When I worked in a bank from 2010 to 2015, the large tech companies mostly kept out of retail banking. Recently, Apple has added a credit card to their portfolio, similar to Amazon. Facebook is working on its own currency, and last week Google announced checking accounts to be launched next year. Margins in traditional banking have been under pressure for some time now, driven by central bank policies. Why is Google entering this business now of all times?
Thanks to its mobile operating system Android and its apps, such as Google Maps or Google Search, the tech giant already knows our preferences and our locations. But what Google only knows to a limited extent is what we really spend our money on. Consumer habits are an often overlooked dataset to predict what products and services users will buy in certain regions.
But if these records are so valuable, why does Google work with Citibank and not set up the checking account project itself?
This shows us two things in particular:
Google has little interest in the banking sector, which has been heavily regulated since the financial crisis of 2008. In biology, a host is an organism that supplies resources to an alien organism. Citibank is providing itself as a host and takes over the regulatory tasks for Google.
It also shows that the value of data has to be determined contextually.
The quantity and quality of a dataset can represent a different value for different companies. Google can link data about consumer habits to other data. Data can diminish or even lose its value if it is not (further) supplemented and processed. Older data on account transactions, for example, can sometimes have little to no value at all. However, if this data is linked to (real-time) location data and search engine queries of the user, for example, it can have a high value.
To give you a more vivid example: Imagine a bank that has customer X who bought an Audi A6 six years ago and paid by bank transfer. Now imagine that Google also receives this data about customer X, as well as the data that X googled “Mercedes S 560” three days ago and was at a Mercedes-Benz dealership two days ago according to his location. He has also added an appointment at this location for the day after tomorrow in his calendar.
While this one dataset about the account turnover is almost worthless for the bank, by linking several datasets it can be concluded that it is about time for a new car (according to the official depreciation table, the usage period of a car is six years). Customer X has already done research about a specific model, both online and offline, and the appointment the day after tomorrow will probably be a test drive.
Therefore, the same dataset has a different value for different participants, here the bank and Google (as well as its advertisers).
When Google offers checking accounts, it doesn’t want money from its customers, it wants their data.