Thursday, May 22, 2014

Hashy Shoe Report: Before & After

By Blair Heckel

Today’s Blog is another Hashy report. We are going to show you how Interstack can change up these reports to fit your needs whether you are in the retail, hospitality, banking or healthcare industry.
Our before and after report showcases our shoe index for the time period of May 14, 2014 to May 20, 2014.
Here is a snapshot of our “before” report.

Here is a snapshot of our “after” report.

You can see the first difference is that we stopped “listening” to the hashtag #boots. We didn’t eliminate the data already captured but we stopped “listening” to the hashtag to show you how your reports can change with the elimination of one single term. Of course, our customers can log back into the dashboard and turn the hashtag back to “listening” to start capturing the data for that term again. No worries… all of your previous data was saved.
Second, you’ll see that the number of Total Tweets Found has decreased from 20,252 tweets to 13,749 and the Total Users Identified has decreased from 6,979 to 4,499.
Next, comparing the top 10 frequent words side by side you can see that the first four words, shoes, pumps, fashion and women are the same even after #boots is eliminated.

The bottom six words switch from sexy, leather, legs, love, style, and stockings to legs, stockings, highheels, dress, love and style.
These are not all unlike. All of the top 10 associated words still remain within the shoe industry.
The sentiment of the overall tweets had altered, starting with a 10.2% negative sentiment for overall tweets to 11.1%
The total tweets found graph shows the weight of the hashtags tracked, day by day, to fit the majority of the tweets.

The percentage of tweets with assigned values compares relevant tweets or irrelevant tweets pertaining to the hashtags being tracked. In the first report, the percentage of relevant tweets is 36% or 7,292 tweets. In the second report, the percentage of relevant tweets is 38% or 5,159 tweets. The number of tweets has decreased dramatically because of the elimination of the hashtag #boots, which implies that #boots carried in a number of irrelevant tweets and therefore, skewed our data.

Scrolling all the way down to the top 50 words associated with the hashtags being tracked, we can clearly see that after the first 15 words or so, the content of the tweets start to change. Not all of the terms are eliminated, but rather shifted, for words that hold more weight.
Here are some differences in the top 50 associated words:
These words were eliminated
  • Latex
  • Fetish
  • Model
  • Pics
  • Spandex
  • Ankle
  • Baby
  • Clothes
  • Shinny
  • Accessories
  • Crochet
  • Makokariny
  • FemdomPicsVideok
  • LateXXtasy
These words were added
  • Thick
  • Pawg
  • Femalecurves123
  • Opening
  • Beauty
  • Collants
  • Pink
  • Nails
  • Gessabel
  • Grande
  • Outfit
  • Amazing
  • MintPal
  • Shoe

I hope you can see after reading this blog, that when you stop “listening” to a certain hashtag, it can have serious effects on your report. Of course when you are tracking similar terms, your reports might not differ as much versus a report where you are also tracking non-hashtags, accounts, and people.

To gain more insight into the conversations surrounding your industry or to use our solution for a more in-depth analysis, contact us at for a demo of our solution.

For more information, visit

No comments:

Post a Comment