How long does it take to build a deal desk?

The first working version — a charter and an approval matrix — takes a day or two to draft and a week or two to socialize. Building the judgment behind it (codifying your real discount norms and term standards from past deals) is the part that takes months if done by hand.

Who should own the deal desk?

One accountable person. In companies under ~$50M ARR that's usually a RevOps or Finance leader wearing the deal-desk hat, or a fractional operator brought in from outside. Larger companies build a dedicated team, often under RevOps or Finance.

What's the difference between a deal desk and CPQ?

CPQ assembles the quote and enforces pre-approved pricing — it's plumbing. A deal desk judges whether a non-standard deal is actually good given your history. You can run a deal desk with or without CPQ; they solve different problems.

How much discount authority should reps have?

Enough to move fast on standard deals without a review — commonly up to 10–15%. Past that, route to the desk. The exact number depends on your margins; the point is to make it explicit rather than case-by-case.

Do we need deal desk software to start?

No. Start with a charter, an approval matrix, and one owner. Add software when review volume is high enough that routing and tracking become the bottleneck — not before.

How to Build a Deal Desk (Step by Step)

Most deal desks aren't built on purpose. They accrete — a discount approval here, a Slack thread there — until one day a bad deal forces the question: who actually owns this? Building one deliberately is not a big project. It's a definition, a document, and a decision rule. Here's how to stand one up without hiring a team or buying software first.

Start with the trigger, not the tooling

The instinct is to buy a deal-desk product. Skip that for now. A deal desk is a decision function, and you can run the first version of it with a document and one accountable person. Tools help once the volume is real; they don't create the judgment. Build the function first, then decide if you need software.

Step 1 — Define what counts as "non-standard"

A deal desk only touches deals that fall outside your standard motion. So the first job is to draw that line. A deal should route to the desk when it hits any of these:

Discount depth beyond a set threshold (e.g., anything past 15%).
Non-standard terms — unusual payment schedules (net-60/90), custom SLAs, auto-renewal changes, non-standard liability or termination language.
Deal size above a dollar threshold where a mistake actually hurts.
Custom commercial structure — multi-year ramps, usage commitments, pilots-to-paid.

Everything below the line stays self-serve for reps. Everything above it gets a review. Writing this line down is 80% of the value — it tells the team what's normal and what needs a second set of eyes.

Step 2 — Write the deal desk charter

The charter is a one-page document that is your deal desk in its first version. It states, in plain language:

What triggers a review (Step 1).
Who approves what (Step 3).
How fast a review happens (Step 4).
What "good" looks like — your real discount ranges by segment, your standard terms, and the concessions you will and won't make.

This is the same artifact we call a Deal Decision Charter at Precedent. It's the difference between guardrails that live in one person's head and guardrails the whole team can see.

Step 3 — Build the approval matrix

The approval matrix is the heart of the desk: it maps discount depth (and term risk) to who can approve it and how quickly. A common starting ladder for B2B SaaS looks like this — adapt the numbers to your margins:

Discount / risk level	Approver	Turnaround
0–15%	Rep or sales manager	Immediate
16–30%	VP Sales + deal desk lead	Within 8 business hours
30%+ or non-standard terms	CRO + CFO	Within 24 business hours, written rationale required

The two rules that make a matrix work: the highest-risk deals need the fewest, most senior approvers (not a committee), and every threshold has an SLA so the ladder speeds deals up instead of stalling them. Review the matrix quarterly — discount norms drift, and an out-of-date ladder gets ignored.

Step 4 — Set an SLA, because speed is the whole point

A deal desk that's slow is worse than no deal desk — reps route around it, and you're back to gut calls. Commit to a turnaround for each tier (the table above), and measure whether you hit it. The goal is to make the fast path the governed path, so nobody has to choose between closing the deal and doing it right.

Step 5 — Give approvers context, not just a request

The single biggest upgrade to a deal desk is what the approver sees. "Can I do 25% off?" is impossible to answer well. "This is 7 points above our largest discount at this ACV, and we granted net-90 once before — it slipped the renewal a quarter" is a decision you can make in thirty seconds.

So every review should arrive with context attached: comparable past deals, your norms for that segment, and the specific risk in the terms. That's what turns a deal desk from a bottleneck into a genuinely useful second brain. (It's also the hard part to do manually — pulling comparables from your own deal history is exactly the job we automate.)

Step 6 — Measure it with a few honest metrics

You don't need a dashboard. Track a handful of things and the desk stays credible:

Average discount (and its trend) — is discipline holding or drifting?
Review turnaround vs. your SLA — is the desk fast enough to be used?
Exception rate — how often deals go outside the guardrails, and who's approving them.
Win rate on reviewed deals — proof the desk isn't just killing deals.

Common mistakes to avoid

A committee instead of an owner. One accountable person is faster and clearer than a group.
A ladder with no SLA. Governance without speed gets bypassed.
Buying software before you have a process. The tool encodes your rules — write the rules first.
Never updating the charter. Norms change; a stale charter loses authority.

The shortcut: rent the function instead of building it

Building a deal desk internally is doable, but it takes a senior operator months to codify the playbook — and most teams under $50M ARR don't have that person to spare. That's the gap Precedent fills: a fractional deal desk that hands you the charter and the guardrails in weeks, with a one-page review brief — cited to your own deal history — on every deal before sign-off.