architecture – FINthinkers

Towards Canary Banking

Services offered by banks are mainly commodities sold with rather opaque pricing strategy. Client experience has become the main differentiator as underlying services are largely the same. The challenge has become creating outcomes which integrate into the life of the customer via respective touchpoints of the client’s eco system. This becomes difficult as the number of potential touchpoints increases significantly (ubiquitous computing).

The focus within banks has moved from operations and backend infrastructure to the touchpoints. Such focus leads to the shift in investments – which leave the backends under-invested hidden from the user by layers of newer systems. This trend started years back when the financial services industry realized that the internet is not a transient effect and changed the way business is fundamentally done. The adoption accelerated with the rise of mobile and the next catalyst with the increase usage of conversational technologies like chat or voice.

Adding more layers does not improve the overall system – instead only to delay and cover up the increasing issues and complexities of the fragile legacy backend which can no longer be sustainable (Measuring complexity in networks). From an opertional perspective the whole architecture becomes more and more unmanagable and risky highlighting the need for a substantial refactoring or more likely a total re-architecting.
Let’s for the moment assume that you own such a “backend” which becomes increasingly hard to change and difficult to operate. What could be a way out of this situation?

Be bold and start greenfield. This allows to get rid of the layers of legacy and build a system which matches current needs.
Try to be clever and refactor the system in an incremental approach. You have to be faster then the changes in the environment as you have to catch up and eliminate debts while adding the functions required to stay competitive. Implementation risk and sunk cost avoidance are the typical arguments backing this option.

Most organizations will go for the second option and try hard. They neither have strong leadership and capability(s) to aim for the first option nor the skills to actually build something new based on current state practices. The organization and the management are only trained to do small changes and the “internal immune system” reacts immediatly against any major change which could endanger the status quo.

Maybe it is better to start with a canary banking approach – where the next version of the bank is built up while the first is operational.

It is similar to driving a car – you choose one, drive it, maintain it and at some point you switch to another model. The new model may contain some parts already used before – that’s fine as long as they fit into the new model. The approach is rather different than continuously adding/ changing external cosmetics elements to an outdated Ford-T backend.

Measuring complexity in networks

People like simplicity. They prefer linear systems where an input translates in a well understandable way into an output or result.

The world is a mesh, a complex network of elements which influence each other. Small effects on one side may lead to a chain reaction in the network and cause a big change.

Many organizations have big difficulties to deal with the increasing complexity which is inherent to the shaping mesh economy. Established rules and patterns become less effective or even contra productive. It has become difficult to stay lean while things change fast – typically organizations and their supporting infrastructures grow continuously amalgamating approaches and technologies over many years.

The typical reaction is to launch some form of a simplification program. Such programs require good KPI’s to understand the benefits of the measures taken. But meaningful KPI’s are hard to find.

Organizations typically start counting elements as a measure and aim to reduce complexity by reducing the number of similar elements. Reducing e.g. the number of applications in use feels right – but it may come with side effects. This measure may decrease agility. Agility is required in order to deal with complexity. It may also lead to less productivity as broadly used applications may make it harder to deal with complicated situations where specialized tools provide a better support.

All problems are different and a best practice only fits to a specific set of problems. One approach which I find useful to measure complexity in the network is based on a heuristic which goes back to Robert Glass.

“For every 25 percent increase in problem complexity, there is a one 100 percent increase in solution complexity”

Let’s make a thought experiment. Let’s assume we want to measure complexity of a system with a set of functions which must interact with each other in a defined way.

One extreme is to have all functions in a single block – we have now a complicated block which hides all interdependencies between functions from the outside world.
The other extreme is to have a block for each function and then connect the blocks based on the functional dependencies. Now each block is simple but all interdependencies have been externalized which makes the network complicated.

This heuristic above means:

If we have 4 functions in a block and add another to it, then the solution complexity of this block doubles.
If we add an additional connection to 4 existing connections between blocks, the complexity of the network doubles.

The heuristic allows to find a setup where the functional bundling into nodes and network connections are in a good balance. Given this heuristic it becomes possible to start measuring and optimizing the complexity of networks. Obviously not all functions are equal neither are all connections. Such factors can be included into a more detailed approach which allows measuring and optimizing an application landscape towards a low complexity.

