By Katie Keating
Brands live in fear of the errant tweet, the insulting Facebook post, the lewd Instagram photo. And sometimes they are hacked. Sometimes a disgruntled employee does act out of spite. But all too often those social media gaffes come down to user error: a human juggling several channels – including their personal accounts – and hitting the wrong button, for the wrong site, at the wrong time.
A colleague of mine, software developer Aaron Quirk, and I submitted the patent “Preventing Messages From Being Sent Using Inappropriate Communication Accounts” to prevent just such embarrassments. Using natural language processing and the Bluemix cloud service, the invention monitors your activity across your email, social channels and more and “learns” your voice over time.
Social business managers, the folks like me who manage a company’s social media presence, typically use aggregation tools to see every site, channel, and app on one multi-columned screen. It’s convenient, but all too easy to hit “send” on one account, when you meant another. Our invention will know your tone and typical language for each site, and more importantly, from each account.
Meant to tweet about a referee’s poor eye sight on your own handle, and not your company’s? The tool will pop up and ask, “Are you sure? This looks personal.” You will then get the chance to stop or allow the post.
The tool runs as a background agent on laptops and mobile devices, monitoring your social media activity. The agent connects to an analytics service, for example, the Watson Analytics service running on IBM Bluemix, to understand usage trends on those connected social media accounts. When the cloud service detects unexpected activity, it alerts the agent so the user has a chance to recover from their mistake.
What secured this patent filing was its originality in the realm of social media tools and the value it offers to social media managers. I know social media and its pain points. Aaron knows software and how to take advantage of the cloud to scale the idea. Going forward, I would love to see this idea take shape as an IBM product so that everyone has a chance to avoid a social media faux pas.