AVA Flavours: Distribution Models in Auto Vendor Allocation

The Auto Vendor Allocation (AVA) system in Rentlz provides flexible distribution strategies to efficiently assign bookings to vendors. Based on the operational setup, city, booking type, and cab category, you can configure AVA to use one of the following models:

1. Percentage-Based Distribution (Fair Share Model)

Overview

This is the most commonly used AVA model where workload is distributed proportionally among vendors based on predefined percentage values.

How It Works

• Admins assign percentage allocations to each vendor for every City + Booking Type combination.

• These percentages are converted into allocation ratios.

• The system uses weighted random selection and a modified round-robin approach to avoid consecutive assignments to the same vendor.

Example

If Vendor A is assigned 60% and Vendor B 40% for Bangalore – Airport Pickup, bookings are automatically distributed in approximately that ratio over time.

Best Suited For

• Cities with consistent demand

• Balanced and fair vendor workload

• Situations requiring proportional distribution

2. Priority-Derived from Percentage (System-Prioritized Model)

Overview

Though it functions under a "priority" concept, this model is configured exactly like the Percentage-Based model. Admins define percentages, and the system interprets higher percentages as higher priority.

How It Works

• In the UI, admins still assign percentage values for each City + Booking Type.

• The system calculates implicit priority based on these values.

• Vendor with the highest defined percentage receives booking preference.

• If that vendor doesn’t accept within the allowed response window, the booking is passed to the next vendor, in order of descending percentage.

Important Notes

• No manual "P1", "P2", or "P3" labels are used.

• Priority is dynamically derived by the system, based on the configured allocation percentages.

• This is ideal when vendor preference is important but workload balancing is not strictly required.

Best Suited For

• Cities where certain vendors should be preferred

• Scenarios requiring fallback mechanisms

• High-performance vendors who deserve more allocation

3. Quota-Based Priority Distribution (Targeted Fulfillment Model)

Overview

This model is used when vendors are assigned specific daily quotas for a defined combination of City + Booking Type + Cab Type. Here, priority is explicitly configured and only comes into play when vendors have remaining quota.

How It Works

1. Quota Setup

   • For each City + Booking Type + Cab Type, a numeric quota is assigned to each vendor.

   • Example: Vendor A – 15 trips, Vendor B – 10 trips

2. Eligibility Check

   • Only vendors with a remaining quota greater than 0 are considered.

3. Explicit Priority

   • Among eligible vendors, bookings are assigned based on manually configured priority levels (P1, P2, etc.).

   • A vendor’s priority is ignored if their quota is exhausted.

4. Reassignment & Reset

   • If a vendor doesn’t respond within the configured acceptance window, the booking is passed to the next vendor in the priority chain.

   • All quotas reset at midnight, and historical data is logged.

Example

For City: Hyderabad, Booking Type: Airport Pickup, Cab Type: SUV

• Vendor A is considered first.

• If Vendor A doesn’t accept, the booking moves to Vendor B.

• Vendor C is skipped as quota is exhausted.

Best Suited For

• Vendor SLAs with guaranteed trip volumes

• Controlled ramp-up for new vendors

• Fixed distribution targets based on vendor capacity

Summary of AVA Distribution Models

Configuration Notes for Admins

• In both the Percentage-Based and Priority-Derived models, you will only configure percentages.

• The Quota-Based Priority model is the only one where you explicitly assign both quota and priority.

• Ensure that for each configuration (city + booking type or cab type), the total percentage equals 100%.

• Use the AVA monitoring tools to track:

  • Vendor performance

  • Acceptance rates

  • Quota fulfillment

• Adjust configurations regularly based on vendor capacity and operational demand.