Our average deal size dropped 12% in a quarter while bookings kept climbing. Nobody lied on a forecast call and nobody missed a number — my reps were just discounting more, quietly, because nothing in our process told them not to.
That's the moment I put a discount approval process in place, and it wasn't a deal desk platform or a new hire. It was three tiers, one Slack channel, and a CRM field nobody could skip. Here's the exact structure, why it works, and the signal that tells you when you've outgrown it.
Why "just check with me first" doesn't scale
Every founder starts with an informal rule: reps can offer a standard discount, and anything bigger needs a quick check-in. It works for the first five deals. Then a rep gets busy, decides a 22% discount "felt like a 15% situation," and closes it without asking, because the rule was never written down anywhere they'd see it under deadline pressure.
The bigger problem isn't the one rogue discount. It's that you have no record of how often this happens, which reps do it, or which deal types trigger it. By the time it shows up in your numbers — smaller average deal size, thinner gross margin, a board member asking why ACV is flat despite headcount growth — you're reconstructing the pattern from memory instead of a paper trail.
The 3-tier discount approval process
The fix isn't more approval friction everywhere. Most deals don't need it. The fix is calibrating friction to dollar impact, so the 80% of deals that don't matter move fast and the 20% that do get a real check.
Tier 1 — 0 to 10% off list price: rep-approved, no sign-off required. This covers the overwhelming majority of deals and keeps your sales cycle exactly as fast as it was before you had a process. Every discount, even at this tier, gets logged in a required CRM field — percentage and one-line reason. That's the only new step for most deals.
Tier 2 — 11 to 25% off list price: requires written approval from a sales lead or founder before the deal is sent, not after. This happens in a Slack thread, not a meeting: rep posts the deal, the discount, and the reason; approver responds same-day. The point isn't to slow the deal down, it's to make someone besides the rep look at it once.
Tier 3 — anything above 25%, or non-standard terms bundled in: extended payment terms, multi-year discount stacking, custom SLA language. This requires founder or CEO sign-off, documented in the deal record before the contract goes out. These are rare, but they're where the real margin damage happens, and they're exactly the deals a rep under end-of-month pressure is most likely to push through fast.
Running this without deal desk software
You don't need a platform to run this. A public Slack channel called #discount-approvals and a required "Discount % / Reason" field on every opportunity in your CRM covers all three tiers completely. The channel creates a timestamped, searchable record. The CRM field means Tier 1 deals still get tracked even though nobody has to approve them.
The habit that matters more than the tooling: review Tier 2 and Tier 3 approvals as a batch, once a week, for fifteen minutes. Looking at them one at a time, in the moment, you'll approve almost everything — each individual case sounds reasonable. Looking at ten of them together, patterns show up that no single approval would have revealed.
The pattern you're actually looking for
In practice, discount leakage rarely comes from ten different reps making ten different bad calls. It concentrates: one rep who discounts to hit a personal number late in the month, one segment (usually renewals, where the customer already knows their leverage) that gets discounted by default, or one deal size band where reps assume the buyer expects to negotiate.
Once you can see the pattern instead of individual approvals, you fix the actual cause — a comp plan that rewards closed revenue over margin, a renewal playbook that leads with price instead of value, a lack of coaching on how to hold a number under pressure — instead of adding another approval step that just slows down the next deal.
When you need an actual deal desk function
The three-tier process holds until volume outgrows a founder or sales lead reviewing approvals in Slack. The signal isn't revenue or headcount, it's approval volume: once you're running more than roughly 15 to 20 Tier 2/3 approvals a month, or a rep is regularly waiting more than a day for a response, the bottleneck you built to fix leakage becomes its own drag on the sales cycle. That's when a dedicated deal desk function — even a few hours a week from someone in RevOps or finance, not a new full-time hire — earns its keep.
Frequently asked questions
Won't this slow down my sales cycle?
Tier 1 covers most deals and adds zero approval friction, just a logging step. The friction is deliberately concentrated on the discounts big enough that a day's delay is worth it.
What if a rep discounts without approval anyway?
Treat it as a process failure the first time, not a discipline problem: the deal gets flagged in the weekly review, and you check whether the tier thresholds or the approval speed are the actual issue before assuming the rep ignored the rule on purpose.
Should the tiers be different for renewals versus new business?
Often yes. Renewal discounts tend to run higher by default because customers already have leverage, so some founders set a slightly higher Tier 1 ceiling for renewals while keeping Tier 2 and 3 identical, so the real exceptions still get a second look.
None of this requires new software or a new hire. Set your three thresholds this week, add the discount and reason field to your CRM, and open the Slack channel before your next deal closes. The board deck problem isn't next quarter's discounting — it's this week's, and it's the easiest one you'll ever fix.