It’s not news that spam plagues social media, but Networked Insights (via Marketing Land) recently put out some research looking into how it affects brands, and the conclusion is that data is pretty “dirty”.

According to the report, 9% of all Twitter users who tweet in English are non-consumers. That’s not just spammers, but includes bots, celebrities, brand handles, and inactive accounts. Spam is a bigger problem on blogs and forums, where 28% of all posts are written by non-consumers, it finds. Still, over 90% of social media posts in conversations around some brands are classified as spam.

Brands today are becoming more sophisticated about analyzing social media to discover consumer insights. But, if social data is filled with spam, brands’ analyses will not be accurate or actionable,” the report says. “The industry recognizes data is inherently dirty. According to a recent New York Times article, 50% to 80% of a data scientist’s time now involves cleaning data. Efforts are being made to clean up social spam, but it’s an uphill battle. Twitter is actively finding and removing social bots, But, according to the Wall Street Journal, the black market is able to create fake accounts quicker than Twitter can remove them.”

The report defines social spam as including tweets, reviews, or blog comments, which contain coupons, product listings, contests/giveaways, adult content, or general spam, which may contain “gibberish or nonsense”.

Older forms of social media like forums and blogs are still the heaviest sources of social spam, based on the report’s findings.

Different categories have different percentages of spam posts, with the most appearing for the shopping (13%), finance (11%), and technology (10%) categories compared to real estate (7%), automotive (6%), pets (6%), sports (1%), science (1%), and religion (1%).

“There are branded posts and conversations about brand names happening across social media all the time,” the report says. “Although overall spam rates are less than 10% across social media platforms, conversations for some brands are dominated by spam.”

Elizabeth Arden and Rite Aid are among the brands being dominated by spam, each with 95% of posts about them falling under the label. Other troubled brands include Visa (81%), MasterCard (76%), Ulta (75%), American Eagle Outfitters (73%), Best Buy (72%), Fisher-Price (70%), Lancome (68%), and Victorinox (68%).

The main takeaway from the report is that when you detect and remove all the spam from brand conversations, the conversations look very different. For example, when they removed the spam from the food and drink category, they found “more nuanced conversations” about healthy food, food cravings, etc. This would enable brands to better interpret what consumers are really talking about.

Once you’re able to eliminate (or at least cut drastically) the spam noise out of the data, you can focus on legitimate data, which you can then use to personalize your marketing, which is key to increasing brand loyalty.

Images via Networked Insights