Case File 06 · Product Managing

Building a tiered approval engine for corporate booking discounts

A three-layer, rules-based approval system for the Edit Selling Price page, built to automate corporate booking modifications and cut support ticket volume in half.

TYPE
SaaS Product Creation · C2
DOMAIN
B2C Enterprise / Corporate Booking
DATE
May 13, 2024
ROLE
Product Managing
Case file cover — approval workflow
Product brief

Manual discounting was flooding support with tickets

The Edit Selling Price page needed to automate corporate booking modifications and offline invoice processing. Manual handling of discounts and booking modifications had led to high support ticket volumes — so the fix combined process automation with new self-serve features: tiered discount approval, cancellation options, and improved booking capabilities.

Headline result Corporate support tickets dropped 50%, from 11,726 to 5,863, after the automated tiered approval flow went live.
Approval logic

Three layers, each with its own approver and threshold

Every discount request routes through up to three approval layers depending on how far the requested price falls below the property's standard selling price. Black arrows carry approval actions; red arrows are notification-only and don't block the flow.

2nd and 3rd layer approval routing diagram
2nd & 3rd layer approval routing
1
1st Layer — Auto Approved / Site Discount Rule

The request goes straight to the CSE (Customer Support Executive) for immediate approval — no manual review needed.

Direct flow → CSE
2
2nd Layer — Manual Approved / Corporate Limit Percentage

If the request needs manual intervention or hits a corporate limit percentage, it goes to the Manager first, then flows back to the CSE for final approval. Red arrows here are notifications only.

Manager → CSE
3
3rd Layer — Manual Approved / 3rd Layer Bulk Upload

Bulk uploads or deeper discounts escalate to the Central/Regional team, then to the Manager, then to the CSE for final sign-off.

Central/Regional → Manager → CSE
Pricing logic

Each layer is a progressively deeper discount threshold

The requested price is compared against three descending thresholds. Meeting the 1st layer price approves at the top tier; falling short pushes the request into deeper, more scrutinized discount layers.

Threshold example (IDR)
Price Request1st Layer2nd Layer3rd Layer
146,718116,15181,510
130,000YES
90,000YES
147,000YES
100,000YES
117,000YES
Layer functionality
  • A request must meet a layer's threshold to be approved at that layer.
  • Falling below a layer's threshold flags the request for review at the next level.
  • IDR 130,000 clears the 1st layer and lands at the 2nd; IDR 90,000 needs the deepest discount and skips straight past the 1st layer.
Technical workflow
  • Input the requested price into the system.
  • Compare it against the 1st, 2nd, and 3rd layer thresholds.
  • Auto-approve on match, or flag for manual approval.
  • Each subsequent layer applies progressively deeper discount logic.
Success metrics

What automation actually moved

50%
reduction in corporate support ticket interactions
30%
decrease in manual handling of bookings & offline invoices
25%
improvement in user experience & operational efficiency
Also shipped Cancellation options and enhanced booking capabilities were introduced alongside the approval engine, compounding the efficiency gains.
Preview

The approval flow in the field

Sales agents submit an edit-selling-price request with proof of owner approval; each layer's manual reviewer can approve or reject with a reason, and every request logs its original price against the auto/2nd/3rd layer thresholds for transparency.

Request status changed to rejected
Reject confirmation
Specify reject reason modal
Reject reason capture
Sales agent requests list with pricing breakdown
Requests & pricing breakdown
Edit selling price upload proof screen
Upload owner's proof
Edit selling price conditions screen
Selling price conditions
What's next

Where the tiered layer goes from here

  1. Use the tiered layer to oversee broader business operational problems, beyond pricing.
  2. Use the tiered layer to minimize discounting on high-tier properties specifically.