Peerreach Blog

28 Oct

Ugandan Elections 2021 Survey

Kyagulanyi gets over 60% percent of the votes, but people don’t think elections will be honest. Election Observers from the EU have not been invited.


By: Nico Schoonderwoerd (Ph.D.)

Published: 28 October 2020

In February of next year, elections are going to be held for the presidency and parliament of Uganda. President Yoweri Museveni and his party NRM came to power in 1986 and in the 5 elections that were held since, he was always declared the winner. In 2016 the elections were marred with irregularities, as reported by observers of the European Union. For the 2021 elections the Ugandan government has decided not to invite European Election Observers. Such an invitation is necessary for them to do their work, according to European Union official Isabel Martinho.


In most recent elections the main opposition candidate was Kizza Besigye, but he decided not to run for his party FDC. In the last 2 years a Ugandan lawyer & pop star turned politician , mostly known by his artist name of Bobi Wine, came to the spotlight. Where in the past all opposition leaders were from the same generation of Museveni, Robert Kyagulanyi , Ssentamu (his real name) is only 38 years old. This is considered an advantage to attract votes among a population that is 78% below the age of 30, which is extremely young. Other opposition candidates are Mugisha Muntu (ANT), Henry Tumukunde (UPDF), Patrick Amuriat (FDC) and a large number of independents.


The opposition suffers from random arrests by the government of Museveni and does not get access to media that is controlled by people around Museveni, limiting voters to make an informed choice. Surveys that poll political preferences are scarce, the last one that became known was in March and performed by RWI. In that poll Kizza Besigye got 17% of the votes, but he isn’t even running. A lot has changed since March, and we think independent election surveys are important for a sound election process. This is the reason why PeerReach, in collaboration with Freedom Uganda, has decided to step in.


In the past PeerReach has published various analyses of the Twitter population (references at the bottom), now it’s polling the people of Uganda through Whatsapp.


Kyagulanyi defeats Museveni


The survey was held from October 1st to October 27th by a team of 3 people with mixed political preference. First we constructed 12000 fully random telephone numbers. Among those 12000 numbers were some 1200 people that were found in the Whatsapp database, and we sent them a message asking them for their availability for some questions. There were 304 responses. We then asked their age, home district and who they would vote for. Respondents happily gave their personal details but were reluctant to share their political preference. Asked if it’s dangerous to share their preference, many confirmed this.


  • Lets hope you are not investigating me, there is a lot of kidnap here.


  •  I’m fearing of my life


We also discovered that people were more willing to respond if you mentioned the names the candidates first, so that’s what we did. Still, we found that we needed to build trust with the respondents before they gave an answer. Some video called us up to verify who we are, some asked us to send a picture. The fact that we operate from outside of Uganda made it easier to convince people we work independently. The fear of potential consequences for sharing their preference is a likely reason why surveys in the past have turned out to be inaccurate.


Robert Kyagulanyi receives 61% of the votes with a standard deviation of 5%. The probability of him acquiring more than half of the votes is 97.5% (in scientific terms: the p-value is 0.025, which means it’s scientifically significant).


Kyagulanyi voters talked very enthusiastically about their candidate and shared videos and other marketing material.


  • Kyagulanyi Ssentamu is growing up his percentage of people


  • (..)  I really like his political ideology


  • To me i would say a number of people are making up their minds for Robert kyagulanyi sentamu becoz our current president is basically a dictator


34% of the votes were for Museveni, also with a standard deviation of 5%.


  • I would vote for museveni. Kyagulanyi needs more time to understand leadership dynamics. We may change for worse, something Uganda is not ready for. We need to change for better.


  • We want Museveni as president only, no change


  • Kyagulanyi has good ideas for the country but he is still politically  immature and he is surrounded by intolerant people who may become violent in a way of preventing against  the impunity inflicted on them by the security forces of president Museveni

Mugisha Muntu receives almost 3.5% of the votes, but with a considerable margin of error of 2%. Asked for a motivation respondents said:


  • (he) has been active in Ugandan politics for more than 30 years making him one of most experienced candidates. He is the one of the most disciplined top army officers with a clean background.


  • We need a change but  a change with justifiable reasons not excitement. Kyagulanyi looks popular because of music background but he lacks a number of qualities  of a leader  Uganda needs if we’re to have a new president Uganda needs now


  • Muntu lacks the political face, he might not score much. It is a race between a cheetah (Museveni), a deer (Kyagulanyi) and a chicken (Muntu).


Other candidates received 1.5% of the votes. 16% of the respondents indicated they would not vote, 11% were undecided and 3% wouldn’t share with us.

Consistent support among gender and area.


Although only a small majority of the Whatsapp accounts we used were male, among our respondents two thirds were male. However, if we compare results among males versus results among women, we find that Kyagulanyi received about 60% of the votes among both genders.


People in cities are more likely to have phones than people in rural areas. 70% of our respondents were from urban areas, the other 30% from rural areas. However, for their votes it doesn’t seem to matter. This needs to be confirmed by further research. This is what one of our respondents had to say about it:


  • In urban centers they prefer Bobi Wine while in villages prefer Mzee. Villagers want money. Like they give them money when campaigns are running. Yes coz whoever goes just aim on satisfying their stomachs. In Uganda leadership is all about grabbing money and be happy

The average age of the respondents was 30 years. Below 35, a huge majority supports Kyagulanyi, above 35 a small majority of our sample supports Museveni.

Most Ugandans don’t think elections will be honest


We asked our respondents if they expect elections to be honest. 58% indicated No, 35% said Yes, the remaining 7% didn’t know. A big majority of Kyagulanyi voters don’t believe in honest elections, but also a small part of the Museveni voters.


  • I am afraid of constant rigging of votes by NRM government which is very very very rampant (…)

  • it can’t be honest cos its dictatorial regime

  • No honesty in Uganda.

  • I barely think of ya getting a healthy election cuz we know the after all person who’s going to win.

Uganda is worse, especially corruption

We asked the respondents how things have changed in 5 years with respect to corruption, health care, education and their personal financial situation. Below you find the results.


73% of our respondents think that corruption is worse.

  • Corruption everywhere just know that’s why some of us are jobless. Jobs are offered to their fellow tribemates. Me am scared of 2021. I think I will run away


  • M7 in this term has tried to fight corruption only that it is deeply rooted. He tried to fire leaders of companies & put investigators that report to him directly but it’s still not enough.


  • corruption is the order of the day in Uganda, a lot of unemployment even if you are educated (..)


  • Worse (..) in some offices to give u a job a man want to first use u as a sex material


  • They are all for their own finances and improving there standards of living plus relatives’

Concerning health care the responses are mixed; 29% think it’s worse, also 29% it’s better.


  • Health care is the same. Treatment is very costly


33% of respondents think that education is worse.


  • Education has slightly improved in urban centers


  • Education is getting better as the syllabus changed recently with interest in more hands on courses that can create self employment. Would be better if not for Covid 19


  • Very worse



In 5 years the personal financial situation of respondents has deteriorated. 63% say it’s worse, only 16% say it’s better.


  • my financial status is messed up…I work for my stomach.


  • Am a student but my financial status are not all that good (..)


  • Our government has not helped us at all ,the poor is very poor and the rich is very rich


This survey shows that life in Uganda for our respondents has become more difficult in the last 5 years, and they have seen corruption increasing. Education is also worse. Health care is perceived as the same.

Although most do not expect an honest election, a big majority is planning to vote. The biggest wish of the respondents is change, and they expect to find this with their votes on new opposition leader Robert Kyagulanyi (61%). Even many of those voting Museveni (34%) are asking for change.

Statistics give a probability of 97.5% to Kyagulanyi having the majority of the votes. The support for Kyagulanyi is consistent among males and females, and those living in urban and rural areas. Most Besigye voters in 2016 are now voting Kyagulanyi. Two thirds of the Museveni voters from 2016 stick to their favorite. But the new voters, in overwhelming majority, vote for Kyagulanyi. If these young voters will show up in the polling stations then the change they are asking for might indeed be realized.




For readability some typing errors in quotes were corrected.



  1. European Commission Observer Mission 2016

  2. RWI poll

About the author:

Nico Schoonderwoerd has a Ph.D. in physics. He published 4 scientific articles about Quantum Physics. He was co-founder of PeerReach, a social media analytics company. PeerReach research was published in various outlets.



LA times:

Algemeen Dagblad:

No comments • Written by Nico
25 Jul

Are celebrities buying fake followers?

Sites like ours and clearly display that many celebrities have huge amounts of fake followers. Are they buying those fake followers or is there an alternative reason?

Indeed, some celebrities (and non celebrities) are buying fake followers. I did publish about this before on A company offering fake followers has to create those fake accounts. They write a computer program that automatically signs up new accounts and follow the same steps that you and I follow when we create a Twitter account. This morning I created a new twitter account @nicotest1234, and nowhere did I have to fill out a Captcha code. So it’s easy to write such a program.

When you create a new account, Twitter suggests accounts to follow.

