Customer-centric innovation: How good are we really at understanding customer needs?

At Finthropology we are happy to see that customer-centricity is gaining traction in financial innovation.

Customer-centricity refers to the capacity of people inside a business to comprehend the circumstances, viewpoints and expectations of customers in order to generate customer pleasure, loyalty and advocacy. 

There is a large body of literature—both academic and from consultants—on how to build and sustain a customer-centric organisation. Less is written about how best to learn about customer preferences in finance. Customer-centricity is, however, often an important goal of digital transformation. Indeed, according to a recent Harvard Business Review survey it is the main goal. Many observers, including companies like EY and McKinsey, have highlighted the potential of advances in data technology to serve customers better.

Digital transformation is a broad agenda optimising processes, meeting increasingly digital customer expectations, embracing real-time data, and taking advantage of new technologies like artificial intelligence, machine learning, virtual reality, tokenisation and more. It also means coping with increased competition and changing business models, including the building of partnership in open banking structures. 

It is not surprising that the combination of digital customer expectations and increasing competition fosters customer focus. Neither is it surprising that the potential of new technologies in collecting, analysing and distributing data focuses the development of customer knowledge on the use of these—predominantly using quantitative data.   

But how good is quantitative data really at helping us be customer-centric? What do we gain and what do we miss when we take a quantitative approach?

Quantitative data  

Bankers and fintechs alike argue that the advanced use of artificial intelligence and machine learning allows them to collect far more customer data than before. A recent study from Italy analyzes how both traditional banks and fintechs combine transactional data (demographic data and product usage) with relational data (lifestyle, preferences and habits) and alternative data (from external sources like social media). 

New technologies also allow them to analyse data in new ways, including pattern recognition and combination with data from the customer context. This takes customer knowledge way further than the learnings obtained through occasional or recurring surveys.

Financial aggregators especially have the potential to analyse data across financial service providers, including information from all bank accounts as well as insurance and health care services. 

There is no doubt that the Big Tech industry has a great advantage when it comes to working on platforms with deep knowledge of customer shopping patterns and preferences. And a very recent example from Klarna shows that more can play this game. Klarna is plugging in ChatGPT to provide shopping advice to consumers buying through their platform.

The use of quantitative techniques opens new perspectives on how to engage in co-creation with customers to adjust and develop products and services. Combining quantitative analysis with open customer fora, analysis of countless comments (and complaints!), beta testing and digital dialogue all point to a better understanding of customer perspectives and needs. The Italian study mentioned above finds that the most advanced companies (both banks and Fintechs) use this type of collaboration.

Quantitative data collection, however, has several drawbacks. First, it most often focuses on existing customers—although customer data can of course be compared to broader consumer surveys that are becoming more usual. Among existing customers it is likely that more active customers will dominate insights. This could suppress insights on diversity, equity and inclusion agendas. 

As we have shown in Finthropology research, women still feel condescended by their bank advisors and prefer to ask friends or relations rather than to go to the bank. Many women also feel that they need to be qualified in financial management and investments to take part in the financial system. They feel that they should learn more about stocks and market to invest their savings. They also feel that they should be better at managing spending. Why not just accept that some things can be left to professionals? 

In 2020 Oliver Wyman estimated that financial service providers miss out on USD 700 billion in revenue per year because they don’t meet the needs of female customers. And quantitative data doesn’t really help us to understand them.

A second drawback is that while automatic data analysis captures the actual use and decisions of customers, it misses a lot of context on what drives decisions and what consumers are actually doing behind the scenes. For example, data analysis doesn’t show that customers frequently use what we call “workarounds” to make financial tools meet their needs. 

In several of our studies in the Global South we have seen examples of people sharing their credit cards. One of our interviewees in our payments study in Brazil explained that her cousin would help her get a computer with a discount. He would pay for it with his credit card paying a lower interest rate as he had a better credit rating, plus he could earn points by shopping. She would later refund him in cash. Another example could be downloading a new payment app, because of having forgotten to bring a purse.

Finally, data analysis may interpret data poorly or incorrectly, or fail to answer the question of why customers do what they do. A very classic example is this from the use of artificial intelligence in fraud detection: Should a sudden high spending on a credit card be seen as caused by a thief in which case the owner will be happy to receive an alert—or could it be caused by the owner buying an expensive gift for a lover, in which case an alert might offend. 

Qualitative data

At Finthropology we base most of our studies on listening to and observing people. Granted, our insights aren’t based on statistics. Neither do they result in statistical insights like what percentage of customers value safety over convenience or vice versa. But our methods often reveal important insights into customer preferences and pain points. Where quantitative studies can answer questions on how and how many, qualitative studies are great for answering why.

We mostly collect qualitative data through personal interviews and observation. Beginning with a number of broad questions can often uncover great stories, which in turn can generate answers to quite specific questions. This means that we collect extensive data about the context of people’s lives as well as answers to the kinds of questions that might be in a survey. 

When we ask about money and payments, we often start with questions about how the interviewee learned about money, how well they feel they are managing their spending, and how they would like to feel about money and spending. This brings out interesting stories and background information on later decisions about choice of bank, money apps and payment solutions. 

When researching payments in particular we often use a “portable kit” method where we ask people to show us the things they carry with them on a daily basis when away from home. We specifically ask about the money-related items they use daily: cash, cards, receipts, apps on their phones, and so on. This method is a brilliant way to gain insights into how people actually manage their finances in a real-life context.

Apart from learning about interviewee’s conscious decisions, we often experience that interviewees start reflecting on their practices and discover that what they prefer they do and what they actually do are not the same. One interviewee in our Swiss study of women and money told us: “You know, I really prefer to carry only my phone and to pay with the apps that I have installed… However, quite often I carry a credit card in my purse for shopping, because the internet connection in the shops is sometimes poor.” Our research has uncovered countless customer pain points like these—some of which might be quite easy to solve. 

Interestingly, digital transformation is affecting how qualitative data collection is carried out. It is possible to do personal interviews via digital media, like we have done for a number of studies during the pandemic (such as our study of women and payments in Switzerland). We found that this remote method worked pretty well, with the exception that interviews tended to be shorter. This was partly because we made fewer observations about the physical context and possibly because we asked fewer questions about items in the bag or the mobile apps on the phone. 

Another option is the study of digital fora, known as “digital ethnography”. We have published a report on open and anonymous user groups of Buy Now, Pay Later (BNPL) solutions that provided the interesting insight that customers may not be as loyal to providers and their merchants as expected. User group participants shared many good tips on which BNPL solutions best to be used for specific purposes. 

Granted, this approach has some of the same drawbacks as quantitative studies in involving only existing and active customers. It could, however, be a good way to pilot a study.

Choosing the best way to learn

Coming back to the starting point of this blog, we are first of all happy that customers are now increasingly at the centre of development and innovation in finance. We would invite financial services organisations to consider what kind of methodology will give them the best knowledge to work with in their quest to meet customer needs. 

We see great potential in exploring how to combine different methods, as we are doing in a study of digital finance in Laos and Cambodia. Qualitative and quantitative approaches both have their pros and cons, so why not use the best of both? This also has the potential to unite knowledge from different parts of an organisation or a partnership ecosystem.