an analytics blog for public use: February 2016

Sunday, February 28, 2016

Sentiment analysis of e-Commerce firms in India: February 2016

Dear readers,

UPDATE: Download complete report here.

Ranking of following 6 e-commerce firms based on the consumer sentiment derived by analysis of tweets at their Twitter handles has been done:

Amazon ~ @amazonIN,
eBay ~ @ebayindia,
Flipkart~ @Flipkart,
Jabong ~ @JabongIndia,
Myntra ~ @myntra &
Snapdeal ~ @snapdeal

10,000 tweets from each handles have been solicited between 01/02/2016 and 27/02/2016 and parsed.
Analysis results are as under:

Sl. No.	e-Commerce firm	Total tweets	very.pos. count	very.neg. count	very.tot	Score (100)
1	Amazon	10000	673	115	788	85
2	eBay	1470	40	76	116	34
3	Flipkart	4628	631	135	766	82
4	Jabong	5961	874	28	902	97
5	Myntra	3112	757	28	785	96
6	Snapdeal	8467	1103	385	1488	74

New rankings are:

Rank 1	Jabong
Rank 2	Myntra
Rank 3	Amazon
Rank 4	Flipkart
Rank 5	Snapdeal
Rank 6	eBay

(click on the image to enlarge)

Detailed data dumps are available on request.
Please send your requests in comments below.
Feedback is solicited.

Best,
N. Singh

Sentiment Analysis of telecom operators in India: February, 2016

Hello readers,

How are you? As TRAI forces the operators to improve the connectivity and reduce call drops, we take a look at how some of the telecom operators fared in their social media presence. We requested 10,000 tweets and measured the response of active Twitterati at the official hashtags of the following:

Airtel ~ @airtelindia,

BSNL Mobile ~ @bsnlmobileindia,

Reliance Mobile ~ @RelianceMobile,

Tata Docomo ~ @tatadocomo,

Idea Cellular ~ @ideacellular &
Vodafone ~ @VodafoneIN

(Click on the picture to enlarge)

The Twitter Sentiment Score is calculated as follows:

TSS = 100* (Total positive tweets/ Total tweets)

A higher score is better and shows an overall positive sentiment among twitter feedback left with respective company's official hashtag. Positive tweets have score more than +2 and negative tweets have score less than -2.

(Click on the picture to enlarge)

Word clouds are available here: REPORT

The individual data files are available on request. Leave your requests in comments below.

Best,

N. Singh

Shell Ideas360

Cleared the STAGE 1 of Shell Ideas360!

Saturday, February 27, 2016

Sentiment analysis of Indian travel planning companies- February 2016

Hello readers,

February has flown past in the Northern hemisphere and spring has set in completely.
In this weather before you plan for that great Indian trip, we bring to you a look at how have some of the travel firms in India been doing on Twitter this month!

We have carried out the sentiment analysis of four of the top Indian travel planning firms, viz.

1. MakeMyTrip - @makemytrip
2. GoIbibo - @goibibo
3. Cleartrip - @cleartrip
4. Yatra.com - @YatraHolidays

For this analysis, 1000 tweets between 01/02/2016 and 27/02/2016 using the above tags have been parsed using twitteR package.

Score for a tweet is the number by which the positive words outscore negative feedbacks.
Any tweet with a score more than 2 has been considered as very positive and any score less than -2 has been considered as very negative.

The results of analysis are as under:

MakeMyTrip is again number 1 with a very high score of 94/ 100.
GoIbibo is second with a score of 70/100 and is followed by Yatra and Cleartrip.

Yatra also needs to increase the use of its hashtag if it wishes to compete on social media presence in terms of volume. Its hashtag returned only 176 tweets when asked for 1000.

So, there is a lot of catching up for GoIbibo to do on social media platforms at least!

Sl. No.	Travel	Total tweets	very.pos count	very.neg count	very.tot	Score out of 100
1	Makemytrip	1000	383	25	408	94
2	Goibibo	486	62	26	88	70
3	YatraHolidays	176	12	10	22	55
4	Cleartrip	299	8	18	26	31
(1000 tweets requested from dates between 01-02-2016 and 27-02-2016, both inclusive)

very.pos. count = tweets which were very positive

very.neg. count = tweets which were very negative

Here are the word frequency tables and wordclouds of the tweets parsed:

1. GOIBIBO

GOIBIBO
Word	Freq
goibibo	216
thanks	56
hotel	47
will	40
booking	36
makemytripcare	33
customer	32
can	31
get	31
team	29

(Click on image to enlarge)

2. CLEARTRIP

CLEARTRIP
Word	Freq
cleartrip	167
axisbank	35
cards	30
deals	30
travel	29
days	28
experience	28
axisbankoffers	25
cancun	25
cavern	25