Tag Archives: #topsy

Who has better sentiment, Nike or Reebok?

In the past we’ve covered the social presence of Nike versus its much smaller competitor Reebok. Now that we have new social analytic tools at our disposal, we can continue the comparison with some cool insights. I used both TOPSY and socialmention*.

Starting off with tweets per day

While it was clear Nike would have many more mentions on Twitter, . I took a look at the past month and found an interesting pattern. Nike is much more volatile. It goes from ~70K to almost ~200K with a day (new campaign launch), dropping back just as fast to around ~90K for the rest of the week. The news just doesn’t stick, it gives only a temporary bump. Reebok on the other hand is much more stable, sure it has some ups and downs, but they tend to take a while to build up and then decrease. Funny enough, when comparing to the generic term “sneaker” – we can see that it has consistently outperformed the Reebok brand. So is the brand weaker than the product, or is Nike the outlier by being strong than its product (like “Frigidaire” used to be compared to “Refrigerator”).



In this case – it’s a tie. Topsy reports that over the past month Reebok had a positive vs negative sentiment score of 70 (from 154K tweets) versus Nike’s 72 (from 2.5m tweets).

Socialmention gives us a bit more in-depth information. It rates Reebok as heaving a 39% reach and a 12:1 sentiment ratio – meaning 12 positives for each negative mention and a huge neutral majority. Nike o the other hand, has an ever so slightly larger reach of 46% with a better sentiment ratio of 16 positives for each negative.


Leave a comment

Filed under 2014 - Post 3, Analytics

Is the next big thing already here?


Samsung’s advertising campaigns suggest that the next big thing is already here (their phone of course).  Jump past the break to see what the social media world thinks…

Continue reading

Leave a comment

Filed under 2014 - Post 3

Social Mention vs Topsy on Ameritrade vs Scottrade

Once again, Ameritrade and Scottrade are selected for my social media research. We looked at their social media channels and cultures in my last write-up. Let’s analyze their social media activities through Social Mention and Topsy. Apparently Social Mention gives you a more “general” view of social media activities while Topsy puts more emphasis on tweets monitoring, though they provide links, photos and videos social presence analyses as well.

Social Mention (1st left) and Topsy (2nd and 3rd left) for Ameritrade:


Social Mention (1st left) and Topsy (2nd and 3rd left) for Scottrade:


Topsy offers a rolling tweet report for the last 25 days while Social Mention shows MTBM (Mean time between mentioned). Unfortunately, these numbers are not directly comparable, due to:

  1. Social Mention rolls up mentions from various social channels, while Topsy shows only number of tweets.
  2. It is impossible to re-construct the actual “mention data” from MTBM alone.

Next, let’s see if the sentiment scores from Topsy and Social Mention can give us some useful information. Social Mention and Topsy are consistent to yield a conclusion that Ameritrade had more positive social presence reports than Scottrade. Topsy sentiment score goes from 0 (completely negative) to 100 (completely positive), and is at 50 to be neutral. Unfortunately, Topsy doesn’t provide further information on how this score is calculated. Social Mention provided tallies of positive, neutral and negative social “pulses”, and the ratio of positive to negative mentions. Does a high positive to negative ratio a good metric? Does a high ratio bring false sense of security to social marketers? Superficially one would conclude that Ameritrade is doing much better on social channels. How can we know the “strength” of these negative comments by comprehending this data? Here’s my recommendation: Add a trend of sentiment score over a period of time, for example, last 24 hours. The rationale is that a powerful negative comment is highly contagious in social world, therefore, social marketers should be able to infer what is going on by looking at sentiment trend.



Leave a comment

Filed under Uncategorized