Studying large audiences with PYLON for Facebook topic data

Richard Caudle | 18th April 2016

Release 1.7.1 of PYLON for Facebook Topic Data introduced a number of new features. One key feature is the additional sampling options you now have at your disposal when looking to study large audiences.

Analyzing large audiences inside recording limits

As a PYLON customer you have a daily recording allowance which you cannot exceed. If you are looking to study a large audience, such as the audience engaging with a global brand, you can quickly hit your recording limits.

Sampling allows you to record a representative random sample of a large audience. You can analyze this audience to produce accurate results and at the same time stay within your account limits.

Read our best practice guide for an in-depth look at sampling.

Sampling an entire industry

For example, imagine you're looking to study everyone discussing and engaging with automotive topics. Perhaps you're looking to find out which demographic groups are posting stories about and engaging with well-known brands.

If you attempt to record everyone discussing cars you would quickly use up your recording allowance. You can instead record a representative audience using 'story-level sampling' using the interaction.sample target.

Filtering on the interaction.sample target allows you to capture a sample of stories and all the related engagements.

Blog---Using-sampling

Your interaction filter might look as follows:

(fb.topics.category in "Cars, Automotive" 
OR fb.parent.topics.category in "Cars, Automotive") 
AND interaction.sample <= 0.5

Here the interaction.sample target is used to capture 0.5 percent of stories that mention cars and all engagements on these stories.

Even though you've only captured a sample of the interactions you can still perform valid analysis. For example here's a typical age-gender breakdown analysis on both authors posting stories, and authors engaging with these stories:

brand-ag_0

Sampling for a brand

Let's look at another example. Imagine you're working with a global brand identifying stories which are receiving a lot of engagement and are therefore impacting the brand's reputation.

If you attempt to record every story and engagement relating to the brand you could exceed your recording limit, particularly if a story suddenly goes viral. In this case it is critical that you record every story, but you only need to record a proportion of engagements across these stories to identify which are receiving a large amount of engagement.

You can use the fb.sample target to do just this, capturing all stories and a sample of engagements across these stories.

engagement-sampling

Your interaction filter might look as follows:

fb.content contains_any "BMW" 
OR ( fb.parent.content contains_any "BMW" AND fb.sample <= 10 )

This filter will capture all stories that mention the brand and uses the fb.sample target to capture only 10 percent of the related engagements.

If you use this filter for your recording you will be able to analyze trending and viral content. For example the top domains of links that are being shared, based on the engagement they receive.

top-auto-domains

Learn more…

Sampling is a useful technique, but there are subtleties you need to consider.

Read our best practice guide for an in-depth look at sampling.


Previous post: Announcing PYLON 1.7.1 - Introducing Enhanced Sampling Support

Next post: Using Topic Graphs to Understand an Audience