Social Monitoring. Garbage In. Pretty Garbage Out.
Saturday, September 5, 2009 at 09:31AM ![]()
If you've done a bit of of work in the digital marketing arena lately, then odds are you're familiar with the various social monitoring tools available to help bring metrics and clarity to a chaotic stream of data. The number of options is impressive, but realize that the majority of them also share the same weakness.
The reality is that we're in a f a "garbage in, pretty garbage out" situation. While the tools are becoming increasingly detailed and complex in terms of chopping up data, visualizing trends, and even gauging sentiment, the majority of them still are populated by the same simple back-end data: RSS.
Challenges with data coming from RSS feeds:
1. Lots of unfiltered spam. Although tools are being implemented to block this out, certain categories like pharma and e-commerce are bombarded with keyword spam.
2. Repetition. In the age of the the retweet and digg link there is a vast amount of RSS data that is redundant. While it's important to know how much a message is being rebroadcast, by simply taking a feed vs. grouping items as original vs. repeat monitoring tools can often create a false sense of importance around a conversation.
3. Text Heavy. Text is core to how a RSS works. While social media channels are primarily text based, there are some big players like YouTube and Flickr that have the potential to distribute viral content in a visual manner which monitoring tools often cannot see. Obviously tagging exists, but this assumes that the someone took the time to implement tags (which often isn't the case.)
The only current solution to the above issues is implementing human power. Unfortunate most of us simply don't have the time or resources to delegate teams to do this kind of work. There has not been enough monetary justification to do anything above and beyond spot-checking the work of robots and algorithms. The point here is, take the output of social monitoring tools with a grain of salt. Understand the specific business needs you're trying to address before pulling piles of data from these tools.
If you're looking to expand on your ability to respond rapidly during a crisis, are you measuring the time and volume of your press releases? Are you correlating the growth of your following to the subsequent volume of rebroadcasts?
If you're launching a discussion forum are you tracking overlaps between the outside social web and your domain? Shortened URLS leading back to the forum?
Social Monitoring is an extremely young enterprise. It has its pitfalls but also has amazing opportunities. As brands are jumping into the social web (and naturally wanting the metrics to justify that decision) they need to remember that data streaming in is going to evolve and setting baselines and expectations now should be handled carefully. We're going to see monitoring start to focus more on quality not quantity as filtering improves and algorithms move closer to mirroring human analysis. Until (if) that happens, be careful when investing a great deal of money, time, and trust into clumps of data that have been reformatted into a somewhat more organized and visually appealing package.
Brands,
Feeds,
Metics,
Monitoring,
Reporting,
Social Media,
rss in
Measurement 