The most decisive factor may finally however be the solution approach and the clarity of the design. If there is a simpler way to solve the problem then this is the first step to do. And this leads back to the starting point – in order to truly simplify first the problem space and the solution approach need to be challenged. They change continuously.

Emotional trust in an hyperconnected world

Trust is an often used term in financial services. What is trust, how is it built, gained or lost? Has trust building changed in the last few years of emerging digital hyper connectivity and will this have an impact on banking? (see “Banking evolution: Service Innovation,” “No Off Switch“) Let’s together explore this a little bit …

There are at least four dimensions of trust:

predictability – ability to predict actions of others and situations which might occur
vulnerability – giving others the chance to take advantage of vulnerabilities
value exchange – exchange of values even though there is no full knowledge about the peer
delayed reciprocity – giving something now with the expectations to be compensated at some future point

The trustor has logical and emotional expectations against the trustee. The logical expectations are often contract related. In case of a loan a payback including interests is a logical expectation. The emotional expectations include the level of comfort and the experiences made during the time where the loan is granted and beyond.

The following little example explores these dimensions. Let’s assume you want to make a ride home and you call a cab.

An ordinary cab arrives with a smiling driver. Before you enter the cab you need to trust the driver that he knows the place, has serviced the car properly and will not crash the car while you are in. This quick assessment is nothing simple but humans have developed senses during the evolution which support this interpersonal check.
The cab arrives – but nobody is in. There is a screen showing a friendly face in an office telling you that he is your driver. The cab is remote controlled in a way that it feels for the driver like being in the car. You can again perform the quick assessment described above based on the reduced amount of information and available senses.
A self driving car arrives with a smiling man in it. He has been mandated by law to sit in the car to intervene in critical situations. You may be tempted to make the quick assessment as in the first scenario but then you notice that this person has limited chance to intervene and influence the sequence of events in an emergency situation as the available time to react would be to short. In essence you notice that you need to trust the system, its sensors and the algorithms.
A self driving car arrives – completly empty. That’s a different story – the interpersonal element and the usual base for quick asseementis is completly missing. Maybe you should do a short ride first to see if this is safe and then, once you gain confidence into the car, its sensors and algorithms go for longer trip. With good experience, trust is built.

There are futher factors influencing your final emotional assessment – the taxi could be dirty, the driver may have an unpleasant driving style or the climate control may be broken. Even when the target is reached on time, the experience may not great and you may decide not to rely on the services of this company again.

I guess it is rather clear what follows now. All these situations also occur in financial services today. The chance that you meet a banker which is an entrepreneur and personally engages in the trust relationship with you are rare. Such a banker would stand up with his name for the agreement made and would do the best to meet the logical and emotional expectations.

So let’s explore the other three scenarios in a little bit more detail.

The secenarios have all one thing in common – the ‘driver’ has limited skin in the game.
Remote meetings with specialists who can come up with creative solutions for complex problems are quite the norm in business and personal live today. Finding the right specialist may already be a challenge and arranging a physical meeting may be close to impossible.
You may have an assigned employee representing the bank as a sales clerk or relationship manager. The relationship manager will talk with you and then key in the data into some engine which finally processes the agreed business. You may build up a personal relationship to your relationship manager. If this is strong, then you will be tempted to follow him if he moves to another bank. If you trust more the brand, its system and processes, then you will stay and engage with a new relationship manager.

You may also be routed to a customer services desk which is used to deal with requests like the one you have. With each call you get to know another person – building an interpersonal relation is not intended.
You may also interact through an electronic channel with the system. A hopefully cool user interface guides you through the necessary steps to get things done.

The objective of most companies is to operate with standard processes leaving the relationship manager very limited flexibility. Hyper connectivity leads to more transparency, the logical element of the trust relationship is performed by an engine and the emotional one is more and more an outcome of the digital experience.

Trust is still a key element in many things. In banking the logical element of trust is more defined by processes, algorithms and infrastructure while the emotional aspect becomes more and more a result of a great digital engagement.

Trust is shifting from personal relationships to systems and experience.

