A Bankers Guide to Implementing Personalization

What is ‘Personalization’?

Personalization is about sending the right information, to the right person, at the right time.  Whilst banking personalization has long been touted as the future of financial services,  and is considered a critical step in differentiating the digital proposition, it still hasn't gotten the attention it deserves from incumbents or challenger banks. In fact, many bankers are still unaware of what it really means. In the context of messaging or notifications, when banks try to be everything to everyone, they end up delivering it to no one.

The ‘segment of one’ has long been the ambition of innovative bankers and marketeers for years, but few have the understanding, imagination and the know-how on what it takes to deliver the concept, particularly when it’s ‘digital’. What’s more, with many non-banks like Amazon and Google delivering convincing personalized experiences today’s customers are ultra-savvy—if a gesture is insincere or not applicable, they’ll see right through it.

Whilst most banks are likely grateful for their customers, they frequently need to remind themselves to act honestly, and avoid using a tone of voice that oversells their products or services or sounds like something that their brand, would never say. Today, if you want any chance of elevating the engagement of an existing or potential customer, it’s not enough to create one message that speaks to an entire segment. Banking personalization requires laser-focus precision—as if you’re engaging to the exact financial needs of each individual customer at the precise time and place that they need that advice. Let’s be clear, that is very difficult.

A Practical Example of Personalization

Deriving likenesses to “real world” examples of ‘digital’ customer engagement is a useful way to highlight just how detached and impersonal banking can be. For example, imagine walking into a bank branch and being greeted by a customer service representative upon entering. Rather than asking what you would like to do in the branch, or whether they can assist you with the right service or advice, the meeter-greeter launches into a sales pitch of the bank’s recently launched products. Whilst it’s entirely possible that one of those products might be exactly what you were looking for, it’s far more likely the bank staff member lost your attention the moment it became obvious that their spiel wasn’t relevant to you. Likewise, with digital, where attention span is at a premium, a bank will never have a chance of truly engaging customers if they treat them like a catch-all “lead,” rather than treating you like … well, you.

The Role of Technology

Despite a raft of freely available, open source artificial intelligence tools, and an abundance of data it would seem that very few bankers know how and where to execute personalization. The purpose of this article is to provide some advice on the foundation they should be building, and a few use cases to inspire some creativity. 

Personal Finance Management (PFM)

PFM was the first mainstream form of personalization introduced to banking consumers, but little else has transpired in the years since. Understanding your spending habits and causing a behavioral shift was a novel idea 5 years ago, but in reality, was just a simple analytics task of categorizing spend categories from the merchant codes of your acquirer’s transactions and reporting them in layman’s terms for customers to understand.  The problem with PFM is that it doesn’t tell you what to do, just indicates that you should probably do something. Arguably, some of the financial aggregators like Credit Karma are taking things to the next level, offering actionable insights on how to improve your credit score and then subsequently recommending better products and solutions than the ones you have currently got. Whilst this is effective at a macro level of ensuring you optimize your choice of financial service providers, it offers little value into your day to day lifestyle needs. It’s these needs where bank’s can offer true value and personalization.

Creating your Customer Experience Database (CXDB)

The first thing banks will need to do is to build a Customer Experience Database (CXDB). I say “build” because there isn’t a product out there today that does this even remotely well, and if you want something that is truly unique to your organization’s strengths then you are going to need to build it not buy off-the-shelf. The CXDB will be a work in progress entity, constantly evolving and improving over time, but importantly will give the bank a baseline for its personalization journey’s and will naturally increase in complexity as the organization culture adapts to becoming a dynamic, personalized service provider. 

CXDB v 1.0 “Moments of Truth”

The initial version of your CXDB will probably be a simple rules-based engine in which the bank will identify actionable tasks and the corresponding data triggers for each task. A simple If-This-Then-That (IFTTT) tool. The easiest actionable tasks to identify for your first version of the CXDB tend to be the scenarios that are triggered at “moments of truth”.

Moments of Truth, a term first coined in 2005 by A.G. Lafley Chairman, President & CEO of Procter & Gamble is the moment when your customer interacts with either your brand, product or service to form or change an impression about that particular brand, product or service. Over time there has evolved three moments of truth. The first moment of truth occurs when a customer is first confronted with the product, taking place either offline or online. This typically occurs through the part of your brain called the posterior cingulate cortex which decides within the first 3-7 seconds of a consumer encountering the product and it is during this time that you have the capability of turning a browser into a buyer (I have written about this before here, in the bankers guide to digital onboarding). Consider this to be the "moment a consumer chooses your product over your competitors”. The second moment of truth happens after fulfillment, when a customer is acquired and experiences the quality of the product or service as per your brand’s promise. Finally, the third moment of truth represents the overall impression you have made on your customer and garners their feedback or reaction towards your brand, product or service. The ultimate reward being that your customer becomes a brand advocate and promotes you via word of mouth or through their social media channels.

