Social monitor Brandwatch adds the ability to detect ‘something’s not right’

You know that feeling you sometimes have, when something’s not right?

Tomorrow, Brandwatch is launching that as a feature in its social intelligence platform.

Called Signals, the new capability looks for “something that is happening that’s different from normal,” global chief marketing officer Will McInnes told me.

It could be that new hashtags are showing up in discussions about a brand’s new product, or maybe different trends, subjects, or issues are emerging. Perhaps the discussion in question has jumped to other social networks.

Like the indescribable feeling that a storm is coming on, the unusual activity could signal a crisis in the making. Brandwatch says Signals can detect an early warning for an incipient customer service problem, a security threat in progress, or negative stories gaining steam.

Signals emails such alerts often, with the idea being that, if something is brewing in the social ocean, you need to know about it right away.

Even Brandwatch can’t predict what unusualness Signals will find.

That’s because the new feature utilizes the increasingly popular machine learning technology, which is designed to get smarter and more accurate from results.

But, like an experienced card player, it doesn’t always know why it senses the things it does.

Social monitoring platforms — Sprout Social and Mention.com, just to take a couple of examples — commonly allow you to set up alerts for search terms, and they often follow the thread of the discussion.

Signals, McInnes said, “alerts you to things you didn’t know you need to know.”

“No one else is doing this,” he claimed.

VentureBeat is studying social media marketing
Answer our survey now and we’ll share the results with you.

He pointed to an internal Brandwatch tester during beta, who was alerted by Signals to social buzz about a product that was unusual because it was taking place in another country than would normally be expected. No crisis has yet emerged from that, perhaps because it was a very early warning that was nipped in the bud.

Human conversation has its hidden sides, and social listening and intelligence platforms are now trying to differentiate themselves in their ability to parse and evaluate those mysteries.

Sentiment analysis, concept modeling, and now “something is happening here but we don’t know yet what it is.”

Social platforms have entered the age of ineffable tools, when you can see the signal but you can’t hear the sound.