How to Analyse a Facebook page

Understanding the impact that the content put out on your brand's Facebook page has on your audience is a critical component in objectively analyzing the performance of your Facebook efforts. Conducting the analysis, however, is often easier said than done.

Facebook Insights provides a nice interface which allows you to see a number of metrics related to each post, but there are quite a few other post level metrics that are beneficial to the analysis yet are not visible in the interface. Gaining access to the additional metrics can be done by either using the

Facebook Graph API or exporting the data from Facebook Insights.

Once you have the data, it's time to get organized.

There are many ways organize the data for analysis, but the focus here will be on three top level segments and four metric categories, each of which can also be segmented.

After the analysis, you will have insights into questions such as "what type of post or mix of posts should I utilize in order to achieve a specific objective?" and "when is the ideal time to post?"

Segments

The top level segments are the type of post, such as link, photo, video or status, the time of day that the content was posted, and the day of week that it was posted.

Each of these will provide insights on its own or can be combined with other segments to provide even more in-depth insights. This information is provided by Facebook for every post.

You can also refrain from segmenting at all and instead compare all posts against each other. This can often be a good starting point to get a high level view of your individual posts.

Metric Categories

The next level down from the segments gets into the metrics that can be used to analyze the performance of each segment. The metrics can be organized into four categories, each of which can be seen in the figure below.

1. Exposure: This one is pretty self-explanatory. It shows the number of impressions that were generated by your posts. This can be viewed as either total impressions, unique impressions (reach), or impressions by users who have liked your page. These metrics are useful on their own, but become even more useful when used as part of derived calculations, which will be touched on at the end.
2. Consumption: This one moves beyond exposure to having some sort of meaningful interaction with your post. This category is called consumption since it is an indication that the user went from being exposed to your post to consuming the content of the post. The sub-metrics of consumption include video plays, photo views, link clicks, and other clicks on the post. Similar to exposure, consumption can also be looked at in aggregate or segmented by the number of unique users who consumed the content and the number of consumptions that came from users who have liked the page.
3. Engagement: This is the step you hope users take after being exposed to the post and consuming the posts content. The sub-metrics here include likes of the content, comments about the content, and shares of the content. Engagement can also be viewed in aggregate or segmented by unique users who engaged and engagements that came from users who have liked the page.
4. Negative Feedback: This is one that you hope to minimize since it is an indication that the content you're releasing isn't relevant or viewed as spam. Examples of sub-metrics for this category include reporting the post as spam and clicking the X button, which hides the post from the users view.

All the metrics described above are easily accessible using the Facebook Insights export feature or for more advanced users via the Graph API.

You're probably familiar with some of these metrics just from using the Facebook Insights interface, but you need to bring the data out of Facebook Insights in order to perform the real analysis.

Once you have your data together in a spreadsheet, you can supplement it with your own derived calculations that will add value by providing greater insights into the performance of your content.

Just like the metric categories and sub-metrics described above, the derived metrics should also be segmented to give greater insights.

Some insightful metrics that can be used in aggregate or segmented include:

• Percent of page fans reached by each post
• Page fan engagement rate vs. non-page fan engagement rate
• Consumption rate of videos vs. consumption rate of photos
• Negative feedback rate for each post type

These are just a few basic examples of derived metrics. There are many others that can also be computed depending on the objectives of your Facebook page.

Donald Knuth - Father of Algorithms

Donald Knuth - the father of Algorithms at a reception for the Open Content Alliance, October 25, 2005

Donald Knuth is a renowned computer scientist and Professor Emeritus of the Art of Computer Programming at Stanford University.

Author of the seminal multi-volume work The Art of Computer Programming ("TAOCP"), Knuth has been called the "father" of the analysis of algorithms, contributing to the development of, and systematizing formal mathematical techniques for, the rigorous analysis of the computational complexity of algorithms, and in the process popularizing asymptotic notation.

In addition to fundamental contributions in several branches of theoretical computer science, Knuth is the creator of the TeX computer typesetting system, the related METAFONT font definition language and rendering system, and the Computer Modern family of typefaces.

A writer and scholar, Knuth created the WEB/CWEB computer programming systems designed to encourage and facilitate literate programming, and designed the MMIX instruction set architecture.

To read the full article click here

Probably a mathematical theory of surprise!

Computer scientists in California have built a mathematical theory of surprise, working from first principles of probability theory applied to a digital environment, according to this news release from the University of Southern California (USC).

And the results of experiments recording eye movements of volunteers watching video seem to confirm it. Beyond vision applications, this new Bayesian theory of surprise could lead to new developments in data mining, as it can in principle be applied to any type of data, including visual, auditory or text."

This new mathematical theory of surprise has been developed by Laurent Itti, of the USC’s Viterbi School of Engineering, his colleagues at his lab, and by Pierre Baldi, of the University of California Irvine’s Institute for Genomics and Bioinformatics.

Before looking at their theory, here are some key definitions given by the computer engineers.

"By analyzing streams of electronic data making up a video image,]
researchers can isolate stimuli with visual attributes that are unique in the
mix by breaking down the signal into "feature channels," each describing a
particular attribute (i.e,, color) in the mix. Such features are called
"salient."

"A parallel analysis performs similar operations, but does so
over time, not space, looking for new elements suddenly appearing. This approach
is said to model "novelty."

"Finally, an analysis can be done purely in
terms of Shannon’s original equations, which can measure the level of
organization or detail found in the data flow, its entropy."

If you want to take a closer look at the theory of surprise, then you will find the full article here.....