Vendors love to throw around fancy lingo in their branding and marketing. And we can forgive them some of their enthusiasm, after all they’ve got NEW! and IMPROVED! products to sell.
We provide our No Bullsh*t PR Measurement Glossary to guide you through some of the hype. It is a more accurate translation of the most commonly distorted or abused buzzwords. I’m sure there’s lots more jargon out there that you will encounter as you navigate your way to the right tool for your job. (You will want to check out our Vendor Selection Guide: The Best Vendors to Solve 24 Common Communications Measurement Problems in 2022.) If I’ve missed one that is plaguing you, please don’t hesitate to reach out.
The No Bullsh*t PR Measurement Glossary
Media Content Analysis– At the core of almost all earned media measurement today is a process known as media content analysis. It’s been used in academia for decades but started being used by corporations in the 1980s. (See The Princess and the Clips… The True Tale of the Birth of Modern Media Analysis for its origin story.) It became commercially viable with the arrival of the Internet and the digitization of media in the 1980s.
Automated Content Analysis– Once media was available in digital form content was easier to find. Humans no longer had to physically go through every page of every magazine and newspaper to find clips and read them. You could create a computer script that would search for specific companies and analyze digital media. This became known as automated content analysis.
Automated Sentiment Analysis– While humans could be (and still are) trained to accurately determine whether an article leaves a reader more likely to do business, invest in, donate to, or work for an organization, it is impractical to do so for huge brands in a reasonable time frame. So, in the late ’90s a query technique known as Boolean search was applied to media content. Boolean search could be used to analyze words, assess the proximity of the words to your brand, and assign a “sentiment” to content. And thus automated sentiment analysis was born.
Machine Learning– In the late 2000s people realized that you could teach a computer how to analyze an article for sentiment, based on human coding examples and other rules. Today, in most modern tools Boolean logic has been replaced by machine learning that more accurately identifies sentiment, topics, and issues.
AI-driven Media Analysis– Within the past few years machine learning has been replaced by artificial intelligence (AI) that automates the learning process and can easily be tweaked and taught to be more accurate in its definitions. As a result, more accurate variations on media analysis are now available. In an ideal world (that as yet has little to do with today’s reality) the AI learns from accurately coded media that has been audited.
Social Listening– Social listening is a fancy new term for “media analysis for social media.” A bot goes out to Facebook, Instagram, Reddit, Twitter, etc. and looks for mentions of your brand. It then uses some form of AI or machine learning to “listen” to the social conversations. It categorizes those conversations by what “issues,” “topics,” or “subjects” are being discussed.
Most of these bots assign sentiment to posts based on the words used. The accuracy of said sentiment is open to debate. The accuracy of everything else depends on the social listening tool’s ability to learn from its mistakes. If you are considering using a social listening tool that can’t be corrected and improved, don’t. Typically we find that they have learning capabilities appropriate to their age. Which is generally around three years old.
And if LinkedIn is important to you, good luck. No one that we know of (besides us) is scraping LinkedIn.
Social Monitoring (see Social Listening)– There are numerous tools out there that “monitor” social media chats for whatever you think might cause a crisis (e.g., bad customer experiences, bad or incorrect references to your brand, or inappropriate conversations your CEO may have had). They are all just doing social listening, but with a slightly different purpose.
Social Engagement– Social engagement is generally defined as some form of interaction with your social content. That interaction might be a robot programmed to like and follow anything that mentions your brand. Or it might be an influencer liking, commenting, and sharing your content. But the platforms can’t tell the difference between bots and people; they just treat all these actions as “engagement.” We strongly recommend you figure out what actions actually help your brand, and only count them.
Media Reach– Reach, we’ve learned, is in the mind of the beholder. It used to refer to the actual number of people who subscribed to a given publication. When social media came around, that got translated into some definition of “views,” either Monthly Unique Views, Weekly Unique Views, or—given that no one seems to follow the standards—whatever definition you or a vendor wants to put on it.
These new social media “reach” numbers too often start to head for the stratosphere. Which leads people like me to remind clients that, even if your platform says you reached “15 billion people,” you need to remember that there are only 7 billion people on the planet. And many of them don’t have electricity, never mind internet access or the ability to buy your product.
So now we use terms like “estimated reach,” and use domain authority and actual URLs (not the full domain URL) to calculate slightly more accurate numbers.
Media Impact– Media Impact is the latest fad in quantifying the reach and/or impact of a particular news story. We all know that impressions are mostly fictitious these days, so vendors and brands have been searching for a valid substitute. Of course there are no standards for defining impact. You can read about all the varieties and our recommended approach here.
Reputation Analysis– Analyzing a brand’s reputation generally involves a survey of key stakeholders to determine what they think of your brand (e.g., Edelman Trust surveys and RepTrak). However, the hot new moniker among the social set is “reputation analysis,” which uses social listening to determine your “reputation.”
If your entire stakeholder set indeed gets all of its information through social media, then this method might be very accurate. But if not, then “reputation analysis” is just rebranded social listening. It sounds so much more important (and expensive) than just “listening” or “monitoring.”
Reputation Management– Reputation management used to be what corporate communications professionals did. And, in fact, it still is. However, a number of social media monitoring companies have decided that they can help you “manage” your reputation by listening to and then intervening on your behalf on social media. It seldom works, just ask the University of California.
Crisis and/or Issues Management (see Reputation Management)– Again, this is what communications professionals are trained to do, and spend most of their time doing. But a number of social listening firms have decided to specialize in listening for a crisis or an issue that might erupt and make you spend your weekend in the corporate war room.
Don’t get me wrong, it is a very good idea to listen for crises on social media—in fact it might be considered corporate malfeasance if you don’t. And some platforms have been designed to assist you in deciding what to do about whatever crisis is about to erupt.
But to actually manage a crisis cannot be done by a bot, a machine, or AI. Managing a crisis, as any good PR person will tell you, requires the patience of a kindergarten teacher, the corporate knowledge of a company founder, the people skills of Oprah Winfrey, the negotiating abilities of Nelson Mandela, and the political savvy of Nancy Pelosi. A platform that includes those features has yet to be invented. ∞
Big thanks to Nick Fewings on Unsplash for the photo up top.
