The main idea is that contracts include instruments that bind the government and a contractor to perform, even when the agreement isn’t finalized in a full formal contract yet. A letter contract fits this because it authorizes work to begin before the final contract terms are written, creating an immediate binding obligation to proceed. A job order fits because it is a specific instruction to deliver particular goods or services under an existing arrangement, which also establishes a binding obligation to perform. Grants and notices of award are financial assistance, not procurement contracts; cooperative agreements are also assistance instruments, not contracts. Blanket purchase agreements and imprest funds are procurement tools and cash-management means, not the type of binding contract instrument described here. So the items that are included in the definition of contracts in this context are letter contracts and job orders.

In a two-step bidding process, the first step is all about evaluating technical capability and how well a bidder’s approach meets the project requirements. It uses a technical analysis to determine which proposals are capable and compliant, without considering price. Only bidders that pass this technical screening move on to the second step, where financial proposals are opened and price is evaluated in combination with any remaining technical merit. This separation helps ensure that cost does not disqualify technically capable bidders and that the competition for price occurs only among those who meet the required technical standards.

The needed concept is that government contracts enforce a standard records-retention period to support audits and reviews of costs and compliance. The default requirement is to keep all contract records for three years after the final payment is made. The clock starts when final payment occurs, not when the contract is completed, because final payment marks the end of the contract’s financial obligations and the point from which post-closeout audits may reasonably arise. If final payment happens on a given date, retain the records for three years from that date unless the contract or a governing regulation specifies a longer period. Longer retention periods (five years from final payment or from contract completion) are not the default and only apply if the contract or regulation explicitly requires them.

Bundling should be justified only when market research shows measurably substantial benefits to the government. This means the decision to bundle is based on concrete evidence that consolidating requirements will create real value—such as lower life-cycle costs, reduced duplication, or better integration—rather than on guesswork. The emphasis is on substantial, demonstrable benefits rather than absolute guarantees. Reducing the number of vendors to one might be a consequence of bundling, but it isn’t the justification itself and can raise competition concerns. Likewise, assuming time savings will always occur or promising best value in every situation are not proper grounds for bundling; the justification must show substantial benefits, not guarantees.