Banks for data – are we there yet?

While IT has been my profession, Economics has been a deep area of interest (to the extent of doing a post-graduation in it!). Sometimes I start thinking of ideas from one and apply it to the other. I have been thinking about this one for over five years now. As months and years have gone by my belief in it has become stronger. I thought it is now time to put it out there and get some reaction. Banks have a framework for helping us secure and manage our money. Is there an equivalent for enterprises to deal with their data? And is the timing for such a service about right? This is the key question I want to explore through this write-up and the reactions.

First some basics from the world of Economics. Financial intermediaries such as banks and insurance companies are absolutely essential ingredients to the success of an economy. They serve several economic functions. I am outlining the key ones here, especially the ones for banks. (1) Security and accounting (2) Payment mechanisms and Liquidity (3) Lowering transaction costs and (4) Reducing information asymmetry between depositors and loan takers. Let me elaborate briefly each of these.

1. Banks are an obvious example for the safekeeping of money in accounts. They also help in accounting for it.
2. Liquidity refers to how easily and cheaply an asset can be converted to a means of payment. Banks make is easy to transform various assets into a means of payment through ATMs, checking accounts, debit cards, etc. They facilitate the payment mechanisms to accomplish what we need.
3. Lowering transaction costs – Banks provide the ability to withdraw money from ATMs, write checks all at low transaction costs primarily due to economies of scale.
4. Borrowers know much more about their likelihood of success than potential lenders and investors. This asymmetric information causes one group with better information to use this advantage at the expense of the less-informed group. Banks help resolve this asymmetry (let’s not talk about 2008 again!).

Now let us see if there is such a need when it comes to enterprises and their data (think of data as money).

1. Enterprises need to secure their data from data breaches and disasters. They also need to maintain "accounts" of their data - what data was produced, when it was produced, when it was consumed and by who.
2. Enterprises also have “data liquidity” needs. This means there are volume and flow differences in which the data comes in and how it is consumed. The volume is it comes in is very "transactional", while it is aggregated and sliced/diced as it is consumed (in that sense opposite of money!). It could also come from internal and external sources.
3. The data gets consumed by many stakeholders – operations, management, other internal systems, external third party applications, external data aggregators/consumers. The transaction costs of these can be very high (guess how many times you called the data team to produce a report).
4. Finally as the world is becoming connected and complex, there is an absolute need to share and consume information. There is always an information asymmetry between consumers and suppliers. An enterprise can be both a consumer or a supplier. There is an absolute need to enable risk profiling on the data being consumed.

Thus, in my opinion, there is definitely a strong need for “banking-like” services for your data. The next question that comes to mind - what can potentially enable and unlock this? What are they key drivers? To me the key drivers are (1) cloud computing maturity, (2) big data technologies (especially the notion of schema-on-read), (3) security frameworks and emerging technologies in this space, (4) connectivity and (5) last but not least the most important – “Trust” in outsourcing (no, not love!).

1. Cloud computing technologies are now mainstream. With the big boys now fully committed to this space and with their quality of service and compliance, the cloud is something which enterprises are beginning to trust more than their internal infrastructure teams.
2. Big data technologies. I know this is a subject of much ridicule if one is talking at fifty thousand feet levels. But having successfully implemented some key initiatives in this space, I can say with confidence that not only one can get significant performance improvements, data variety, scalability and redundancy but also “read what you like” at run-time. As you can guess I am a huge fan of schema-on-read paradigm and it can fundamentally change the way we think of warehouses.
3. Maturity in security frameworks like HIPAA, FISMA, PCI DSS and others ensure that you get compliance/coverage from various perspectives. Also this technology space is hot including encryption-on-rest technologies which provide you the last level of security protection for your data.
4. Connectivity and costs associated with it, are dropping fast. The reliability of connectivity is no longer a subject of discussions. This goes to show that there is an underlying acceptance of this. Can it handle enterprise volumes across continents, you bet!
5. Finally the above four factors will start leading to “Trust” in outsourcing. Outsourcing has been in play since the 90’s. But it has been limited to back-office. Can I outsource my warehouse to a “bank” who can maintain, manage and provide that data back to me and external parties. It is the “Trust” that needs to be created for this to happen. This is a more softer issue, but with strong investments in this space and good branding/marketing this will come.

In conclusion, I think the idea is at a state where I feel it is worth a discussion. Do I think it will be a reality? Yes, absolutely. The timing is the only question!