Leveraging on AI to improve Customer Service

For financial institutions to thrive in the current economic clime, customer service delivery must adequately align with customer demands and expectations. Beyond seeking lightning fast responses and resolution of complaints, customers are beginning to tend towards banks and payment solutions that incorporate AI into their CS process to ensure excellence in service delivery to customers and also enable them (the organization) make the ultimate switch to cognitive banking.

With the massive influx of MVPs and product lines in the financial services space in the past 2 years, it is becoming glaringly apparent that beefing up CS staff strength in proportion to the volume of interactions received will not be sustainable in the long run as running that model means that companies must be ready to build stadiums to accommodate their growing staff strength at some point. 

What companies need to start looking at is a solution that allows them rapidly conclude monotonous and repetitive tasks, provide cognitive banking services, and do so for up to a thousand customers simultaneously.

Consolidating the efforts of physical agents with AI and ML to support scaling ambitions will be the future of CS. With the present focus on financial services, AI will enable institutions effectively handle increasing interaction volumes, cater to dynamic customer needs and explore options that revolve around predicting the best value propositions for customers.

The need for human operators cannot be underplayed as some customers will still prefer the traditional methods of communication (email, telephone, SMS, walk-in etc) over chatbots to have their issues resolved. Over-reliance on AI is never a good idea and so the ideal mix of human intervention and AI must be concocted very carefully. Measures must be put in place to ensure that calls which can no longer be handled by AI are escalated to human operators. In handling email interactions, human operators will need to proofread responses drafted by AI until the system becomes foolproof. This will undoubtedly take some time as it will need to be trained with Nigerian/specific country peculiarities and contexts to ensure usability and relevance.

Added benefits to the financial services company that adopts AI will include reduced time and effort spent on training agents as AI requires one-off training, winning points on brand perception and maintaining consistent performance levels instead of the inevitable fluctuations that occur when a new set of agents are hurriedly recruited. Adopting AI also significantly lowers interaction abandonment rates, improves capacity for handling high traffic periods and ultimately ensures 24 hour real time support.

To properly launch into intuitive/cognitive banking services, AI will need to learn from transaction patterns, existing tickets logged on the organization’s CRM system, conversation histories, help content uploaded etc and leverage on these to provide spot-on recommendations for customers (after ensuring that their initial complaints/enquiries have been sorted out) 

To optimally drive these initiatives, special attention needs to be placed on the data which AI will be feeding off. This is because the more specific, detailed and viable the data, the more intelligent the system becomes. If properly executed, we may well be on our way to breaking ground in mimicking the human thought process in providing financial services.

*AI=artificial intelligence

*CS-customer service

*ML=machine learning

Managing customers according to type

“Customer is king”

“The customer is always right”

These two are probably the most popular phrases that companies incorporate as slogans especially during Customer Service Week. This seems fair enough as without these customers, they’ld be out of business. 

I work in FinTech and trust me when I say that this industry demands that you put your money where your mouth is. Here, the surefire way your customers take you seriously is when the efficacy of your customer service/complaints resolution team is second only to that of your Tech team. I believe that the best way to deliver on the customer service bit of this is by dealing with customers according to the stratum which you conclude that they belong to within the first 20 seconds of conversing with them. Remember that it’s not just about solving their problems, it’s also about how you go on about it.

In the next part of this post, I talk about the customer categories I have come up with so far and a little bit about how some highlighted traits help in better managing them and ensuring that their funds are safeguarded in the long run.


This customer thinks he knows all about bank processes and resolution timeframes but his info usually falls short of being accurate. He might have garnered his info through assumptions, misinterpretation or misinformation. It is also not uncommon for this customer to expect less stringency with documentation, I mean it’s a Digital Bank afterall. This customer usually does not mind being corrected as long as these corrections are not made in a rude or aggressive manner. 


These customers are extremely knowledgeable and are not shy to let you in on this within the first 30 seconds of speaking with them. They usually want exact details and specifics. They are very inquisitive. I’ve learnt that it’s best to treat them as “adults” as they are quick to take offence if they perceive that they are being babied or if you turn overly apologetic. You also have to be careful not to overpromise as they often come across as intimidating. 


Sometimes you wonder why this customer bothers with contacting customer service for resolution. He usually dwells on naming the higher-ups that are close friends instead of stating exactly what his complaint is. In dealing with them, it’s best to “just respect yourself”


It is not uncommon for this class of customers to volunteer their PINs, PANs, CVVs, passwords, OTPs etc when speaking with their bank. This occurs especially when they are reporting failed login, dispense error and card functionality issues. It is pertinent that this customer is properly educated on the dangers of divulging his details over virtual channels. You may also want to advise him to visit a branch when unsure of what info to give online or on a call. If not properly educated, they easily get themselves into the next category.


If you remember this message, “Dear customer, your ATM card has been deactivated due to BVN. Call Mr John on 080XXX” then you know this category. It’s quite fascinating how easily this SMS gets them to call these fraudulent individuals and provide their card details only for them to turn around and contact their bank 5 minutes later. Needless to say, these calls end in card hotlists. Fortunately, measures such as Card Control and periodic email and SMS blasts advising on non-disclosure of banking credentials seem to be reining in this group somewhat. 


This post will become a Long Form if we dwell on this. Just know that when your gut tells you something is off, something usually is.


This customer is usually this way out of desperation, delayed resolution or sheer impatience. Most times, it is the thought that someone is doing something differently or going above and beyond for them that helps.


Although rare, this does happen. It dawns on you 3 minutes into the conversation that this customer doesn’t have any enquiry or complaint. The dicey part is gently saying your goodbyes as these interactions always create an impression. Besides, cross-selling and upselling never hurt anyone. 

Always one for intelligent conversation, I’m definitely the customer that actually knows it all. 

What kind of customer are you? Let’s discuss in the Comments section.