FSCA monitoring Twitter to measure banks’ TCF compliance

Posted on 2 Comments

The FSCA is using social media to supervise whether financial products and services promote fair customer outcomes, and it will be looking into the banks’ low response rate to “priority tweets” as revealed by BrandsEye’s 2021 South African Banking Sentiment Index.

BrandsEye defines a “priority conversation” as mentions that presented a risk to the bank, provided an acquisition opportunity, a chance to mitigate churn, or included a service query. Overall, more than a fifth of banking industry conversations contained a priority theme.

BrandsEye collected more than 2.7 million consumer social media posts about South African banks from 1 September 2020 to 31 August 2021. About 500 000 of these posts were analysed for sentiment and conversation themes, including the Treating Customers Fairly (TCF) outcomes.

Absa had highest overall Net Sentiment ranking in the index and FNB the lowest. The banking industry experienced more negative conversation than positive this year, resulting in an industry Net Sentiment of -7.5%.

The final sections of the Conduct Standard for Banks came into full effect on 3 July, and banks must demonstrate to the FSCA that they deliver on and report on the six TCF outcomes. This includes the banks’ behaviour and interactions with their customers on social media.

Among the final sections implemented from July is section 8, which relates to the regulation of customer complaints, which are increasingly being lodged on social media platforms, Kedibone Dikokwe, the FSCA’s divisional executive: Conduct of Business Supervision, said in the foreword to report on the index

“To ensure that the extent and nature of these social media complaints is accurately captured, the FSCA has adopted a data-driven strategy that includes the monitoring and identification of all TCF-related conversations on social media. In addition to gaining valuable insights on an industry-wide level, this strategy has proven useful in helping to highlight areas of concern among individual banks,” Dikokwe said.

“Considering BrandsEye’s Index finding that only 43% of priority tweets received a public response from banks, the FSCA is looking into this issue, while bringing it to the banking sector’s attention,” Dikokwe said.

However, the report noted that a large component of the 57% of priority tweets that did not receive a response were cases where customers did not tag the bank’s Twitter handle directly in their query and only referred to the bank by name. This means the banks would not be notified of the post, making the mention more difficult to find.

Less than half of priority tweets receive a response

Over the past year, banking customers posted about 188 649 priority Twitter posts, a 28% growth in service conversations. However, 57% went unanswered.

Nedbank, at 63.1%, performed far above the average response level of 43%, followed by African Bank, which responded to 61.1% of priority queries.

TymeBank was the least responsive bank, offering only 28% of priority tweets a response, followed by Absa at 30.2%.

Capitec Bank and Discovery Bank responded to similar proportions of their priority queries, at 36% and 35% respectively, although Capitec’s queue was significantly bigger than Discovery Bank’s.

FNB responded to 46% of priority queries, and Standard Bank to 44.3%.

Large number of complaints about Discovery Bank’s service

The index found that 50.1% of consumer mentions about the banks that expressed either positive or negative sentiment contained one of the six TCF outcomes.

Half of the banks in the index had between 49.5% and 52% of conversations speak to one of the TCF outcomes.

Discovery Bank had highest percentage of conversations related to TCF outcomes, 60.6%.

Other banks with above-average TCF-themed conversations were Capitec (50.8%), FNB (52%), Standard Bank (53.5%) and TymeBank (51.4%).

Discovery had a high volume of complaints about customer service (outcome 5b). Customers who reported being unable to cancel their products caused Discovery to have higher-than-average complaints about switching (outcome 6c), according to the index.

Instances of system outages caused Standard Bank to have a high proportion of complaints related to channel performance (outcome 5c).

TymeBank received nearly as many complaints about product performance as customer service and almost double as much as the rest of the industry. The speed of transfers to and from other banks caused frustration among customers. Transferring money between Everyday and GoalSave accounts also presented a challenge.

Capitec’s channel performance complaints were higher than Standard Bank’s, with instances of downtime presenting a considerable conduct risk. Capitec’s affordability was also questioned, with several authors saying that the bank’s individual transactional fees added up, despite low monthly fees.

Customers criticised FNB and Absa for having high fees. This is reflected in their higher-than-average conversation volumes about product design.

Accusations of racist business practices impacted FNB, as shown by the high volume of complaints about fair treatment (outcome 1a).

Most TCF-themed posts about performance and service

TCF outcome 5, performance and service, was the most notable conduct theme in customers’ conversations about the banking industry, accounting for 85.2% of market conduct conversations.

This outcome included conversations relating to customer service (36.7%), channel performance (20.9%), and product performance (14.2%).

These themes also had strong relationships, as about a fifth of mentions comprised at least two, if not all three, sub-categories, according to the report.

Although most of this conversation was negative, some customers said they had good experiences.

Absa and Nedbank both received the most positive feedback from customers, who said that staff helped to resolve their issues.

TymeBank and Standard Bank did not fare as well, with customers complaining about long turnaround times and attempting to contact the bank multiple times.

Product suitability (outcome 2), which focuses on product design, was the next greatest conduct theme (14.1%). Affordability was frequently discussed on social media. Customers who were let down by products or services criticised the value for money.

Claims, complaints and changes (outcome 6) was the third-largest conduct outcome (6.9%). It comprised nearly equal parts of conversations about accessibility and claims, complaint handling, and switching.

There was a strong relationship between accessibility and claims and complaint handling. Customers who failed to access their funds or make an insurance claim found themselves in drawn-out complaint-handling processes.

2 thoughts on “FSCA monitoring Twitter to measure banks’ TCF compliance

  1. I wonder what software BrandsEye uses to collect more than 2.7 million consumer social media posts about South African banks on Twitter?

    1. According to BrandsEye, they use real people to select randomly 565278 posts from the 2.7m posts, which are obtained using AI.

Comments are closed.

