Page 136 - Towards_a_New_Malaysia_The_2018_Election_and_Its_6146371_(z-lib.org)
P. 136
Big Data Campaigning 121
parliament, Noriah Kasnon, died in a helicopter accident. In GE13, Noriah
Kasnon had won with only a tiny majority of 399 votes, of which 66 per cent
were Malay and 31 per cent Chinese (Malaysiakini 2016). In the aftermath of
his victory, Budiman expressed ‘surprise’ he won so convincingly, and thanked
his ‘election machinery’ (Chan 2016). Azrin said the point of being involved in
Sungai Besar was ‘a case study to prove to the PM we can win a larger contract.
It was a showcase of behavioural data. We build psychological pro les of the
voters. We start analysing their behavioural pattern. How data could be used
to persuade votes.’ CA presented this work to Najib prior to 2018, asking for
a price of USD12 million. eir nal pitch was similar in Malaysia as it was
elsewhere in the world post-2016: ‘We won Trump’, says Azrin. ‘Despite all
the odds, we won Trump. When everyone else said Trump would lose, we were
con dent that he would win. We did the data-crunching, we knew about a
week before that Trump would win. And we were spot on’ (interview, Azrin,
Shah Alam, February 2018). But the pitch did not work. Even before the CA
scandal broke out, Najib never responded to their o er. 1
In the aftermath of GE13, newly re-elected Prime Minister Najib realised
he needed to do more work in the social-media realm. He employed the
company ORB Solutions, which would later be renamed Resonate Asia.
is company hired around 30 sta , mostly developers and programmers
in their twenties. ey reported directly to Najib, providing social-media
sentiment analysis and state-by-state predictions of election outcomes, and
gauged voter groups by examining Facebook content in at least ve languages:
Malay, Hokkien, English, Iban, and Tamil (interviews with BN sta ers, Kuala
Lumpur, February and April 2018).
Exactly what kind of information was gathered and used to target
undecided voters? Obviously social media were central, but BN sta ers talked
of having signi cant amounts of data that could be analysed to target voters
(Boo 2017). Azrin Zizal explained that, ‘We already have a good stash of
scattered data. We just need to get it organised. ere are ways to purchase
data. We are slowly building the data landscape’ (interview, Azrin Zizal,
Shah Alam, February 2018). Other BN sta ers I spoke with discussed the
wide range of data possibilities for BN—hypermarket cards collect data
on what people purchase, cable-television companies provide data on what
people watch, telecommunications companies Maxis and Telekom apparently
provide some information and data that can be purchased. One big data
campaigner claimed that, ‘if we need to we can buy through the side door’.
BN’s Tun Faisal noted, ‘ ere are so many sources that they can mine the
information from—like handphones, GPS, social media. You can see trends.
This content downloaded from 139.80.253.0 on Fri, 06 Nov 2020 04:22:11 UTC
All use subject to https://about.jstor.org/terms

