The Downsides to Strategic Sourcing

Do Strategic Sourcing and Category Management Really Bring Best Value to the Government?

As the Chief of Acquisitions for the DoD Joint AI Center (JAIC), I was responsible for recommending to the JAIC Director the acquisition and procurement strategies needed to acquire AI technology for various DoD missions. I constantly struggled when choosing the most effective procurement vehicles for acquiring AI capabilities.  I knew we wanted something repeatable and fast. Multiple award ordering vehicles were an option (i.e. IDIQs, GWACS, BOAS, BPAs, etc.). Sometimes, these are referred to as best in class vehicles. They can be repeatable and fast (“can be” are the key words – the contracting office must properly utilize FAR 16.5, 8.4, etc.)

But… I also needed to ensure that the vehicles allowed for a growth in the U.S. commercial market. Our commodity was an emerging technology. I wanted to make sure the DoD dollars fostered an influx of new ideas and new companies. In short, I wanted maximized competition with little restriction, while maintaining the streamlined ordering provided by best-in-class vehicles. I needed flexibility. I needed vehicles that can expand and grow as technology expands and grows.  I needed to provide accessibility to start-ups and the upcoming inventors of winning ideas. Winning ideas that can protect our nation.

This is where the current application of ‘strategic sourcing’ and ‘category management’ created problems for me.

Common Procurement Methods for Strategic Sourcing/Category Management

First, let’s walk through some definitions, taken from GSA.gov:

Strategic sourcing is a “the structured and collaborative process of critically analyzing an organization’s spending patterns to better leverage its purchasing power, reduce costs, and improve overall performance.” (GSA.gov)

Category Management is the practice of buying common goods and services as an organized enterprise in order to improve the efficiency and effectiveness of acquisition activities. (GSA.gov)

To implement both of these initiatives, the Federal Government has recently gravitated toward an entire repository of multiple award ordering vehicles, each with a finite selection of awardees. Sometimes the pool of contractors is small, like 20 to 30. Sometimes the pool numbers in the hundreds or (very rarely) the thousands.  The idea is that by selecting a limited pool of highly qualified companies, very fast and orders can be competed at discounted rates.

The main intended beneficiary of these initiatives (and the application in procurement) is the American taxpayer. Better purchasing power for the government can further prevent frivolous and wasteful spending. Plus, the benefits to these models are evident: creating ordering vehicles with limited vendors can streamline acquisition lead times and provide discounted prices for requirements; and they allow agencies to prevent redundant spending by consolidating and tracking recurring needs among different offices. All very good things.

So, what’s the problem, you may ask?

My problem rested in its application to emerging technology, and how these ordering vehicles can restrict and unintentionally impact the attainment of new companies -- and new ideas that could counter the efforts of our foreign opponents.

In his book An Insider’s Guide to Winning Government Contracts, Joshua Frank warns his readers that “category management is the future of government acquisition” and highlights the large percentage of the commercial market that will consequently be “cut off” from competing on government requirements as a result of these best in class vehicles (Frank, Joshua P. “Chapter 7. Category Management.” An Insider’s Guide to Winning Government Contracts, 2nd ed. RSM Federal, 2020).

Side note – he outlines steps that companies can take to get ahead of this curve, I highly recommend this book for any company striving to win government contracts!

While consultants like Frank are helping companies prepare for these kinds of accessibility restrictions, I had to ask myself – as a Government professional – are these restrictions the best way to obtain best value? Should this limitation on competition be overlooked? How will limited competition affect the market? What is the long-term impact of restricted ordering vehicles? Will prevention of new companies prevent new ideas? Will this ultimately bring the best value for the government? Will it really improve performance? Or benefit the American taxpayer? Or the very real technological conflict against foreign adversaries?  

 “Can We” vs. “Should We”

In the JAIC, we learned some lessons the hard way. When we started, we did not have contracting authority, so we used other offices. Most were great offices and we built good relationships. But some of these great offices had rules that the contracting officers had to follow. One of their rules was the mandatory use of best in class vehicles if there were at least 2 companies within that limited pool who were able to compete. Our emerging technology needs often fit within the scope of these vehicles. So we used them, and we learned that they are great tools for some requirements. But not so great for others. 

We shifted from a ‘mandatory use of best in class’ model to a “can we vs. should we” model. Yes, we can compete this AI requirement among this pool of 20 companies. But should we? Should we instead open up the aperture of possibilities and see if we are missing something? 

This shift from mathematical procedural compliance (almost always localized) to critical thinking allowed us to embrace and better analyze the unique characteristics of the various procurement tools in the tool belt.

Current Techniques that Maximize Industry Competition

The good news is that there several procurement options that allow for streamlined orders and provide continuous opportunity and accessibility to new companies.  

One method is a Partnership Intermediary Agreement (PIA), an agreement using statutory Federal Lab Authority. Intended for technological transfer and commercialization of government prototypes, a third party Partnership Intermediary can facilitate large scale competitions with any willing company – traditional or not.  An example of this is the great work of Special Operations Command (SOCOM) and DEFENSEWERX are doing with SOFWERX in Tampa, Florida. Or the "Tech Grove" PIA between University of Central Florida (UCF) and Team Orlando DoD community. These can be fast, repeatable, and open to multiple new vendors.

Agencies can also use Other Transaction Authority (OTA) to create competition ecosystems similar to a PIA. JAIC/Chief Digital AI Office (CDAO) used this model when we released the Tradewind prototype environment (kudos to the JAIC acquisition team there on their excellent progress with this project). Again, streamlined ordering is feasible while opening up to a wide range of new and old industry players.

In both of these ordering environments, membership is free and always open. It takes about 10 minutes to join and participate in the competitions.

Finally, agencies can utilize the Commercial Solutions Opening, or CSO (championed by Defense Innovation Unit (DIU) and used by many other agencies, such as GSA FEDSIM and JAIC/CDAO). The CSO is a solicitation that allows for very streamlined source selection of commercial vendors.

There are more vehicles and methods than the examples I mentioned (and through interagency contracting, agencies that don’t have federal lab or OT authority can utilize them). All of these methods serve to strike a balance between streamlined procurement and maximum accessibility to new companies and new ideas.  There are no “ramp in” periods. No companies have to worry about missing the opportunity to join.   Large amounts of proposals (a consequence of greater competition!) can be managed through phased down-selects that start with a lower burden of entry for proposal submission (one page). The success of these procedures have already been demonstrated.

Conclusion

In the realm of emerging technologies, we need more industry players, not less.

It is true that Category Management is focused mainly on common goods and services, maybe not so much emerging tech. But, regardless of your mission and contracted services/products, everyone will soon be implementing emerging technologies into their requirements, products, services, and missions. Everyone will eventually need to redefine the meaning and application of “Strategic Sourcing” and “Category Management” principles.   And: our foreign adversaries are striving to harness this powerful tech at great speed and with large resources. Get ready, and good luck!

Mr. William (Will) Roe Roberts is Acquisition Lead for the DoD Chief Digital and Artificial Intelligence Office (CDAO) and was the first Chief of Acquisitions for the DoD Joint Artificial Intelligence Center (JAIC). Will is a member of the Florida Bar and has over 15 years of experience in Government Procurement. His notable work includes the creation of guides, trainings, and contract vehicles to assist the DoD in acquiring and delivering emerging technologies, such as artificial intelligence.