How to Run a Deal Desk Process

What does a deal desk actually do?

A deal desk applies consistent pricing and terms rules to non-standard deals fast enough that the rules do not cost you the deal. It is a speed function with a governance mandate, in that order.

Teams get this backwards. They stand up a desk to control discounting, staff it with approvers, and never define a response time. Reps learn the desk is where deals go to wait, so they escalate directly to a VP instead. Within two quarters the desk reviews a small share of exceptions while the rest of the discounting happens over Slack.

Design the desk around three commitments: a published routing rule so reps know when to use it, a complete intake so reviewers never chase information, and a response SLA the desk is measured against.

Which deals should route to the desk?

Route by trigger, not by deal size. A 200,000 dollar deal at list price with standard terms needs no review. A 15,000 dollar deal with a custom indemnity clause does. Publish the table where reps work rather than in a policy document. If the share of deals hitting the desk exceeds roughly a third of volume, the problem is your standard configuration, not rep behavior. Widen the standard band and the desk gets its time back.

What belongs in the intake request?

Enough for a reviewer to decide without asking a single follow-up question. Incomplete intake is the largest source of deal desk delay.

Require these fields before a request enters the queue:

- Account, opportunity, and current close date - Requested pricing against list, with the discount percentage calculated - Competitive situation, including who else the buyer is evaluating - Business justification for the exception, stated as what changes if it is denied - Contract term, ramp schedule, and payment terms - Any prior exception granted to the same account

The justification field carries the most weight. "The customer asked for it" is not a justification. "The buyer has a competing quote at this price and the close date is inside the quarter" is one, and it tells finance what they are actually trading.

Keep the form short enough that reps complete it correctly the first time. Every field beyond these six raises the return rate, and a returned request costs a full day in a quarter-end week. Where a field can be populated from the CRM, populate it automatically instead of asking the rep to retype it, because a manually entered discount percentage is a number two people will later disagree about.

How does the deal desk feed the forecast?

Every approved exception is forecast information, and most teams throw it away. Discount approvals predict what deals will close for, and end-of-quarter volume predicts how the quarter is being manufactured.

Two connections matter.

The first is deal value. Most deals close for less than the amount sitting in the CRM, and the gap can be large. A pipeline carrying an average deal size of 80,000 dollars while closed-won deals average 40,000 dollars is a forecast that will miss even when every deal closes. Feed approved discount data back into how you value open pipeline so the forecast reflects the price the market pays rather than the price the rep entered. A weighted pipeline built on unadjusted amounts inherits the whole gap.

The second is pull-forward cost. Deals pulled from future periods to rescue a current number usually arrive with discounting attached, which means you are borrowing revenue from next quarter at a price. The desk sees that trade before anyone else does. Report it monthly: how many approved exceptions were tied to deals with close dates moved earlier, and what the aggregate discount cost was.

What rules keep the desk from becoming a bottleneck?

Four, and each removes a category of waiting.

- Approval thresholds, not approval chains. One reviewer per dimension. Adding a second finance approver above a certain value doubles the wait and rarely changes the answer. - Standing pre-approvals. Publish a band that managers can approve alone. Anything inside it never touches the desk. - A named backup for every reviewer. Vacation should not stall a quarter. - Auto-expiry on approvals. An approved exception is valid for thirty days. Beyond that the deal is re-reviewed, because a price approved for a deal that has since slipped is priced against conditions that no longer apply.

The expiry rule does double duty. It catches the deal slippage signal that matters most, because a close date change is the strongest predictor that a deal is in trouble, and a deal that slips from one quarter to the next is less likely to close even when it stays in commit.

How do you measure the deal desk?

Four metrics, reviewed monthly, none of which is request volume. Track average discount by segment over four quarters rather than month to month. A rising discount trend at flat volume usually means competitive pricing pressure entered the market, which is exactly the kind of change that breaks a forecast built on last year's assumptions. The desk is often the first function in the company to see it, so route that observation into the monthly revenue review instead of leaving it in a queue report. It belongs alongside the pipeline coverage picture rather than underneath it.