RevOps playbook · AI Employee: Rev

Discount Approval Routing

Discount approvals resolved in < 4 hours

The problem

Discount approvals are the deal-desk bottleneck that ships every rep to Slack DM the CFO on quarter-end. The approval matrix exists ("under 10% approves at manager, 10-20% at VP, 20%+ at CFO") but nobody executes it consistently. Deals stall for days waiting for approvers who are travelling; approvers rubber-stamp because the context wasn't ready; audit later finds discounts approved by the wrong authority. Sales blames deal desk, deal desk blames the matrix, everyone loses.

At a glance
Trigger
Form
Approvals
Manager, then finance for above-threshold
What it does
Writes to your systems
Systems
Salesforce · Slack
How it feels in production

An hour-by-hour walkthrough.

AE John submits DEAL-8842 through the quoting tool with a 24% discount request. Rev picks it up immediately, resolves the required approval chain (24% is over 20% → CFO), and drafts the approval packet: - Deal shape: $180K ACV, 3-year term, Acme Corp (existing customer, $60K expansion) - Discount: 24% off list ($56K discount value) - Business justification: John's note + auto-generated competitive context (Acme evaluated three vendors, decision by Friday) - Precedent: median discount for this segment + term is 12%; deals over 20% in the last 4 quarters: 8 (of which 5 won, 3 lost) - Risk flags: this rep's win-rate on discounted deals is average; last deal this rep discounted 20%+ closed within a week - CFO's own guardrails: this discount tier requires 3-year term (met), NRR floor above 0.90 (met) The packet goes to the CFO in Slack with approve / decline / conditionally-approve buttons and a suggested condition ("approve to 20% with 3-year prepay commitment"). CFO reads for 30 seconds, chooses conditionally-approve with the suggested condition. Rev writes the counter-offer back to the quoting tool, notifies John with talking points for the customer conversation, and logs the audit record with the CFO's decision and reasoning. Cycle time from AE submit to CFO decision: 12 minutes. Everyone saw the same numbers. Approval matrix enforced by mechanism, not memory.
How it works

Step by step.

  1. 01

    Detect the discount request + resolve the chain

    Read the quote / CPQ event for discount amount, term, product mix, customer profile. Match to the approval matrix; identify every required approver in the chain. Skip approval where the discount is within self-approve threshold.

    CPQ · Salesforce · Approval matrix
  2. 02

    Draft the approval packet with precedent + risk

    Deal shape, discount value, business justification, precedent from similar deals (segment + term + industry), risk flags (rep pattern, customer pattern, competitive context). Approver reads for 30 seconds, not 5 minutes.

    Historical data warehouse · CRM · Competitive intelligence
  3. 03

    Route to approver with actionable options

    Approver DM: approve / decline / conditionally-approve. Conditional approvals include suggested conditions ("approve to 20% with 3-year prepay"). Approver can edit the condition inline.

    Slack · Teams · Web UI · Approval flow
  4. 04

    Chase, escalate, and delegate

    Non-response nudge at 2 hours, escalate to delegate at 4 hours (approvers can pre-designate a delegate for OOO or busy periods), CFO cc at 8 hours if still pending. No deal stalls beyond a single working day.

    Slack · Teams · Escalation ladder
  5. 05

    Write the decision back + log for audit

    Approved / conditioned / declined outcome written to the CPQ + CRM. AE notified with talking points. Audit record with approver identity, reason, and decision timestamp. Weekly digest to CFO with approval-volume trends.

    CPQ · CRM · Audit log · Analytics
Systems and wiring

What you connect to make this run.

Salesforce CPQ · DealHub · Conga · HubSpot

read+write

Read the quote + discount event. Write the approval decision + any conditions back to the quote so the AE cannot ship a quote without the required approval.

Salesforce · HubSpot · CRM

read+write

Deal shape + customer history + rep history for precedent + risk. Write the approval-outcome audit trail as related records.

Snowflake · BigQuery

read

Historical won / lost by discount tier, segment, term. Feeds the precedent line in the approval packet + the CFO's monthly "where are we discounting" analysis.

Slack · Teams

read+write

Approver DM with buttons. AE notification with next-step talking points. Deal-desk channel for visibility on the queue. Delegation set via profile config.

What changes

Before and after, honestly.

Time from AE submit to approver decision
Before
1-5 business days
After
Under 4 hours (median 30 minutes)
% of approvals routed to the correct authority
Before
60-80% (matrix violations common)
After
99%+ (mechanism enforces the matrix)
Quarter-end approval backlog
Before
40-100 deals waiting
After
Under 10 deals
Average discount granted
Before
18-25%
After
12-18% (precedent + risk visibility tempers requests)
Frequently asked

Answers about this playbook.

What if the CFO wants to override the matrix for a strategic deal?

Override is a first-class action, not a workaround — CFO can grant an out-of-matrix approval with a reason. The audit trail captures "CFO override with rationale: X" so the exception is visible, not hidden.

How does it handle multi-approver chains (e.g. VP Sales + VP Finance)?

Parallel or sequential per the matrix. Parallel approvals wait for all; sequential proceed in order with rejection at any stage killing the chain. AE sees the current stage + who's holding the queue.

Can approvers see other pending requests in their queue?

Yes — approver dashboard shows their queue with SLA countdown, allowing batching if useful. Executives who prefer everything through Slack can stay in Slack; those who prefer a dashboard have one.

What if the AE wants to negotiate the condition (e.g. accept lower discount)?

Counter-offers are supported. AE can accept the conditional-approval as-is, or reply-negotiate through the same flow (draft a new proposal, route to the same approver). Each iteration is one round-trip.

How does it integrate with our deal desk?

Deal desk sits in the loop for complex deals (multi-product, custom terms, ELA). Approval chains route through deal desk first when the deal shape triggers it — Rev doesn't replace deal desk, it makes their queue actionable.

See it run on your data.

Free plan, no credit card. Connect the systems this playbook needs and run it against a past event first.