Tortoise or the Hare: Social Media Sampling
Posted by Vovici Blog on Wed, Mar 09, 2011
At yesterday’s NewMR Listening event, Annie Pettit, Chief Research Officer of Conversition, contrasted the tortoise and the hare approaches to social media research across four methodological steps.
- Twitter searches vs. comprehensive searches. The hare will just hit http://search.twitter.com and find comments there, forgetting that only 7-8% of online users utilize Twitter. The tortoise will conduct a search that cuts across social media, including blogs, forums, media-sharing sites, Q&A sites, news sites and more. The comprehensive strategy is necessary because sentiment and attitudes vary by type of social-media channel. The top-two-box sentiment for Walmart, for instance, ranges from a low of 25% on video-sharing sites to a high of 38% on Q&A sites.
- Keyword searches vs. brand searches:
- Errors of inclusion. The hare just searches social media for a keyword and analyzes all the returned results. The tortoise crafts a query that will exclude stock words and phrases. For instance, a search on “Target” will return references to “above target”, “likely target”, “target market”, etc. What’s the difference? Well, the top-two-box sentiment for the brand “Target” is 48%, compared to 32% for the keyword “target”.
- Errors of exclusion. The tortoise will take it a step further and include brand synonyms: “McD”, “Mickey D”, “MickeyDee” and “Golden Arches” would be part of a McDonalds brand search, for instance; each has different sentiment scores, reflecting different attitudes about the brand.
- Keyword searches vs. constructs. The hare wants to see the sentiment surrounding its new brown version of its product and does a keyword search. The tortoise will use constructs that weed out false matches: comments about Jackson Brown, Mae Brown, Charlie Brown, Chris Brown, Joe Brown, Les Brown, etc. (Charlie Brown has a top-two-box score of 43%, compared to 21% for Les Brown.)
Annie has shared her slides as well:
To extend her analogy, I do fear that the hare is going to win some early social-media research races. Only then, as consumers of such research dive in and discover the flaws, will they be willing to embrace the slow-and-steady pace of the tortoise. As Annie said, “Quality is the most important part of the work: without quality there is no validity and no ability to generalize from the findings.” Here’s to the social-media tortoises of the world.
See also: