Digging (Deeper) Into the Data: 3 Media Monitoring Pitfalls to Avoid
Last week, we couldn’t help but notice the commercials coming on like clockwork featuring sparkly diamond rings. Just like with the Christmas season, it seems the jewelry industry likes to encourage their target audience to pop the question on Valentine’s Day. But how many couples are actually getting engaged on Valentine’s Day? It seems like there would be no element of surprise on this day of all days in the year to get engaged, right?
What started off as a fun idea to monitor just how many couples are getting engaged on this Valentine’s Day, turned into a great lesson in reading deeper into the social media tea leaves. We used Cision’s software to track and analyze all of the social media conversations swirling around about getting #engaged. Bring on the gleeful and gushy posts, with a healthy serving of engagement ring selfies!
Out of the more than 44,000 social media posts created during the holiday discussing getting engaged, nearly three out of four were more likely to be negative in sentiment (70.1 percent) than positive (28.5 percent). If this is supposed to be the most romantic holiday celebrated each year, and people are specifically talking about getting engaged, then why all the negative posts? Maybe people are getting upset seeing all the lovebirds, and are feeling left out or jealous?
“WRONG.” — President Trump
Cision’s software gives us the ability to sift through these social media posts and analyze them at a granular level. We wanted to get into the weeds of what’s really upsetting everyone who’s talking about being engaged this Valentine’s Day.
Out of the social media posts with a negative sentiment, some of the top words used were:
- “sugary drinks”
The common theme of these top words used clearly indicate these negative social media posts were not coming from people upset about having a bad date on Valentine’s Day. Rather, they show that many social media users were talking about political matters, such as proposed government legislation, federal regulations or policies. Specifically, it looks like much of the negative social media posts were about a study from the University of Melbourne that proposed taxing sugary, salty, or fatty foods, claiming that it would “add an extra 2.1 years of healthy life for every 100 Australians,” according to the Sydney Morning Herald.
What started off as a quest to find out how much true love was really in the air this Valentine’s Day, turned into an interesting and unexpected learning experience: Junk food might cost you more in Australia in the near future, and big data needs to be refined (just like those sweet, sugary drinks!).
After observing how conversations about this major holiday were unexpectedly overtaken by another event, this illustrates how when brands are monitoring, they need to take into consideration external factors that may affect their data and analysis to ensure their data is as clean as possible. Here are three monitoring pitfalls you can encounter when gathering social media data, that may leave you vulnerable to dirty data that’s inaccurate and misleading:
- Having too broad of search criteria
- They key is to find the right balance between using a search parameter that’s tight, and restricted, and one that’s loose-ended and not specific enough.
- In this case, we could have considered including monitoring of the words “Valentine’s Day” or “#valentines” in conjunction (using an “AND” statement, for you boolean nerds out there) with all of our “engagement” words we were monitoring. We didn’t include Valentine’s Day-specific words in our monitoring because we didn’t want to potentially limit mentions of “engaged” or other related words because not every single person talking about being engaged may necessarily also mention “Valentine’s Day.”
- Only looking at the surface
- When looking at the raw data, you can churn out numbers and draw conclusions very quickly. If you don’t thoroughly examine the numbers though, you could be missing a very different reality behind the data.
- In this case, we could have seen the high number of negative social media posts in our monitoring and stopped there. What a great headline that would have been if we concluded: “Valentine’s Day: The Unhappiest Holiday Ever.” This would not have matched reality. Always thoroughly investigate the data!
- Forgetting to take into account other trending events during your time of monitoring
- The event you’re monitoring might share keywords or phrases with another event occurring at the same time as yours. Sometimes you can predict this, and plan your search terms accordingly, but not all events are planned. The unpredictable can happen, and you need to keep this in mind when examining your data. You should try to identify any patterns or trends developing in your data that might suggest another conversation occurring outside of your specific event.
- In this case, the University of Melbourne “proposed” a tax, and the internet was not happy about it. We were looking for mentions of “proposed,” and we got them. But there were two very different conversations occurring around that word on social media.
Social media monitoring is an excellent way to identify conversations occurring about a specific event, topic or even your own brand. The key takeaway from this social media monitoring event is that being able to dig deeper into the data is critical to discovering the reality of what the raw data is presenting. If you’re not able to go in depth with the data you’re relying on to ultimately make informed decisions, you could you run the risk of misrepresenting the facts to your stakeholders.
Communications Best Practices
Get the latest updates on PR, communications and marketing best practices.
Cision Product News
Keep up with everything Cision. Check here for the most current product news.
Thought leadership and communications strategy for the C-suite written by the C-suite.
A blog for and about the media featuring trends, tips, tools, media moves and more.