Schermafdruk 2015-07-25 14.30.50


By default, ‘Popular accounts’ is checked. The following accounts are in this group

Schermafdruk 2015-07-25 14.31.43

Schermafdruk 2015-07-25 14.31.51


Using the Audience Insight section of PeerReach (displayed on the bottom of a profile after signing in using Twitter credentials) I found out that all those accounts are quickly gaining followers since the summer of 2014, when Twitter changed the sign-up process.


Schermafdruk 2015-07-25 15.08.40


Schermafdruk 2015-07-25 15.09.40


Schermafdruk 2015-07-25 15.10.09


In the Audience Insights for @robwijnberg, @alexanderNL and @fritswester we see that all of them are gaining followers quickly since mid 2014, and that most of these accounts are spam (fake) accounts.

Are @robwijnberg, @alexanderNL and @fritswester paying for those followers? Very unlikely, especially since I know two of them personally. However, some people are buying fake followers. The accounts created start following the accounts that Twitter is suggesting. Therefore also sincere Twitter users end up having many fake Twitter followers.



No comments • Written by Nico
05 Feb

More active twitter users after IPO, but they tweet less.

We’ve updated some of the numbers from our big Twitter research from last November to give you some insight on Twitter’s first quarter as a public company. Twitter’s official number will be published later today.

The first metric we’ve looked at was the number of Monthly and Daily active Twitter users. The number of accounts tweeting at least once a month (MAtU) has grown from 117 million in October 2013 to 122 million in January 2014. The number of accounts that tweet on a daily base has actually fallen from 45 million to 41 million  over the same period.

About 40% of Twitter users actually don’t tweet but only read tweets, that’s why number published by Twitter are higher than ours.


Twitter is approaching the magic number of 1 billion registered accounts. In July they had 825M, in October 904M and now 923M. This looks like a strong flattening but it should be noted that Twitter is very actively removing spam accounts. Some months they remove more than others.

Languages on Twitter

We looked at the most important languages spoken on Twitter and found some interesting results.

English is the main language on Twitter, but the number of MAtU that have English as their main language is going down from 62M to 61M.

Spanish is growing from 16M to 18M and Japanese from 11M to 12M.

Portuguese, which Twitter users are mainly from Brazil, has declined from 4.5M to 4M.

Noteworthy is that Twitter usage is spiking in arab (mainly Saudi Arabia). Usage went up 25% to 5.6M.

6 Comments • Written by Manfred
07 Nov

4 ways how Twitter can keep growing

The 7-year old microblogging platform Twitter will be listed today for an initial price of $26 and investors worldwide have to decide whether to invest or not. Almost 1 million new users register daily, but the number of active Twitter users is not growing as it used to. In this article we will explain why, and in which areas new growth can be realized.

Twitter has caused a revolution in mass communication. Top politicians and famous artists don’t have to contact journalists to broadcast their message, they can just post a tweet. Also for the average person it’s much easier to spread important messages. When an airplane landed on the Hudson river, a photograph was spread worldwide within minutes. In the Arab spring revolution, Twitter and Facebook had an important role to amplify the voice of the people. Twitter also plays an important role as a conversation platform in congresses and for television programs.

Twitter is the only big Social Media Platform that allows 3rd parties to use the Twitter Platform. This allows other companies to create services useful for tweeps and help the Twitter ecosystem.

PeerReach is such a company in the Twitter ecosystem.  It identifies influential accounts and relevant content, and therefore it constructed a database of 130 million active twitter accounts. Based on our database we can determine the way that Twitter is used. Twitter does not provide very detailed information about its usage.

July 2013
October 2013
Sample error
Registered accounts825M904M9M
Monthly active users*218M232M0
Monthly active users USA*53M
Monthly active tweeting users114M117M3M
Daily active tweeting users42M45M2M
Timeline views151000M158000M
New accounts30M28M2M

It’s important to note that we use Monthly Active tweeting Users, and Twitter uses Monthly Active Users as a metric. About 40% of the Monthly Active Users actually don’t tweet, but only login to the platform.

Below we see how Twitter grew in time. In 2009 the rapid growth of Twitter started and it passed the milestone of 100 million registered users. In April 2009, Ashton Kutcher was the first to reach one million followers. Now there are more than 1600 of such accounts.

Although Twitter generated revenues of USD 422M in the first three quarters, it’s still having a net loss of 133M. Here are our 4 suggestions how Twitter should continue to grow.

1. Generate revenues from non-US users

Twitter had revenues of 121 million in the second quarter, of which 75% are generated in the USA. However, only a quarter from the Twitter population actually comes from the USA, and Twitter should generate more return on its user base in Japan, the United Kingdom or Brazil. 1000 timeline views in the US generate $2.58 in revenues, whereas 1000 international timeline views generate only $0.36 in advertising income.  