FSCA monitoring Twitter to measure banks’ TCF compliance

Posted on 2 Comments

The FSCA is using social media to supervise whether financial products and services promote fair customer outcomes, and it will be looking into the banks’ low response rate to “priority tweets” as revealed by BrandsEye’s 2021 South African Banking Sentiment Index.

BrandsEye defines a “priority conversation” as mentions that presented a risk to the bank, provided an acquisition opportunity, a chance to mitigate churn, or included a service query. Overall, more than a fifth of banking industry conversations contained a priority theme.

BrandsEye collected more than 2.7 million consumer social media posts about South African banks from 1 September 2020 to 31 August 2021. About 500 000 of these posts were analysed for sentiment and conversation themes, including the Treating Customers Fairly (TCF) outcomes.

Absa had highest overall Net Sentiment ranking in the index and FNB the lowest. The banking industry experienced more negative conversation than positive this year, resulting in an industry Net Sentiment of -7.5%.

The final sections of the Conduct Standard for Banks came into full effect on 3 July, and banks must demonstrate to the FSCA that they deliver on and report on the six TCF outcomes. This includes the banks’ behaviour and interactions with their customers on social media.

Among the final sections implemented from July is section 8, which relates to the regulation of customer complaints, which are increasingly being lodged on social media platforms, Kedibone Dikokwe, the FSCA’s divisional executive: Conduct of Business Supervision, said in the foreword to report on the index

“To ensure that the extent and nature of these social media complaints is accurately captured, the FSCA has adopted a data-driven strategy that includes the monitoring and identification of all TCF-related conversations on social media. In addition to gaining valuable insights on an industry-wide level, this strategy has proven useful in helping to highlight areas of concern among individual banks,” Dikokwe said.

“Considering BrandsEye’s Index finding that only 43% of priority tweets received a public response from banks, the FSCA is looking into this issue, while bringing it to the banking sector’s attention,” Dikokwe said.

However, the report noted that a large component of the 57% of priority tweets that did not receive a response were cases where customers did not tag the bank’s Twitter handle directly in their query and only referred to the bank by name. This means the banks would not be notified of the post, making the mention more difficult to find.

Less than half of priority tweets receive a response

Over the past year, banking customers posted about 188 649 priority Twitter posts, a 28% growth in service conversations. However, 57% went unanswered.

Nedbank, at 63.1%, performed far above the average response level of 43%, followed by African Bank, which responded to 61.1% of priority queries.

TymeBank was the least responsive bank, offering only 28% of priority tweets a response, followed by Absa at 30.2%.

Capitec Bank and Discovery Bank responded to similar proportions of their priority queries, at 36% and 35% respectively, although Capitec’s queue was significantly bigger than Discovery Bank’s.

FNB responded to 46% of priority queries, and Standard Bank to 44.3%.

Large number of complaints about Discovery Bank’s service

The index found that 50.1% of consumer mentions about the banks that expressed either positive or negative sentiment contained one of the six TCF outcomes.

Half of the banks in the index had between 49.5% and 52% of conversations speak to one of the TCF outcomes.

Discovery Bank had highest percentage of conversations related to TCF outcomes, 60.6%.

Other banks with above-average TCF-themed conversations were Capitec (50.8%), FNB (52%), Standard Bank (53.5%) and TymeBank (51.4%).

Discovery had a high volume of complaints about customer service (outcome 5b). Customers who reported being unable to cancel their products caused Discovery to have higher-than-average complaints about switching (outcome 6c), according to the index.

Instances of system outages caused Standard Bank to have a high proportion of complaints related to channel performance (outcome 5c).

TymeBank received nearly as many complaints about product performance as customer service and almost double as much as the rest of the industry. The speed of transfers to and from other banks caused frustration among customers. Transferring money between Everyday and GoalSave accounts also presented a challenge.

Capitec’s channel performance complaints were higher than Standard Bank’s, with instances of downtime presenting a considerable conduct risk. Capitec’s affordability was also questioned, with several authors saying that the bank’s individual transactional fees added up, despite low monthly fees.

Customers criticised FNB and Absa for having high fees. This is reflected in their higher-than-average conversation volumes about product design.

Accusations of racist business practices impacted FNB, as shown by the high volume of complaints about fair treatment (outcome 1a).

Most TCF-themed posts about performance and service

TCF outcome 5, performance and service, was the most notable conduct theme in customers’ conversations about the banking industry, accounting for 85.2% of market conduct conversations.

This outcome included conversations relating to customer service (36.7%), channel performance (20.9%), and product performance (14.2%).

These themes also had strong relationships, as about a fifth of mentions comprised at least two, if not all three, sub-categories, according to the report.

Although most of this conversation was negative, some customers said they had good experiences.

Absa and Nedbank both received the most positive feedback from customers, who said that staff helped to resolve their issues.

TymeBank and Standard Bank did not fare as well, with customers complaining about long turnaround times and attempting to contact the bank multiple times.

Product suitability (outcome 2), which focuses on product design, was the next greatest conduct theme (14.1%). Affordability was frequently discussed on social media. Customers who were let down by products or services criticised the value for money.

Claims, complaints and changes (outcome 6) was the third-largest conduct outcome (6.9%). It comprised nearly equal parts of conversations about accessibility and claims, complaint handling, and switching.

There was a strong relationship between accessibility and claims and complaint handling. Customers who failed to access their funds or make an insurance claim found themselves in drawn-out complaint-handling processes.

2 thoughts on “FSCA monitoring Twitter to measure banks’ TCF compliance

  1. I wonder what software BrandsEye uses to collect more than 2.7 million consumer social media posts about South African banks on Twitter?

    1. According to BrandsEye, they use real people to select randomly 565278 posts from the 2.7m posts, which are obtained using AI.

Comments are closed.