Use Cases for Facebook Topic Data

Kester Ford
17th March 2015 0 Comments

In my last blog we looked at how PYLON works and how developers can use it to create insights from the billions of posts and engagements that take place every day on Facebook.

Getting insights from topic data
Now that you know the basics of how PYLON works with Facebook topic data, we’ll look at what you can do with it.

Your index will contain what your target audiences are saying on Facebook about your events, brands, subjects and activities. With this, the next task is to extract anonymized insights into that data.

The PYLON API allows you to extract summary counts of interactions, be they stories or engagements – not the actual posts themselves.  The summary results can then be delivered through tools that you develop to allow marketers to interpret the information.

The next step in understanding how PYLON works is to look at what we call analysis queries. These allow you to query the topic data stored in the index, and return aggregate results.

PYLON supports two kinds of analysis queries (time-series and histogram).

  • Time Series – Allows you to analyze the volume of interactions, or unique users who have created interactions over a period of time.
  • Frequency Distribution – Allows you to count the number of occurrences of each value of a data field, to return the top results.

Each can be executed against a subset of data as specified by optional analysis filters. Filtering data into smaller subsets is the key to getting the most insight from your recorded data. However, we apply an audience-size gate to all results, so that no result represents less than 100 unique authors.

When you are below the audience-size limit your results are redacted. To avoid that possibility you need to work with a larger audience and with a little practice you will soon be digging into your data and returning the summary results you are looking for.

When you first start to explore your results, it’s best to add conditions one at a time. For example first specify your time period, then add one condition to your filter, then add remaining conditions until you’ve dug as deep as you can.

Using query filters with your analysis queries allows you to further segment the data by filtering for a subset of information in the index. Multiple reports can then be generated from the same index data by changing the filter parameters.  Query filters can be based on attribute values or on VEDO classification that you have applied to the index. The start & end parameters allow you to specify a time period.

Remember the smaller the time series interval you choose, the less data you will have to analyze. When you start to explore your data set, it’s best to initially define a long time interval, a week for example versus an hour.  This will help ensure you meet the minimum audience threshold and get results to analyze.

What can I do with the insights I get from topic data?
The use cases for topic data are endless.  The only thing to remember is that personal information is stripped from the data:

For market research you can find out what people are saying about your brand, (brand health) as well as what they think about your competitors.  
Another obvious use case here is share of voice analysis, to surface how often your products are mentioned, against mentions about competing products.

Then there are a wide range of marketing optimization use cases where you can look at profiles of target audiences talking about a particular subject (audience research) and break them down further by analysis of their demographics.

Other good examples include customer experience, or measuring customer satisfaction, operational efficiency uses cases including demand forecasting around product or resource availability when launching a new product or service.  And also worth a mention are use cases like M&A research and product development.

This selection of use cases is only the tip of the iceberg of what is possible with Facebook topic data. If you are interested in gaining access to Facebook topic data within your organization, please get in touch and tell us more about your use case.

Kester Ford

Written by Kester Ford

Kester Ford is DataSift's Director of Product Marketing

Share This