Interview: Carnegie Mellon Professor Ari Lightman On How Students Are Empowered By Learning To Use Brandwatch Consumer Research
By Kara FinnertyJun 10
Published March 19th 2019
As we’ve already written, fake news and its detection goes beyond the realms of technological solutions and can often lead us down a philosophical rabbit hole.
But, practically speaking, if falsehoods are being spread about your brand you’ll want to know about it fast.
While a silver-bullet solution might never exist, what steps can brands take now to ensure they can tackle misinformation at its root?
Whatever people are saying about your brand, you should be tracking it.
That said, especially when your brand receives thousands of online mentions a day, it’s important to filter the data so that urgent matters are dealt with quickly. If something negative is brewing, the sooner a brand is able to act, the better.
There are a few ways you can go about surfacing negative mentions, as well as ways to find potential sources of fake news around your brand.
The most basic way to track negative mentions is to break conversations down using automatic sentiment analysis. You’ll quickly be able to see the positive, negative and neutral.
Spikes in negative conversation are obviously bad, while spikes in positive conversation are obviously good. AI assistant Iris will be able to tell you what’s driving particular spikes, and you can click around within the spikes to see what topics are driving the sentiment.
Of course, using customizable filters you’ll be able to get deeper than that. Different industries will have different words associated with negativity (in the gaming industry, “sick” is often a positive adjective), and automatic sentiment analysis won’t always recognize this. Therefore, customizing rules around sentiment can help ensure that mentions are properly categorized as positive, negative or neutral. “The bomb” used to discuss a new music album is a high complement. “The bomb” discussed in relation to a train station is a huge red flag.
So far we’ve not discussed fake news – instead, we’ve discussed capturing negative mentions. In my opinion, whether a disparaging story is true or not, it warrants investigation and action. That said, as we move into discussing sources we’ll begin to start to target sources of misinformation that can be flagged.
I think source segmentation is one of the most powerful ways to surface potential misinformation spreading about your brand. This takes research and ongoing checks, but it’ll allow you to flag mentions coming from particular authors or sites that have a history of spreading fake news.
This is a messy area – deeming a site or author as a purveyor of fake news is quite the accusation and I’m not going to suggest any sites or names here. That said, a seasoned social media manager who’s experienced fake news being spread in the past will easily be able to find the most vocal proponents of a fake story and the sites that spread them. Using rules in Brandwatch Analytics, analysts can easily segment conversation about their brand by author name, specific links, and sites. In this way, conversation around your brand that’s coming from previously problematic sources can be flagged and quickly reviewed before appropriate action is taken.
Many brands will be aware of common misconceptions around their products or services. This can often be pretty harmless, but even so it’s probably worth keeping an eye on.
Let’s say you’re aware of a particular pain point that customers are often complaining about – perhaps a particular taxi firm only allows passengers to pay in cash. Once this policy changes and cab drivers allow card payment, not everyone will hear the news immediately. By tracking any conversation relating to the misconception that the taxi firm doesn’t allow card payment, social managers can engage with those discussing it and spread the good news.
A more serious example might be misconceptions around harmful ingredients in food products or policies around staff behavior. Setting up rules to segment conversation relating to misconceptions can allow for the social team to engage with and correct any untruths that are being shared.
When negativity or fake news buds, it’s important to be aware of it before it blooms.
Using customized Alerts or Signals within Brandwatch, all relevant stakeholders can be instantly alerted to potential problems 24/7 without anyone being logged into the platform.
Alerts can be set up so that potentially problematic mentions are immediately sent to the group’s inboxes. These can be customized to pick up pretty much anything – verified Twitter users sharing your brand name along with a particular phrase, any shares of a link to a popular false news story, and more.
Meanwhile, Signals will help alert teams to changes in the data that aren’t anticipated with custom rules. Perhaps a story relating to your brand is gaining popularity, or sentiment around your brand is rapidly declining. Brandwatch will send you the signal along with context around what’s happening.
These notifications will be sent via email to all relevant stakeholders, so a plan of action can be quickly formed.
Brandwatch can’t tell you whether something is true or false, but we can tell you if a particular topic or type of sentiment is rising in volume and alert your team so that action is taken swiftly.
Knowing about potential crises early means social teams can communicate with the relevant departments to determine whether a story is true and create a plan accordingly.
Like I said above, whether conversations are “fake news” or not isn’t necessarily important. Problematic stories need handling as soon as possible.