Late last month, the Government Accountability Office overturned a $950 million contract award to a small Virginia firm to provide cloud services for the Defense Department. The deal was unusual not just for the extraordinarily high price tag, but for how it was handled. Instead of following traditional, competitive procurement rules, officials used something called a “production other transaction agreement” crafted by the Pentagon’s innovation outpost, the Defense Innovation Unit Experimental (DIUx). This essentially allowed the department to make the award based on the success of a prototype developed by REAN Cloud, thus bypassing the government’s expensive and time-consuming traditional procurement process. It was a bold move, and to many observers, it offered a far more efficient way to purchase technology—where traditional procurements can take years to play out, OTAs can be conducted in a matter of months.
Not surprisingly, some traditional federal contractors weren’t happy. One of them, Oracle America, filed a protest with GAO, and the watchdog agreed with the company that the Pentagon’s use of other transaction authority was improper. GAO’s decision was a blow to many officials desperate to reform the government’s antiquated procurement practices.
But it shouldn’t be.
In rejecting the DIUx agreement, GAO affirmed that its ability to hear OTA protests remains limited; it left the definition of OTA prototypes to the discretion of buyers; and it left the form of follow-on production agreements to the imaginations of their authors. Those are big wins in a case where GAO could have severely narrowed the parameters and limited the utility of this flexible and effective acquisition tool.
Some observers predict that the ruling will rein in the use of other transaction agreements, which are not covered by laws and regulations specific to federal procurement contracts.
Yet, the Oracle protest was sustained on a narrow (and easily correctable) issue, and GAO’s ruling actually reinforced the prototype OTA as a tool that is here to stay and is primed for expansion. The ruling solidifies the agility, viability, and value of OT agreements. Far from discouraging acquisition innovators, GAO’s decision should encourage these insurgents in their quest to field technology faster and balance risk with reward.
Congress has been notably gung-ho about OTAs, expanding their reach in each of the past three annual national defense authorization acts. Legislators have urged the Pentagon to apply the authority more broadly to speed purchases of innovative products and services, especially from nontraditional government suppliers that have been leery of cumbersome government buying rules and processes.
For those who’ve responded by experimenting with OTAs, GAO’s decision offers three encouraging clarifications:
First is that the GAO maintained its prior precedent, limiting its review to whether the agency’s use of its discretionary authority was proper. GAO further dismissed Oracle’s argument that the government was required to employ a Federal Acquisition Regulation-based procurement unless that option was not “feasible or suitable.” As a result, GAO’s jurisdiction remains limited, as recognized by the watchdog itself.
GAO’s confirmation of the broadest definition of “prototype,” is a second win, indicating that OT authority applies to an expansive portfolio of requirements. The agency’s prototype in this case was defined as a “repeatable process that highly automates the installation of these applications and the [operations] and maintenance of these applications down the road into a commercial cloud environment.” GAO found that that this effort was a proper prototype project. The decision acknowledged and recognized processes, pilots, tests, and demonstration of commercial products as valid prototype efforts under the statute.
Finally, GAO declined to apply a dictionary definition of “prototype,” finding that the agency guidance was published well in advance of the solicitation and the protester did not explain how the definition in the Defense Department Other Transactions Guide for Prototype Projects was improper, ambiguous, or should be disregarded in favor of another definition. GAO’s confirmation of the broadest descriptions of prototype is a big win for the application of other transaction authority to a greater portfolio of requirements across Defense.
With the expansion of OTAs in the 2016 National Defense Authorization Act to permit follow-on production, many in industry and government have questioned what parameters follow-on transaction users would adopt. In this case, DIUx constructed a production-OTA (P-OTA) to function similarly to an indefinite-delivery, indefinite-quantity procurement. The intention was to make the P-OTA available to other Defense entities. GAO did not discuss allowable P-OTA structures in Oracle, so it would appear that they can be wide-ranging and include something as broad as a multi-year multiple-agency ordering agreement.
So, in the wake of GAO’s ruling, organizations seeking to craft P-OTAs should take the liberty of defining what production they are seeking under the provisions of the statute and not limit themselves to pre-conceived notions of what P-OTAs should look like.
The fact that the Oracle protest of the P-OTA was ultimately sustained also should not overshadow the procurement lead time and competition successes of the overall effort. The level of competition received, and speed to production are unprecedented wins. Consider that DIUx was able to leverage its Commercial Solutions Opening (CSO) process to solicit 21 solution briefs for the prototyping effort within a 12-day window from March 10 to March 22, 2017. That is an astounding number of responses in such a short time. Adding to that accomplishment, acquisition officials were able to down select from those 21 to five vendors, receive panel pitches, complete all evaluations, and award an agreement by May 23, 2017. The total lead time from solicitation to award of the initial OTA was 75 days. In the traditional FAR setting, the competitive range determination alone to winnow 21 vendors to five would by itself dramatically exceed 75 days.
After a six-month prototyping performance period, the government set in place its follow-on P-OTA on Feb 2. That puts total lead time for this effort from solicitation to award of follow-on production at 11 months. In the traditional FAR context, the government would be three to four years into this effort before entering into a bid protest procedure. The flexibility of the OTA environment allowed the Government to find a quick solution, and even after the GAO protest decision, to have ample time to take action, or seek an alternate approach. The robust initial competition, flexibility, and speed of this effort, solidify the OTA mechanism as a winning tool for acquisition streamlining.
Benjamin McMartin is an attorney and serves as chief of the Acquisition Management Office for the U.S. Army Tank Automotive Research, Development and Engineering Center in Warren, Michigan. The views expressed here are solely those of the author, and do not necessarily reflect the views of the Army or the Defense Department or any other government entity.