Let me start with this: I get it.
NASA funding is about so much more than space exploration. It’s about expanding the human experience. It’s also political, strategic and quite expensive. The advantage of spending money to boldly go where no man has gone before is not just about exploiting the great unknown but the advances that those of us held down by gravity can enjoy day to day.
NASA has changed the world as they continue their mission to leave it. The smoke detector that alerts your family of a deadly fire is compliments of NASA. Both the CT scan and MRI were greatly advanced by NASA research. Fire-resistant fabrics, your cordless drill and pens that write upside down all owe their success, in part, to space exploration. There are real benefits to discovering how we can put a rover on Mars. I get it, and I’m all in. Warp eight to the first starbase.
What I don’t get is, in a time when budget cuts and waning interest have grounded manned flights into space, how NASA can spend $14 billion dollars on an enterprise-wide procurement system. That’s billion with a ‘b’. Houston, we have another problem.
Forget space travel, let’s talk about time travel. I recently wrote about some epic IT failures, and I wish I could jump in the DeLorean, zip forward a few years to see how this pans out then pop back in time to include it on that list. I’m that confident this procurement system will bust, even if it’s implemented on-time and on-budget. But alas, NASA is yet to invent the flux capacitor so we’ll just have to look at what they got from their last enterprise-wide solution, which cost $5.6 billion in 2007. It focused on hardware and storage systems and allowed NASA to access products from 4,200 manufacturers managed by 42 vendors. NASA reports that they process 1,100 orders a month. That means that in a year about 13,200 orders are processed. So since 2007, close to 80,000 orders have been processed. Divide the cost by the number of transactions and that equals $70,000 per transaction. And the system needs to be replaced faster than my family car.
Automation is designed to make work get done faster -- an advantage that’s spurred the fear that technology will evolve to be so efficient that it will replace the need for people all together. I’d argue that at $70,000 per transaction, the reverse may be true. We could hire 1,100 people, pay them $70,000 a year (which is a decent government salary no matter where you live) and all they would have to do is process one order a month. As long as that one order is processed in a timely manner, we save nearly $200 million per year.
I don’t want to pile on to NASA. First, I still want to live on the moon. Second, they are not alone in the struggle to find proficient ways to procure the goods and services they need. But if procurement isn’t efficient at $70,000 per transaction, I’m worried about the new system. I worry about every new IT system because it seems whatever issue we hope to fix is only getting a new flashy coating. A fresh chrome finish to a kinked up, twisted pipe. In the end, we spend a lot, we do a lot, but we don’t get a lot.
Procurement is close to my heart. The first real project that I facilitated in state government was a procurement project for a medium-sized department. Ironically, it started as an IT project to develop a tracking tool that would allow employees to see where in the process their order was. It was a priority for the department since managers were frustrated over having to call several offices to track down order forms and move them along. The average time to order products ranging from office supplies to vehicles had grown to more than four months -- longer for IT purchases. If orders changed from “I’d like to have a new _______” to “We need ______ to work,” then you had to track it down and walk it through.
The IT portion of this was easy. We mapped the process, knew which people had to do what in which order and how long it took at each desk. We could electronically route the form, allow for electronic approvals, and monitor each order as it went through the system. Easy and not too expensive. People were able to look up their order form, and, as a bonus with the electronic routing all forms should have moved through the system faster. Ten days faster.
Ten days was the time the forms typically spent in the hands of mailroom staff moving through the building. We cut that out. But what the electronic system did not address were the seven various levels of approval needed on each form. It did not address the time it sits on a manager’s desk awaiting Thursday afternoon when they process all the orders that came in that week. It did not address the numerous handoffs. While it did make it much harder to lose an order, it did not tackle the issue of what happens when Approver 7 disagrees with Approver 1 and they have to talk about what was being requested.
I remember sitting with the team, just having heard the report from IT that this project could either wait at least a year for in-house developers to work on it or we could outsource it for around $300,000. ROI, ten days and a little less frustrated agency heads -- hard to justify.
Staring at the process map that literally looked like a maze of pipes, we went to work fixing the things IT could not fix. We went back and looked at why we had so many approvals and where we were batching orders. We started to address the root causes of why orders took so long. We started removing and consolidating steps in the process. We worked to build integrity and quality into the first steps so we could reduce the number of approvals and soon we had a process map that shrunk fourteen handoffs to four. We could process over 90 percent of the orders in two days or less, and it cost us nothing.
I’m not well-versed in how NASA does procurement. Maybe their process needs a $14 billion system to manage. Maybe for that $14 billion, they save a trillion. I’m not suggesting that you throw smart people in a room, say, “This is what we have, how do we make it work?” and you’ll get phenomenal results using just a sock and a box to procure rocket parts. But maybe you would. Maybe the rush to a new IT solution isn’t going to get us to Mars any sooner -- just as the rush to automate our processes isn’t getting our clients benefits any sooner, or shrinking the lines at the DMV, or filling pot holes. IT isn’t the solution to all our problems, and at that price tag, it shouldn’t take a rocket scientist to see we may want to postpone liftoff.