From Pipes to Platforms

As we look at fraud detection systems for financial services, it gets very interesting when we consider what has emerged. Here is a snapshot of what we typically see:

  • New fraud trends routinely pop up. (Think phishing, bust-out, takeover, new account, etc..)
  • When the problem becomes big enough, the institution seeks out a solution to identify this fraud. This is typically a point solution that solves for that specific problem.
  • As new types of fraud emerge, the next point solution is deployed.
  • Many of these solutions are dependent on data and analytics.
  • Over time, we begin to see a maze of overlap with respect to data and functionality.
  • Suddenly, we find ourselves with a highly complex system.
  • In addition, our cost to administer and manage the growing list of vendors increases.

So how do we end the madness? How do we meet the business needs most optimally?

How do stop the fraud, anticipate the next fraud type and not significantly increase the costs of creating such a complex system?

I think the answer lies in both the data and the technology choice. On the data front, we routinely see clients with many data solutions all of which overlap highly. It is not uncommon for us to see clients with 5-10 separate point solutions or more. They have ID Insight for this, Lexis for that, Experian for this doo-hickey, FICO for that gidget, etc…

However – when you back away from the whiteboard, we suddenly realize that each solution is bringing overlapping data to a separate, specific solution. Whether it be credit bureau information, utility records, etc., we see that these solutions are utilizing very similar or even the same third party databases. This should not really be a suprise, as most solution providers are all going to the same root data.

The end result is a system that connects to many very specific point solutions. The typical solution is very rigid. That is, “our solution looks like this and this is how you connect”. Suddenly, we have 8 or 10 rigid pipes, all of which share many of the same components.

So back to ending the madness. In a Utopian world, I would have all the data I need at my disposal in one pipe, and this pipe would not be rigid, but flexible tubing with connections to bring in even more information and data. From a solution side, this utpoian view would also allow me to use the raw data and information to solve for any future fraud types that emerge.

So while today my problem may be takeover, it might be something entirely different next week. Rather than have to go out and find another rigid pipe – why not go to my flexible pipe and just ask to solve for the problem? Under this view, I suddenly realize the enormous benefits:

  • I don’t have to through an RFP to identify the “best” point solution.
  • I don’t have to wait for months (and sometimes years) to solve. I can turn it on immediately and stop the bleeding.
  • I can also turn certain features off allowing me to reduce cost.
  • I don’t have to manage another vendor and bring another through the Security hurdle.
  • I get to sleep at night realizing if something new emerges, I am now equipped to stop it quickly.

We see this Utopian view emerging over the next few years. The madness cannot continue and the various consumer, cost, security and compliance pressures are forcing this.

What we see is moving from the world of rigid pipes to what is becoming known as “Platform as a Service”. Rather than point solutions, we will move to data and analytic platforms that are flexible and able to react very quickly.

Here are the high level compents of what we see in the future Fraud Platform.

  1. Will have the most comprehensive set of data to solve for 90% or more of all existing and anticipate fraud problems.
  2. The ability to quickly bring in more data as needed.
  3. The ability to ask for any data desired for whatever reason.
  4. The ability to customize your treatment of the data (eg. I want a tight or loose match on address, while I want an exact match on Social security number).
  5. Have multiple modules or components that can be turned on or off as desired. Eg. a bust-out module, a custom score, a takeover module, etc..
  6. An easy to use web-anabled dashboard that allows the client to quickly and efficiently enable data, solutions, modules, etc…

We recently launched the next version of CompleteID and in doing so, enabled users to do many of these things. Customize scores, alter rules, changing matching criteria, customize data sources, etc.

This is already resulting in many new customers. While solving for the immediate problem is necessary, clients realize the power of being able to have a flexible system that can react quickly.

Stay tuned as we continue to remove the rigid pipes and move to the Platform.


Date Posted: June 4, 2010 Author: Category:   IDI Blog

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload the CAPTCHA.