Guest post from Kurt Peckman, Program Director, IBM Business Analytics
Follow Kurt on Twitter: @KurtPeckman
Below are excerpts from my interview with Cliff Clavin, a former United States Postal Carrier and a regular customer at Cheers – a neighborhood pub located in Boston, Mass., where everyone knows your name.
No better person to discuss the Fiscal Cliff than an actual Cliff who is familiar with all matters.
Following was our discussion over coffee and egg sandwiches about how decision management is an integral part of Mr. Clavin’s United States Postal Service (USPS) job, as well as the pending Fiscal Cliff:
Kurt Peckman (KP): Thank you for agreeing to answer some questions about the Fiscal Cliff over coffee.
Cliff Clavin (CC): Ah, my “plesh-ah.” [in a thick Boston accent]
KP: Before asking questions related to decision making and the Fiscal Cliff, tell me again how the media ripped you off?
CC: In my years of walking routes for the USPS I had a lot of time to contemplate monetary policy. During my lunch breaks I would give brown bag talks about the corporate tax rate and quantitative easing – and long before “QE” was cool. During one of these talks, I got into a heated debate with a young Alan Greenspan, former Chairman of the Federal Reserve. As a result of winning that debate, I earned the nickname “Fiscal Cliff” from my postal colleagues. Until recently, I was the only “Fiscal Cliff” around.
KP: I’ve heard you did a lot of thinking about the best components of a quality decision management system while walking your routes. Tell me about those components.
CC: Constraint-based optimization, business rules and predictive analytics are a must. Automate the whole thing, and make the system such that people – and systems – on the front line can use the service and manage good decisions.
KP: Seems simple. Can you give some examples? Does the USPS use any of these elements?
CC: The USPS uses optimization to develop its delivery routes. It uses rules applied to postage rates based on size, shape, weight, etc. And, it makes use of predictive analytics to anticipate volume and plan rate increases – especially around high traffic times like the holidays. Admittedly, the USPS was slow to embrace everyone moving to email. But I tried to warn them. I once presented a talk to the Postmaster General entitled “Stamping out the Internet” alongside lobbyists from the pen pal and envelope industries.
KP: So, how does all of this relate to the current “Fiscal Cliff?”
CC: Typically, the ultimate goal for a decision management system is to make high volume decisions operational – especially when the impact of those decisions are intended to support a specific corporate strategy. In the USPS example I mentioned, consider the piece of mail coming in that is initially classified as “very hard” to deliver? How should this piece of mail be routed internally? What rules apply? Is postage due? Is the address legible, missing or wrong? What resources can and should the post office allocate to “help” deliver such a piece of mail? What’s the chance of successfully delivering a piece based on the “very hard to deliver” attributes of the piece?
KP: What does a missing ZIP Code have to do with the Fiscal Cliff, Cliff?
CC: The current situation tagged as the “Fiscal Cliff” is largely about developing rules right now – and these rules have optimization and predictive analytics serving as inputs.
KP: For example?
CC: We see rules in every story about the Fiscal Cliff: threshold level for the debt ceiling; if tax bracket x then tax increase y; Program A should be cut by a, and Program B should be cut by b; and on, and on. Developing these rules requires classic constraint-based optimization and predictive intelligence as inputs.
CC: [sigh] I think I might have better luck explaining this to IBM Watson. How can the Federal Government maximize cost reductions via cuts in programs while minimizing the impact (e.g., loss of jobs) from those cuts? The “formula” for that involves constraints and the predictive effects as gleamed from historical trends and data.
KP: When does all of this become operational?
CC: That’s my point – at this stage it’s purely about rules, policy and governance. Too early to tell.
KP: Should I be worried?
CC: On the contrary, stop to consider the mathematical beauty of Fiscal Cliff negotiations. Classic two-person, zero-sum game theory. I’m writing a paper on it now – my peer reviewers want me to call it “A Game Theoretic Approach to Fiscal Cliff Negotiations,” but I will likely publish it under the title, “Recipe for Fiscal Chicken.”
KP: What other projects are you working on now?
CC: After studying the transition of postal mail to email to text, I’m working on applying technology & automation to decision making.
KP: Oooooh. Tell me more!
CC: I can’t now, but I will in about a week on this blog. Let’s just say I’ve helped develop a cloud-based analytics solution that many organizations around the world (especially the North Pole) can leverage this holiday season to dramatically improve their gift distribution decision-making.
KP: Fair enough. I’ll call you next week for details.
CC: Please do. Waiter, can you make sure he gets the check?
For more information:
· Download a product review of IBM Analytical Decision Management from industry analyst James Taylor
· Watch a short video on IBM Analytical Decision Management
**Cheers and Cliff Clavin are copyright Charles/Burrows/Charles Productions and Paramount Television.