The huge difference in Twitter usage per country is clear if we calculate the Twitter penetration per country. The penetration is defined as the number of monthly active tweeting users relative to the total amount of internet users in that country.

The picture displays the 23 countries with more than 800.000 active tweeting users. Of these countries, there are 14 with more than 5% active Twitter users. The Top five countries are non English speaking countries, with Saudi Arabia even having over 32% active twitter users.

Saudi Arabia and Indonesia are two countries known for the fact that most internet users don’t have a PC, but access the internet through mobile.

2.    Identify growth countries and groups

Our graph with Twitter penetration numbers clearly shows a number of green fields for Twitter. In big countries like Nigeria, Germany and India Twitter usage is still low. China is not listed because Twitter is banned from China and therefore the Chinese microblogging platform Weibo has no competition. According to the SEC filing, Twitter has considerable competion in some local markets, such as Kakao in South Korea and Line Japan.

Using the PeerReach algorithm and turk, we determined the Gender distribution per country. Globally, 49% of Twitter users is male.

3. Target the young and users on mobile.

Twitter users are young, on average 24 years. The average male is 26 years old, the average female 22 years old. Teenagers dominate Twitter. Only 20% of the tweeps are older than 30.

In the graph below we find that the average is quite different from country to country. Asian countries have the lowest age on average. English speaking countries have older Twitter users.

In most graphs we don’t show the statistical error, but here we displayed the error bounds on the average age (1

The driving factor for the average age is the percentage of teens among the total active Twitter population. This is very different from country to country. In 5 Asian countries the teen percentage is above 60%.

4. Improve retention for new accounts

Of the 45 million people that tweeted on October 25th  (the day we did our count) the big majority were already on Twitter on July 25th.  Twitter has a stable group of active users.

However, low retention for new users is the weak spot for Twitter.  Of the 90M accounts that registered in the last 3 months, only 3M turned into daily active tweeting users.

For the accounts that registered in July, we counted the number of tweets until 25 October. 56% of these accounts haven’t placed a single tweet. Only 8% tweeted more than 50 times.

For new users the Twitter experience is very overwhelming. Search results are chaotic, discovery is difficult, and the trends, invented by Twitter themselves, haven’t been improved since many years.

If Twitter can improve retention of new users, it should be able to get past 500M world wide active users and the stock price will sky rocket.



Internet usage:

Twitter booming in Saudi Arabia:

SEC filing:

177 Comments • Written by Nico
08 Aug

The Tweet-spendables

Also don’t forget to check out their full PeerReach profile:

Jackie Chan

Sylvester Stallone

Arnold Schwarzenegger

Terry Crews

Steven Seagal

No comments • Written by Manfred
21 Jul

Who wins the yellow-blue jersey on Twitter? Six most remarkable climbers of #tdf.

Without publicity there wouldn’t be professional cycling and therefore every top cyclist has his own twitter account. During the most important cycling race of the year they inform their fan base about their whereabouts and share inside pictures that you would never see on any television network.

Read more…

1 Comment • Written by Manfred
17 Jul

1 Reason why Gonzalo Higuain’s Twitter Audience increased

Since April, rumors about a possible transfer of Real Madrid’s Gonzalo Higuain have been spreading throughout the world of football. The popular Argentinian striker initially was linked to Juventus, but after the Italians aligned Carlos Tevez, Arsenal coach Arsène Wenger spoke openly about his interest in the player.

Read more…

No comments • Written by Manfred
10 Jul

Australia’s top fashion accounts on Twitter

As promised, the second in a series of short posts highlighting some of our PeerGroups. The second one is Down Under’s Fashion list. Below you’ll find the top 5, check out the full list here.

No comments • Written by Manfred
08 Jul

UK’s top fashion accounts on Twitter

This is a first in a series of short posts highlighting some of our PeerGroups. The first one is UK’s Fashion list. Below you’ll find the top 5, check out the full list here.

No comments • Written by Manfred
04 Jun

Extreme Twitter lists power

Twitter recently increased the number of lists you can create from 20 to 1,000; but creating useful lists by hand can be a time consuming job. That’s why we want to highlight another cool feature of PeerReach: adding our PeerGroup lists to your Twitter account so you can join the conversation with the peers from your interest area. Be it sports, science, fashion or one of our other lists.

Here is short tutorial on how to add one of our lists to twitter.

First find the list of your choice and make sure you’re signed in with PeerReach. You’ll see a “Create this list on Twitter” button on the right side. This will add the top 100 users to a new list within your Twitter account. This process can take up to a minute.

Here’s an example on our music list:

After the list is created you can use the “view list” button to see the list’s timeline or find the list on your Twitter profile or within your favourite Twitter client.

4 Comments • Written by Zlatan