Every very day millions of mostly frustrated consumers take to social media to volunteer feedback about their banks. DataEQ conducted an exploratory study of this volunteered consumer feedback for two Saudi Arabian banks and two United Arab Emirates (UAE) banks. The study analysed public social media posts in both English and Arabic. Using a combination of AI and human crowds, DataEQ was able to surface the main issues driving complaints, identify differing priorities amongst English and Arabic-speaking authors and explore the issues driving digital banking-related conversation.
Similar to other markets studied by DataEQ, the most prominent topics which drove consumers to change banks were accusations of unethical behaviour as well as poor turnaround time of consumer queries. This was common across both English and Arabic authors on social media. Though overall net sentiment towards the banks’ digital services was low, Emirates NBD came out on top as the only banks with a net positive sentiment driven by positive perceptions of its digital banking innovations.
Social media data is a valuable but largely untapped source of customer intention and expectation data. It captures how people feel, without the bias of a questionnaire or a focus group moderator and, if properly evaluated and structured, can help organisations such as banks, improve governance, mitigate risk, reduce churn and improve customer experience.
Historically, the inaccuracies of using AI-only sentiment analysis has limited social media use to reporting on marketing metrics. DataEQ has overcome these challenges with a unique approach that combines machine learning with proprietary crowds comprising human contributors.
Methodology: combining AI and human intelligence
For this study, DataEQ retrieved 336 531 public consumer social media posts about two Saudi and two UAE banks from 1 November 2017 – 30 April 2018. The four banks: National Commercial Bank, Al Rajhi Bank, Dubai Islamic Bank and Emirates NBD were selected based on having the highest volume of social conversation of all the banks in their respective regions over the research period.
A random sample of 59 830 of these posts were evaluated for sentiment by real people on DataEQ’s Crowd platform - enabling an accurate understanding of consumer sentiment. Human verifiers assigned each post a sentiment of positive, negative or neutral, or in some cases, both positive and negative
Drivers of sentiment
Topic wheels illustrate the topics of conversation by volume
A topics analysis was conducted to surface the specific issues driving consumer sentiment. 15 621 of the sentiment-bearing mentions were distributed to DataEQ’s Crowd contributors who identified which banking-specific topics were contained in each post.
The study found that Emirates NBD performed best of the four institutions (-2.8%), followed by Dubai Islamic Bank (-15.5%).
National Commercial Bank had the most negative net sentiment (-45.1%), followed by Al Rajhi (-26.3%).
Comparative sentiment performance
Net sentiment is calculated by subtracting the percentage of negative sentiment from the percentage of positive
The most widespread themes behind customers’ expressing an intent to leave their bank were accusations of unethical behavior and turnaround time.
Major differences between Arabic and English authors – UAE
Arabic speaking authors in the UAE were most concerned about perceived unethical behaviour in their banks (37.8%) while English authors were most concerned about the turnaround time of their queries (23.7%).
Digital performance
Given the increasing investment in digital services and products, the study analysed conversation about four digital banking themes: online banking, banking app, digital security, and business or technological innovation.
Emirates NBD was the best performing digital bank and the only banking with a net positive sentiment (+7.7%) in digital conversation. Consumers were impressed by Emirates NBD’s innovations such as the Braille-enabled account opening service and blockchain services.
Dubai Islamic Bank had the highest net sentiment for their banking app, as a result of the announcement that Trakhees customers would be able to pay directly through Dubai Islamic Bank’s app.
National Commercial Bank had the most negative sentiment in digital safety or security. The bank’s consumers complained about unauthorised cash transfers and withdrawals. A number of consumers speculated that the bank’s website was hacked, and that their data was not safe.
Al Rajhi Bank had the worst performance in online banking conversation. Consumers complained of technical issues with online banking and having to visit the branch to make any changes.
Understanding what issues are driving consumer sentiment to guide strategy and operations
Across the banks studied, consumers had similar concerns, but priorities and expectations differed significantly across Arabic and English consumers within the UAE. By using real Crowds of local language speakers, we were able to identify not just how people feel but what specific issues from apps to credit cards, were driving conversation.
When accurately structured, social data provides banks with an opportunity to understand their customers’ priorities across channels, regions and languages. In a highly competitive market this data is proving crucial to gain a competitive edge and is increasingly being used to make targeted interventions across the business from customer experience to risk mitigation.