Multipurpose Traverse

As today’s business challenges span across boundaries within and external so too must leadership. The ever-increasing complexity of today’s world calls for a critical transformation in leadership from managing and protecting boundaries to boundary spanning ( see Never fail to fail, Giving Direction, Dance on the VUCAno) With that it’s business model reflects towards a multipurpose traverse offerings supporting the client’s dynamic behaviors and journeys ( Banking evolution: Service Innovation, Banking Today)

Under the context of digital offering(s) is its simplicity of a single-purpose business model/ offering/ app the wave of the future?

WeChat, or Weixin in Mandarin, is quickly becoming one of the most popular multi-purpose platforms, not just in China, but the world. Released in 2011 by Chinese internet giant Tencent, With nearly 800 million active monthly users, its user base has grown consistently in every single quarter to date. More importantly the point that I would like to focus is it’s actual embodiment of the app.

It’s safe to say that the most ardent of technophiles have at least 100 apps on their smartphone e.g. Facebook Messenger, WhatsApp, Telegram, Skype, Google Hangouts and Duo for instant messaging. Uber, Lyft, Citymapper, Waze, Tripadvisor, AirBnB and Skyscanner for directions/maps. In addition for gastronomy related: Deliveroo, Just Eat, OpenTable, Zomato, Yelp or Urbanspoon. That’s 19 apps to cover three essential functions. WeChat includes capabilities above and more.

WeChat lets users do everything you’d expect it to – instant messaging, sharing life events and chatting to family members. But its feature list extends far beyond custom emojis and profile pictures. WeChat allows you to arrange a catch-up with a friend, pre-order food from a restaurant, book a taxi to the restaurant, get directions on foot, pay for the meal (or split amongst your friends at the time of payment), check movie times and book tickets, and also purchase other items. All without hitting the home button.

The possibilities for brand-to-consumer engagement on WeChat are almost unparalleled anywhere else in the world, and this is almost entirely due to the way the app manifests itself in as many aspects of daily life as possible. By knowing a person’s current location and when they usually have dinner, all in one app, fast-food brands can hyper-accurately target consumers when they’re most inclined to purchase. And by tapping into the app’s data on payments and money transfers, marketers can get a good idea of when, where, how and why users spend their money, before using this to hyper-accurately target their audience when they’re most likely to buy. With such understanding of a client’s behaviour enables to proactively provide financial wealth services be it from suggesting dynamic relevant payment methods to making recommended investments, wealth management and advisory, etc…

The need for banks to traverse beyond its current boundary is imperative to regain expediency with the new paradigms ( see Digital Tur Tur).

Ubiquitous Computing

The term is not at all a new trend or technology. Previously known as pervasive computing where due to technological advancement and cost feasibility the trend of embedding computational capabilities into everyday objects. This makes them effective in communication as they are network interconnected and performing activities of the end users without a centralised system.

Ubiquitous computing integrates via different devices, industries, environments, applications (e.g. wearable devices, appliances, fleet management, sensors). The goal of it is to make devices “smart” in the form of creating a sensor network capable of collecting, processing and sending data via the context and activity that it is under.

We had seen first phases of such capability involving wireless communication and networking technologies, mobile devices, and RFID tags. With the exponential advancement in internet capabilities, usage of voice recognition and artificial intelligence, the growth and adoption of embedding ubiquitous computing significantly increases now often associated and known to be the internet of things (IOT)

Gartner predicts approximately 8 billion connected objects to be use by the end of 2017 and it appears to be growing. In order to cope with the growth of IOT a heavy incorporation of artificial intelligence (AI) fueled autonomy will be required. An AI-driven era of IOT becomes the key building block to herald an increasingly seamless experience and hyperconnectivity as users and their digital counterparts concurrently transpose from one medium/device to another, between multiple environments, the physical and digital ecosystem.

Conway’s Law – Change or Fail?

We were recently discussing Conway’s Law in the context of the ongoing transition towards solutions which are digital in their core. The law is named after programmer Melvin Conway who first introduced it in 1967 and said: ‘ organizations which design systems … are constrained to produce designs which are copies of the communication structures of these organizations.’ If Conway is right, then this means that implementing a new inherently digital solution requires organizations to change structures or at least their ways to communicate. Sounds like change is inevitable …