that tell you what your friends are up to. If you or your friends add applications, this is where the notification will appear.

So if you are a frequent Facebook user (like myself), then you have probably stumbled across all of these – even if you didn’t know it. The interesting aspect of working with FB Apps is that there are a lot of variances to understand with respect to widget analytics. There are a few different things to be aware of:

  1. FB has two different ways that a widget can be shared within FB – 1) “Add Application” from the application’s canvas page and 2) natively grabbing the widget from another user’s profile page. The former example is how the widget (or app – keep in mind that I am assuming that the app IS a widget) can be spread without using the sharing tools provided by the widget serving platform. The latter is completely dependent upon using the services of the widget serving platform. The result here is that the widget placement that gets the credit for spreading the widget is either going to be a centralized registered application – OR – a specific user’s profile. The granularity between the two options can be critical if you as a marketer are trying to identify a power user who is causing your widget to get grabbed a lot.
  2. Click to activate – just because your widget is being viewed each time a visitor views their profile page, does not mean that your widget analytics data is incrementing in lock step.  FB requires that the visitor “click to activate” the widget before it will load in the content and make outbound connections to send tracking data. Other social networks do not behave this way, so it is a critical nuance to be aware of when analyzing your widget metrics for Facebook.

Not all social networks are alike. In fact, you may find that the analytics reported for views of your widget will FAR EXCEED that of another network – yet the same two networks will show opposite trends for how often and how many times your widget is grabbed or spread.  In many cases, this is dependent on tapping into the “social graph” or lack of social graph of many of these networks.  Those that easily hook into your “friends” and invite them to join the same application (or widget) will generally behave different than just waiting for someone to grab your widget on a static page. Pros and cons to each.  The shrewd marketer will combine these efforts to build a marketing campaign that works effectively across all metrics – views, grabs and interactivity.

What are you noticing that is different between social networks?  Do you see behavior that is completely different from one site to another?  Segmentation is the key in understanding how to measure success in your widget campaign.  Different metrics will apply for different networks depending on their behavior.

Check out my favorite widget of the week – Clear Channel launched a new Jonas Brothers widget that promotes the band’s songs, including a new release. Enjoy!

Advertisements

Posted in widget analytics | Tagged: , | 2 Comments »

eMetrics Summit Toronto

Posted by widgetgirl on March 21, 2008

Toronto

On Monday March 31st I will be participating on a panel at the E-Metrics Summit in Toronto, Canada to discuss the differences in web analytics versus audience measurement tools. The panel description is as follows:

“Are you confused about the number of customers visiting your website? Are the metrics reported by your web analytics tool different from the metrics reported by your online media, or by audience measurement organizations? The WAA invites eMetrics Summit attendees and the local GTA (Greater Toronto Area) business community of web marketers, publishers and agencies to attend this community meeting.

A panel of experts will discuss the value of the metrics, methods and tools used by web analytics practitioners, online advertising media and audience measurement organizations. Find out how-to use these metrics and tools to better understand your customers, your website’s competitive standing and overall website value”.

Sound interesting? What will be even more interesting is the “who” that is on the panel.  In addition to myself, the panelists include:

The group of us have had some very interesting “pre-panel” discussions to prepare and a pretty good understanding of the issues that many web analytics practitioners (and widget analytics practitioners 🙂 ) are facing. So if you are planning to come to this conference (or you are a local Toronto techie), please join us at 4:30pm on Monday March 31st for the WAA member meeting and panel discussion.

And of course, please find me to say hello. Let me know what your organization is doing with widgets and what your analytics challenges are.  I’d love to chat!

Posted in widget analytics | Tagged: , , | Leave a Comment »

Engagement with a widget

Posted by widgetgirl on March 13, 2008

Engagement

This weekend I am hosting an engagement party for two of my close friends. The constant use of the word “engagement” in my conversations during the day (my job) and my conversations during the evening (the party) have had me thinking a lot about this word. Engagement is such the buzz word in web analytics these days. Everyone is either claiming to have the magic formula or debating it’s definition and use. I don’t think that there is a definition for engagement. I believe that it is an umbrella for a set of metrics or tools to analyze how visitors interact with your widget (replace widget with any content type you want 🙂 ).

When I say “umbrella”, I mean the metrics to evaluate engagement are going to vary widely depending on what you are analyzing. Metrics that I would use to evaluate engagement are specific to the content and content type. For example:

  • Video – clicks on “play”, “rewind” or “pause”, percent of stream played, new versus repeat plays.
  • Photo Gallery – number of photos viewed; number of clicks on “Next” or “Back” in the gallery; new versus repeat visitors to the gallery; time spent viewing the gallery.
  • Game – number of games played; games played per visitor; clicks per game.

From a widget analytics perspective, I would of course always consider the grab rate (grabs of the widget divided by views), number of times the widget has been shared and whether or not those new placements of the widget are staying active or churning off. You can read my post on “Outbreak” to get a better understanding of what I mean by active or churning.

Specifically for widget analytics, at Clearspring we have focused on a specific set of metrics to measure how widget viewers are engaging with our customer’s widgets. They break down as follows:

  • Interacted Views – views to the widget where a visitor either moused over or clicked on the widget.
  • Interaction Rate – interacted views divided by views.
  • Clicked Views – views to the widget where a visitor clicked on the widget.
  • Click Rate – clicked views divided by views.
  • Clicks – no explanation needed I hope.
  • Clicks per Clicked View (CPCV) – clicks divided by clicked views (think page views per session if you are a web analytics junky).
  • Avg. Time Spent – the average time spent viewing the widget (time on page).
  • Avg. Interaction Time – the average time spent physically interacting with the widget on a page.

Adding on to the theme of interaction metrics above, the ability to capture custom events is critical in completing the engagement umbrella toolbox. Who cares if someone clicked 5 times on a widget? But if they clicked 2 times on play and 3 times on “send this content to a friend”, then the value of those 5 clicks takes on a whole new meaning.

None of these metrics in isolation are going to answer the question of “what is your customer’s level of engagement?”. The metrics are simply a set of tools provided to help the widget creator analyze their content that is specific to their business goals and objectives. Every business is different and every widget is different. Depending on the objectives of your widget and its content, your definition of engagement will vary drastically from the next widget that comes along.

Staying with the theme of “Engagement”, my favorite widget this week (which I am listening to via headphones as I type) is Clear Channel’s new widget that features Sara Bareilles. The widget has several songs that the site visitor can listen to (she is singing “Love Song” right now – how engagement appropo). Check out this widget and look for two specific things: very cool and prominent “Get and Share” button on the widget AND sharing buttons specific to each radio station’s, private social network. You can see that by clicking on the “Get and Share” button and toggling through the icons to see the radio station that is promoted.

Hats off to Clear Channel – the quality of the stream is amazing!  Cheers!

Sarah Bareilles

Posted in widget analytics | Tagged: , | Leave a Comment »