Why We’re Changing Our Twitter Potential Impressions Calculation
Our goal at Simply Measured is to consistently improve our reports to give you a complete picture of your brand’s digital presence. This means pushing ourselves to be thoughtful about how we calculate and present every one of the 100+ metrics that we deliver. In this pursuit, we recently found an inconsistency in the way Twitter potential impressions were calculated. We’d like to share with you exactly what that inconsistency was, the mistake we made, how we’ve resolved it, and how we’ve used this to create a more flexible experience around the way you report Twitter potential impressions.
We provide 10 different reports that rely on Twitter data. Several months ago, based on conversations we regularly have with customers, we improved the way impressions were calculated. At that point, we began counting all @replies as one impression, rather than counting an impression for every follower of @reply authors. Talking with our customers further, it was clear that this was a more accurate method because @replies are only shown in timelines of people following both the sender and the user replied to. This weekend, however we realized that in three of our reports, there were constraints that had us counting all followers within replies. This led to impressions being overstated against our desired method in the Multiple Twitter Channel Analysis, Stream Snapshot, and Klout Stream and Klout Influencers and Advocates reports.
Measuring impressions is a fundamental challenge in this business. We quickly realized that there wasn’t a simple “catch-all” way to incorporate this metric. Potential impressions provide an important benchmark for your business, and help you place an overall value on your branding efforts. That’s why impressions are important in traditional and TV advertising, and that’s why they’re important in the digital space.
We’ve heard from some customers who wanted to count impressions on @replies more liberally, making more agressive assumptions about the amount of overlap between users. We wanted to address these varied approaches in addition to solving the inconsistency mentioned above.
The Solution: Let You Decide
To make sure that we’re providing you with the type of metric that is most meaningful to your business, we decided to give you a range of options for calculating potential impressions across all Twitter reports. From now on, when you use any Simply Measured report to calculate Twitter potential impressions, you’ll have three different options to choose from. Below, you’ll find a breakdown of all three options, and exactly what they can mean for your measurement efforts.
3 Methods For Measuring Potential Impressions on Twitter
Method 1: This will be our default method. In this more “conservative” calculation, one @reply results in one potential impression. This will likely give you a lower than actual reading, but makes a fair assumption that @replies won’t generate the full reach of the author’s followers. In addition, @replies are typically 1-to-1 interactions where reach/impressions are not as relevant, rather than broader messages. We’ve focused on this approach as the go-to model, and the majority of the customers we consulted support this method.
Method 2: This is the most aggressive method, accounting for each follower of the author as a potential impression. This approach takes the word “potential” at its most literal form, which is important to some of the brands we’ve spoken to. To choose the option in the metric setting, select “As 100% of followers” from the drop-down menu in the report settings.
Method 3: We wanted to allow for customization here. This method will let you define any percentage contribution for followers of @replies from 90% to 10%. For example, if you believe that 20% of the average author’s audience is overlapping with the account replied to, you can use this as a basis for potential impressions across all @replies on all Twitter reports.
If you’re a current customer, this function is available on all previous reports. Simply regenerate after you select your preferred method in report settings. If you have any questions about what this means, how to use the function, or would like to better understand it, reach out to your account manager, or email firstname.lastname@example.org and we’ll make every effort to help you fine tune the way you measure and report on potential impression.
As the Head of Marketing Communications at Simply Measured and generally delightful person, my job is to use data to tell stories to the internet that help the internet get better at telling stories...You're welcome internet.