First there was George Carlin alerting us to the seven words you couldn’t say on TV. Now Coke blocks 3,415 words from user-generated content.

Do I hear you asking, “What’s user-generated content”? Here’s how it works, in this case:

An accessory for Coke’s newly launched “Taste the Feeling” global ad campaign, GIF the Feeling is a website that allows visitors to fashion Coke ads by combining a stock clip with user-provided tagline. The result can be posted to social media or downloaded as a GIF—that au courant looping animation image format.

What would you say in a Coke ad? Would you describe a positive emotion? Or create digital graffiti? Would you try to use Coke to promote your political cause? Or would you try to subvert the product? Given human nature, more than one person would try to troll, deface, subvert, coarsen or impair the campaign. Hence, the “Profanity API” — a technology intended to ward off foreseeable misuses of the advertiser’s brand.

“Profanity” doesn’t capture the full range of the technology’s activity. It blocks words in many other categories:

• drugs and alcohol
• sex
• politics, race and religion
• brand names and proper names
• violence, crime, abuse and insults
• business and enterprise
• health, medicine and chemistry
• food and drink
• soft drink terms
• negative connotations
• miscellaneous words

For a thought-provoking review of what is blocked and why a company would subject its brand to so many potential indignities, check out Ian Bogost’s fine article in The Atlantic.

Incidentally, I’ve seen at first hand what can happen when a “profanity API” is not in place. Years ago, I was part of a team that provided services for a telecom manufacturer. We had been invited to discuss integrating some of our services with a business card ordering system the manufacturer had purchased. Our host beamed as he showed us the easy-order feature which eliminated the need for HR staff to approve the content of the card. His face darkened — from a mix of embarrassment and anger — when he opened the dashboard to show us the cards that had been requested. In more than 70 instances, employees had entered the CEO’s name and assigned him a derogatory job title.

More recently, I worked on a project for a cell phone reseller which aimed a product at the teen and young adult market. Customers were invited to manage their account online and leave brief testimonials about the product. You can guess what happened. When we received the data file for a mailing, we discovered that offensive or derogatory comments littered not only the comments field but also the address fields in a high percentage of the records.

User-generated content is a rough neighborhood. Hats off to Coca-Cola and other brands who venture out, trusting in their “profanity API”.