Measuring social media campaign effectiveness – a statistical method

on Apr 09 in google, hints and tips, social media, statistics, tools posted by

Social media marketers are under pressure to account for their budgets. Accounting departments like numbers. But social media impact can be hard to measure. There are hundreds of metrics to choose from, such as number of comments, registrations, etc. Then there are dozens of online applications that will measure some or other aspect of your social media presence. It’s hard to choose from all the possible tools.

Where to start?
Here I describe a simple, robust method, that costs nothing and will provide an easy to understand result.

The method is based on using Google Alerts and measuring the change in their number over time. The assumption is that the number of alerts correlates positively with the overall social media impact or “buzz.”

Here I describe the process for a campaign as it might appear for a mythical client, Widget Tools, for their campaign “Best Widget Ever”

  • Set up a Google Alert with the key words that are most appropriate to the campaign. The alert needs to specific to the name of the campaign, such as “best widget ever.”

  • You can choose “Comprehensive” for all news items. For a better gauge of social media impact you can set it up just to monitor blogs.
  • Set up the alert to notify you daily. For a high-intensity campaign, you can use “as it happens” but once a day is fine for most campaigns.
  • Be sure to enclose your search term in quote marks, otherwise you will get pages that have all the search terms anywhere, rather than the search phrase specific to your campaign.
  • Set up a spreadsheet with a date column and another for the number of alerts.
  • Graph the number of alerts per day over time. This is called a time series. 

  • Calculate the average number of alerts per day before the campaign. This will serve as a control if your search terms result in false positives, and give you a baseline to assess the impact of the campaign.
  • Calculate the average number of alerts per day after the campaign. This serves as the measurable effect of your campaign.
  • Draw a straight line equal to the average before across the graph, and another line equal to the average after the campaign launch, as shown in the figure.

You can see from the graph that the average after the campaign is higher than the average after the launch. The difference between these two number is the measured impact of the campaign, in alerts per day.

Statistical test
If the difference is not obvious, or if your accounting department needs convincing, you may need to conduct a statistical test. Such a test is not infallible, but it will provide confidence in your results.

Measuring the number of Google Alerts does not by itself, provide evidence of positive ROI. Your accountant still may not be happy. However, it does provide incontrovertible evidence of how a campaign impacts social media.

No Comments Leave a comment

No comments yet.

RSS feed for comments on this post. TrackBack URL

Leave a comment