In Financial Services, these Moments of truth are events that do not happen often, but when they do, represent a real opportunity to sell, upsell or cross sell a product or service, or amend an existing service that is being provided. Below is a typical list of such “moments”:

Each of the above events can lead to a contextual ‘conversation’ between the bank and the customer, providing the bank has acquired the necessary data identified to trigger the conversation. That data is the bread and butter of today’s data scientists and product owners who will be well versed on what that data should be and the sources where it can be acquired from. As one would expect Data definition and Data lineage are core attributes of the CXDB.

CXDB v2.0 “Everyday Banking”

The next phase in the evolution of your CXDB can be a focus on more everyday events. Each of these events will have data points that trigger a propensity rule that corresponds to a service or product you can offer in real-time, or useful advice that will be valued by your customer and reinforce trust. For example, lets assume you are a bank that has a credit card with a range of travel-related benefits e.g. airmiles, lounge access, multi-currency FX, hotel points etc. Each of the following detectable data points could infer a real-time conversation to the customer about how to use a feature of the product, or possibly present a cross-sell opportunity:

In another case, where for instance Travel Insurance is a complimentary service to the credit card, it is possible to extend the data points even further to trigger a policy quote at precisely the right time and location:

“Show me the money baby” – Data is the New Money

As you may have come to work-out, acquiring the data, and ensuring it is clean, structured and with integrity intact, is the crucial component in the start of the personalization journey. Of course, much of the data points can be ascertained through clever analysis of the raw transaction data, but this tends to usually be post facto and the real magic happens when data is acquired in real-time and acted upon from people’s most personal belonging – their cell-phone/mobile.

For a moment, let’s look at what information can be acquired from a standard mobile phone and how that data can be used to provide a personalized service. When combined with a bank’s customer profile data and readily available AI tools, there are several personalization notifications that can be inferred. A mobile phone can tell you the date and time of the day, the location where you are standing, how you got to that location (drive, walk, train etc.), what’s specific to that location (home, office, beach, mall, restaurant etc.). Combine this with what you understand about the customer through their spend analysis and other mobile phone actions like taking a photo of someone or something you can offer advice that is laser focused and timely. Below are a few notification examples of what is possible :

“It’s lunchtime and three of your friends ate at Jim’s Diner this week (1 minute away) for under $7.00, which would allow you to save $3.00 towards your holiday fund. There’s also free parking just around the corner on Main Street.”

“That’s a picture of the new 3 series BWM which costs $32,450 or an auto loan payment of $400 per month which you are already approved for. Would you like more info?”

“You are in Peak Mountain Ski Shop, and we wanted to let you know that your Platinum Credit card qualifies you for a 15% discount there, and includes free travel insurance in February. You also have enough airmiles to fly to Aspen.”

“Nice picture, but FYI you can’t afford new shoes this month, your annual car insurance renewal is due on the 20th”

How much is too far? – The Personalization Privacy Paradox

With so much incentive to become a truly AI driven, hyper-personalized business and break through the noise in the market, there is a danger that too much insight into the personal affairs of your customers can trigger the “spider-sense” in everyone -- frankly freaking them out, and making them less likely to engage with you. We saw a perfect example of this in recent months through some of the advertisements from Revolut in the UK which triggered a backlash from consumers who found the message to be a step too far. The campaign was not much different to that which the popular music streaming site, Spotify does, but music is music, and banking is banking not music!

Revolut’s Data Scientists may have been clever, but did not consider their customers reaction.

Enter the Privacy Paradox, another term coined to describe the manner in which customers today are split between their desire for hyper-personalized advice and their natural instinct to protect their personal details. A major reason for the paradox is simply because many consumers don’t really grasp what the term ‘personal data’ comprises, and what snippets of information financial organizations can pull together to create customer personas. Obviously, bank details, addresses and passport numbers are all clearly sensitive data that should be actively protected, but research has shown that most actions that consumers take online are easily manipulated to uncover sensitive information that customers think is secure.

Conclusion

Firms have been monitoring customer behavior and purchasing relationships for decades from the first supermarket loyalty card schemes to the internet’s earliest tray of cookies. The techniques and mobile technologies over the past ten years are simply acting as a catalyst and evolving, but ever more savvy consumers are becoming cognizant that if a bank really wants to control their data, then they could, whether or not they should is another matter. When trust in banks flounders then consumers always perceive the worst.

Personalization in banking will fast become an art form, rather than a science practice. Building the perfect CXDB will take time, but in the journey to create the perfect financial assistant for your customers do not be so dazzled by the depth of information that modern technology can unveil, and over egg the pudding – remember its quality not quantity of personalized advice that wins the heart over.

If banks are too succeed in creating an awesome personalized experience, they must exercise transparency and practice restraint when acquiring data. My advice when building the CXDB parameters for your bank would be to only seek that data that will elevate the experience of the service you are offering. Often the age-old advice is still the best. Less is more, and if customers sense that a financial institution knows too much about them, they probably do.

Useful Resources:

Techie references for various Personalization Tools:

https://azure.microsoft.com/en-us/services/cognitive-services/personalizer/

https://stackshare.io/stackups/amazon-personalize-vs-google-ai-platform

https://techcrunch.com/2019/06/10/aws-is-now-making-amazon-personalize-available-to-all-customers/