23. May 2016

Data isn’t valuable. Information is!

Data isn’t valuable. Information is!<br/>

Banks spend a vast amount of time researching and collecting data about clients, but often lack the bigger picture of connecting these separate data piles from various systems. Data alone is worthless, but connected and turned into information using an identity database, new possibilities such as reducing the cost per client, increasing quality of service and anticipating a client's actions are possible.

Let me introduce you to Urs. He is a member of the board at a private bank called Bank Lake Zurich. His main goal is to increase the amount of clients handled by his relationship managers, which means reducing costs. He realizes that relationship managers spend a major share of their time researching which client to contact next and what he or she might be interested in. Urs’ new best friend Patrik, the CIO, figures that this problem can only be solved by a data driven approach.

Patrik states that over the years, the bank has collected a big pile of data consisting of transactions, position information, tax statements, but also CRM, click streams, phone calls, emails, written letters and so on. Each pile alone is next to worthless (except the banking stuff of course:-) ). Patrik thinks out loud that once you start to connect all the piles, you’d be able to figure out the answers to questions like:

  • How often was the client in contact with the bank? How did the activity change over time? Is it trending up- or downwards?
  • What is the client’s opinion towards the bank and how does this influence the amount and volume of trades he is performing? How much new money has he brought?
  • Is the client at risk of leaving the bank? 

Urs thinks that this is great news and exactly what he always wanted for his relationship mangers. So he wants to know more about how this might work.

The key: connecting data from various systems to one client using an identity database

In a perfect world, every system would have the same identifier to classify a client. However, the world isn’t perfect. It’s not only not perfect, but a lot of interesting sources additionally don’t have a unique qualifier! Therefore, something more sophisticated is needed to connect each system.

Every person has many identifying properties, for example a name, CRM ID, an account number, etc.

Meet Hans, a typical bank client:

To identify Hans in any system, we can use any of his known identifying properties.

So let’s apply this to an email archive. The bank received the email below and it contains three properties we already know belong to Hans. This means we can - with a fairly high probability – conclude that this email was sent by Hans. Going through all existing data and connecting it to a certain client is what we call crawling.

But wait, the email offers even more information! It contains an email address. Email addresses are a pretty good identifier. In fact, much better than just a first and a last name. The email “h.m@mail.com” is added to the pool of known identifiers. This is called the learning aspect. It can be digitized, which means that you don’t need to manually maintain the identity database! The possibilities of the learning process are infinite.

EXAMPLE: Assume Hans browses your web site for quite some time, all his actions are logged, but they are of no use to you because he has never logged in. One day, Hans sends an email using the contact form. At this point, you are able to trace back every click Hans ever did! Why? Because the contact email will contain the browser identifier, which is then used to match the clickstream history.

Sounds incredibly difficult and expensive?

Urs wants to stay realistic and thinks of the cost of such a solution. But Patrik, the CIO, tells him that he already has a working proof of concept in place, which was developed by ti&m using their “Analytics” product. Patrik invites Urs to see for himself and to stop by ti&m in Zurich for another “Lab-Visit” - for more inspiration and to sharpen the gut-feeling of what is possible.

Legal / compliance / reputational remarks

The next argument is about compliance: Firms in general and banks specifically need to keep the legal/compliance aspects as well as the reputational risks in mind. Data privacy laws allow the use of internal data to profile and classify clients. They also give any client the right to demand to know what data is stored about them, how it is used and to request a deletion. The bank privacy laws guarantee the privacy of the client’s bank relation to any 3rd party. As the data is only used internally, there is no violation of banking regulations.

Knowing what data is available about each client and proactively as well as transparently making it available to the client can lead to a reputation gain! Thus, the client feels in power of what the company knows. Nevertheless, personalized pricing and similar actions are very sensitive and the public has not yet accepted them. Aggregating all data in one place is a high risk. It can be mitigated by making the aggregate anonymous or enforcing technical as well as organizational security.


Urs is convinced. Having such a system in place will give him the full 360° view of his clients. This allows him to reduce the cost per client due to less human time being invested. On top of that, his relationship managers may increase quality of service as they can anticipate the client’s actions. And these are just a few of the possibilities. By using big data technologies and brain power one can come up with amazing opportunities to reduce costs and grow business.

Please get in touch – come by for a Lab Visit at ti&m!

Martin Fabini
Martin Fabini

Martin Fabini has worked in IT for over 20 years. At ti&m, he introduces clients to new business cases with new technologies.

Related Articles

Digitalisierung total - die Vernetzung der Dinge
Digitalisierung total - die Vernetzung der Dinge

Editorial: 1 Milliarde PCs, 5 Milliarden Smartphones und geschätzt zwi­schen 50 und 100 oder mehr Milliarden vernetzte Dinge, das sind die prognostizierten Zahlen per 2020. Die neue Ausgabe vom ti&m special beleuchtet, inwiefern sich reale und virtuelle Welt immer mehr vermischen und welchen Einfluss FinTech-Unternehmen und Start-ups dabei haben.

find more information
mind reader 750x410
Are you a mind reader?

Are millennial men the right target group for a Pampers marketing campaign? Of course - if they have children. Anticipating what the customer wants is like mind reading. This is now possible by analyzing customer behavior and situations.

find more information
Google Anthos – The Cloud Platform of the Future

Niemand Geringeres als Urs Hölzle hat das Anthos-Projekt von Google in Zürich vorgestellt. Anthos soll nichts weniger als zum Linux für die Cloud werden. Unser Senior Software-Ingenieur Fabian Camenzind war dabei und gibt einen Einblick in die wesentlichen Features sowie die Vor- und Nachteile von Anthos.

find more information
Lazy Angular: Writing Scalable AngularJS Apps
Lazy Angular: Writing Scalable AngularJS Apps

There are two major issues I have faced in the past few years, when writing AngularJS applications, and I have seen numerous other teams fighting the same battles. Out of these experiences the “Lazy Angular” approach came to life. It gives us a project structure which works for both, large and small applications. And it enables us to keep a somewhat consistent load time as new features come to life and our app grows.

find more information
Autonome Postautos: unterwegs mit der Zukunft

Autonome Fahrzeuge // Die autonomen Postautos haben keinen Fahrer und können dank ihrer leistungsfähigen Sensoren problemlos navigieren. Zum ersten Mal testet ein Unternehmen diese Technologie in der Schweiz im öffentlichen Raum.

find more information
Setup eines Maven-basierten Multi-Modul-Build incl. RCP- (OSGI-) Komponenten<br/>
Setup eines Maven-basierten Multi-Modul-Build incl. RCP- (OSGI-) Komponenten

find more information