Twitter analytics is an integral part of leveraging social media today. Among certain demographics in Kenya, Twitter is by far the social media platform with the most interactions. Therefore, any company engaged in selling must realize that social media is an additional sales channel. Consequently, a well-devised marketing strategy will lead to a larger market footprint, higher sales, and increased profitability.

The backbone of any social marketing strategy is analytics. Paying an “influencer” to promote products might lead to increased sales; however, analytics is critical in gauging the return on investment. Indeed, analytics can determine whether every shilling spent on marketing generates a sufficient return.

In this article, we analyze about 1.2 million tweets (excluding retweets) posted from Kenya. The tweets cover a week, from 4:25 p.m. on Sunday, March 31 2019, to 10:27 p.m. on Sunday, April 7 2019. Note that, due to technical limitations, Twitter does not provide access to every tweet. (An interesting aside is that, globally, tweeps tweet more than 500 million tweets per day.) Our dataset contains tweets from 114,606 distinct users who have either geotagged their tweets to locations within Kenya, or, correspondingly, have set their profile location to places within Kenya. Because of these factors, we do not claim that this dataset is entirely representative of the Kenyan Twitterverse. Nevertheless, let’s dive in and see what insights Twitter analytics can give us.

 

Twitter analytics: INFLUENCE

A logical first place to start investigating is the most prolific tweeps — although in this age of twitter bots, perhaps a better metric would be the most relevant tweeps. Calculating relevancy is quite difficult, and beyond the scope of this first-pass analysis. Below, we present the most prolific Kenyan tweeps for the week March 31 – April 7, 2019.

Tweep	Number of tweets
@KenyaPower_Care	3,874
@kenyan_digest	2,662
@Ma3Route	1,894

 

In the table below we list the most prolific tweeps over their Twitter lifetimes, as determined from our dataset.

Tweep	Twitter signup	Lifetime tweets	Tweets per day
@googuns_lulz	Feb 21, 2015	3,238,665	2,134
@juanmuriango	Jun 29, 2010	2,431,054	756
@KenyaPower_Care	May 24, 2010	1,328,087	408
@radiomaisha	May 3, 2010	1,020,940	312

The first entry, @googuns_lulz, is a bot that does not even try to mask the fact that it is a bot. On this website, the creator of the bot deciphers the tweets; apparently, every tweet is simply an encoded timestamp. In second place @juanmuriango tweets an average of 756 tweets a day, or, assuming 18 waking hours, 42 tweets an hour. This works out to almost one tweet a minute, for 9 years. Clearly, some people take their Twitter very seriously.

Next, we rank the 5 most influential Kenyan tweeps, again, as derived from our dataset.

Tweep	Number of followers
@PirryOficial	2,533,734
@WilliamsRuto	2,243,557
@RailaOdinga	2,171,865
@ntvkenya	2,065,361
@citizentvkenya	2,059,759

Most Kenyans require no introduction to positions 2–5 on the list. However, few would recognize the first tweep, @PirryOfficial. Further analysis reveals that he is a prolific Colombian tweep, who, on a visit to Kenya, tweeted a video of himself in Tsavo. Since his tweet is geotagged to a location within Kenya, Twitter’s algorithms assume he is Kenyan, though he isn’t. This illustrates just one of the many possible pitfalls in performing geographical analysis with Twitter data.

 

twitter analytics: media And HAshtags

Twitter users can include media (pictures or videos) in their tweets for increased interaction. This is perhaps the easiest way to increase interactions, and will significantly increase the number of retweets, favorites, and replies. In our dataset, 15% of the tweets included media. Below, we present the average number of tweet interactions, based on whether media is present or not. The table shows the importance of attaching media to tweets.

Interaction	No media	Media present	Difference
Retweets	4.1	7.3	78%
Favourites	6.3	14.4	129%

Clearly, attaching media to tweets leads to a marked increase in interactions. In addition, carefully selected hashtags can also spur interest and interactions. Below we list the top five hashtags in our dataset.

Hashtag	Number of tweets
#MainaAndKingangi	5,085
#JSWR	4,841
#StateoftheNation	4,246
#SOTNKe2019	3,877
#Kwibuka25	3,874

 

twitter analytics: source and timing

“How do people tweet?” Is it via an app on a phone, or from a web browser? Answering this question is important, since understanding your audience helps in crafting a tailored message. In the table below, we notice that 57% of the tweets were from an Android device. Once again, Kenya bucks the global trend, where an overwhelming majority of tweets are sent from iPhones.

Platform	Count	Percent
Twitter for Android	676,925	57%
Twitter for iPhone	159,905	14%
Twitter Web Client	107,176	9%
Twitter Web App	80,473	7%
Facebook	39,997	3%

Drilling deeper into the data, we observe that mobile devices send a minimum of 71% of the tweets in our dataset. This is in keeping with global statistics, where 80% of Twitter usage happens on mobile devices. The takeaway is that media attached to tweets needs to be formatted for best viewing on mobile devices, since this is where most interactions take place.

Lastly, we tackle the question, “when do most Kenyans tweet?” Below, we bin the 1.2 million tweets according to the time of posting.

Evidently, tweet volume peaks between 8-9 p.m. In contrast, peak tweet volume globally is noon to 1 p.m, local time. Why the difference in Kenya? One possibility might be Kenyans’ traditional dependence on the 7 p.m. evening news. It is likely that news reports drive Twitter conversations. Another possibility is that Kenyans have more time in the evenings to tweet. A marketing strategist can use this information in several ways. First, since 8-9 p.m. is the busiest tweet time, tweeting during this period might lead to more interactions. On the other hand, this could also be the worst time to tweet — your message might get lost in a flood of competing tweets. Further analysis could clarify this ambiguity.

 

Conclusion

Twitter analytics is a vital component of a social media marketing strategy. By digging deeper into Twitter data, one can extract actionable insights, leading to superior branding, better messaging, and increased revenue. We presented some insights and tips which you can incorporate in your own Twitter strategy. Alternatively, contact us to discuss your social media and analytics solutions.