Platforms are pretty buzzy. The term is all over banking and technology, especially for core back office functions in risk and finance. Nearly every Tom, Dick, and Susie in vendor space uses the term “platform.”
But not all platforms are created equal, and some don’t even qualify as such. This isn’t just semantics; this definition matters. Platforms have powerful features that allow financial institutions to do things better, faster, and cheaper than ever before; the latest generation takes it to another level.
How can one tell a real platform? There are three distinguishing characteristics for a platform.
1. A platform is architected to be broad, flexible, and adaptable.
Adaptability is crucial in risk and finance applications. There have been a myriad of new demands over the past decade, and most financial institutions have struggled to keep up.
A real platform changes the eternal buy versus build question to “buy and build.” Yes, there’s an embedded assumption in this that financial institutions don’t build platforms. There are always exceptions, but in general the pressures facing internal IT departments are simply too great for them to build a real platform. Platforms require significant investments in architecture and time in order to build components that can be used generically. Somewhat ironically, most internal IT departments are too busy managing the myriad of applications they own and reacting to today’s change to be able to strategically build a foundation that can adapt quickly to tomorrow’s change. Often, this leads to shortcuts being taken; a point solution is identified as being substantially similar to a new challenge and attempts are made to try to adapt it for the new challenge. This is a good way of spending a lot of time and money, but not generally a good way of solving a problem.
This is not to say all vendors are competent platform builders, nor is it saying that all who claim to have a platform actually do. Much of what is called a platform is actually a collection of point solutions cobbled together. Beware, especially if these point solutions were built by separate companies and “seamlessly integrated” by another well afterwards. System design matters, and needs to be considered from the very beginning.
A platform is built of components that allow for the creation of higher order software services that can be reused for different things across the enterprise. Rule engines, cashflow engines, workflow processing tools, and so on come together in a way that allows for customization as a core design principle. Describing what a platform does can be tricky because it can often do many things. Thus a true platform is strategic, not tactical.
Here’s an important rule for identifying a platform -- if it only consists of one component it does not meet the definition. A database is a database, a rules engine is a rules engine, a workflow processing tool is a workflow processing tool. None of these are platforms; putting them together in an architected way creating software services (e.g., cashflow, accounting, loan master) that are exposed and work with each other makes it a platform. Many banks in their finance transformation projects try to piece together a bunch of specialized components into a platform of sorts. This is Challenging with a capital C, not unlike Luke Skywalker hitting the Death Star’s exhaust port, but then he -- unlike most IT departments -- had access to the Force. The challenge in transformation projects is because most of the components were not designed to work together; they also need to integrate with point solutions that were not designed to integrate with anything.
The platform must offer an ability to extend existing functionality without going through an entirely new build process. A system built without the ability to adapt with new features, workstreams, and data elements will lead users to add manual processes or IT departments to build surrounding point solutions. Both of these are expensive, risky, and inefficient. The economics of a platform are quite appealing. Having hundreds or thousands of users leveraging the same bit of code saves money. But many institutions have specialized ways of doing things. A platform reconciles the two views -- the core components are bought, while the specialized ways of doing things are custom built using the core components.
2. A modern platform’s data model is domain aware, open, and extendable.
The power of a modern platform begins with the data model. Getting it wrong spells failure. The failure may not be doomsday; it may be a soft failure of lack of adoption and proliferation of external systems and applications to deal with the weakness, but it will be failure nonetheless. There’s no point in investing -- and yes, a platform is an investment in that it will reap benefits -- if it’s necessary to build an entire set of point solutions around it in order to make it work.
A platform handles data from a wide variety of sources. The architecture must support multiple avenues of data flow both in and out. If it doesn’t, we are most likely looking at a point system.
For a risk and finance platform that must support multiple workstreams, the data model must be domain aware. A technology-centric, generic data model will not work. The data model must include all attributes of the instruments it stores, organized in a coherent way. These include credit attributes (both from origination and those that vary over time), contractual attributes, reporting attributes, and accounting attributes.
But the right data model, though necessary, is not sufficient. Thus a data warehouse -- even the mythical one singing the Siren song of addressing nearly every challenge -- is not a platform. If an institution has a good data warehouse, it can feed a platform; if it doesn’t, the data model inside a platform can limit the institution’s need to build one. A key element of a platform is the ability to both capture the base data but also allow the proper manipulation and use of it to provide multiple views on the same data. Trying to put this functionality into the data warehouse has been the cause of many a project failure.
The openness of the data model is key; no data should be trapped in the system -- all data should be transparent, well understood, and transferable to other areas or systems. Modern platforms can be very powerful, but they don’t do everything!
3. Modern platforms leverage modern technology to do things better, faster, and cheaper.
Behind every buzzword lies a fundamentally new methodology. They are better. The new generation of platforms leverages these advances, using a cloud-based infrastructure to deliver open and adaptable solutions that are fundamentally different compared to traditional installs or hosting.
In general, the economics of a platform are appealing, and there is an intuitive sense in not building components that every institution needs. For example, one does not see banks writing database software or servicing systems anymore, but before the vendor space was mature it was common.
The economics of the latest generation of platforms are even more appealing. The SaaS delivery model means that banks don’t have to worry about large server farms (which depreciate faster than cars) or costly upgrades or massive implementations when a regulation or business practice changes. The software constantly gets better. Bugs are constantly fixed. New features are constantly added. Long gone are the expensive three-year upgrade and refresh cycles. Business users no longer have to wait to get improved features and performance, they just come. All the headaches of efficiently operating and constantly improving the software falls onto the vendor, specialized in the area.
For new, general challenges, the platform will change to address it -- that’s one of the commitments most good SaaS vendors are willing to make. For bank-specific challenges, generally new rules, new reference data, new disclosures, or a new process can be incorporated relatively painlessly -- all leveraging existing components and software services.
Today’s real platforms -- with domain-specific data models, well-architected integration of powerful components, and modern delivery models -- can fundamentally change the way an institution performs key functions, allowing them to adapt rapidly to change and do things better, faster, and cheaper than ever before.
John Lankenau is the head of valuation and accounting product solutions at Primatics Financial. He has extensive consulting and financial services industry experience, with an emphasis on complex loan systems integrating risk and finance. Mr. Lankenau is a notable thought ... View Full Bio