Fighting the flack: how keyword analysis can help beat a crisis

October 19 2016
Published in Tips and Guides

In spite of the old “sticks and stones” idiom, words can do a lot of damage to brands online. One angry Facebook post or tweet can escalate rapidly into a social media storm, leaving a permanent stain on a business’s image. Which is why keyword analysis can prove to be a crucial tool for online reputation management. Tracking which keywords and expressions are being used in conversations about a brand can be crucial to spotting the first moments of a crisis, before it has time to snowball.

Here are a few of Visibrain’s top tips on how tracking keywords can help with your day-to-day brand reputation management.

Know what you’re looking for…

Having an unbeatable knowledge of your brand and all of its issues, past and present, is key. Based on any crises your brand may have experienced in the past as well as any current PR problems you may be having, compile a list of words and expressions that you NEVER want to see in a social media post or news story about your brand, and monitor them closely. A good media monitoring tool will feature an alert system that will notify you if any of your chosen keywords come up in a conversation about your brand.

For example, after its severe food poisoning scandal this year, Chipotle should be closely monitoring any mention of E-Coli.

But expect the unexpected

Even with the best planning, it’s difficult to predict every single word that could spell trouble for your brand. Some crises simply arise out of the blue, using words and expressions that are not part of your usual monitoring lexicon.

Microsoft’s chat bot disaster is a prime example of this kind of problem. In case you missed it, Microsoft experienced some major embarrassment earlier this year after Tay - an experimental Twitter chat bot designed to converse with human users and learn from their responses – starting spewing racist and inappropriate comments.

Semantic analysis of Tay’s first moments Twitter showed that words such as “racism”, “rogue” and “sexist” all started to appear in conversations about Tay just hours after launch, a worrying weak signal that could have alerted the brand to issues with the project before they had time to escalate to the extent that they did.

So how can this kind of situation be avoided? As well as setting up alerts for your expected problem keywords, be sure to check your semantic content regularly, especially during key strategic periods for your brand, such as the launch of a new campaign or product. To avoid having to constantly check your monitoring tool, set up hourly or daily reports to ensure you’re not missing anything important. Not only will this help you pick up on negative expressions quickly, but it can also provide invaluable customer insights.

If you want to know more about the different ways of using data to prevent a brand reputation crisis, we’ve written a comprehensive guide on how it’s done. Download it below, and let us know what you think!

Download the full crisis prevention guide

Simply enter your email below to start the download.

This field is not valid

Thank you! Your download will begin shortyly...

Published in Tips and Guides