A likeable Twitter bot?

One news story in particular caught my eye this week. Robert McMillan wrote an article for Wired about a man who created a Twitter bot that people actually like. Back in 2008 Greg Marra created @trackgirl as an experiment to see if she could infiltrate a network of real people; in particular, runners.

The Twitter script was designed to scour Twitter for running related tweets and copy and paste these as her own words. She also followed 5 running-related individuals per day and followed back anyone who followed her. This alone is considered good Twitter etiquette – and she wasn’t annoying anyone with porn-y content or faster than lightening responses and re-tweets.

What’s interesting about this particular bot is the reaction to one tweet about twisting her ankle. People started to DM her asking how the recovery was going and showing genuine concern. All studies in human empathy aside, this could be of genuine interest to brands.

According to Corp Comms, around one quarter of Italian and International Brands Twitter followers are actually Twitter bots, which can significantly increase a brands perceived social media standing. So what if brands were able to use Twitter bots to infiltrate their target audience and spread some subtle marketing messages?

We’re all pretty well versed in how to spot a Twitter bot, but @trackgirl was able to work around this by infiltrating one particular community. If she had been created by Nike, she could have spent time tweeting about her running endeavours, and then casually slipped in a tweet about her new Nike running shoes. Not only did the experiment prove that she could infiltrate running circles, just it also showed that people actually cared what she was tweeting about.

This raises a huge number of legal and ethical questions about the use of twitter bots to create social influence, but if the Twitter bot was using Tweets that were already in the Twitter-sphere, would the use of a bot to promote a brand really be that bad? On the one hand it would simply be a case of aggregating positive brand messages, and on the other it would be deception of a brands’ audience on a mass scale.

In the past we’ve seen the ill-effects of Twitter bots getting involved with politics as they were used to bolster and spread positive messages during the Mexican presidential election. While this technology could spell good news for brands, it would be one step closer to taking the ‘social’ out of ‘social media’.

Would you feel duped if you discovered a twitter user you were following was a bot? And how would you feel if they were spreading marketing messages?