Average Order Value (AOV)
AOV = Total value of all orders ÷ Total number of orders
Here’s the story of how BharatAgri lifted its AOV by nearly ₹200 with one key product feature…
Context
BharatAgri sells agricultural products (such as seeds, insecticides, fertilizers, etc.) to farmers.
Farmers typically purchase these products from local shops as well.
Local shopkeepers recommend product bundles to farmers, resulting in better outcomes and higher order values.
What is product bundle?
Bundle = Primary Product + Secondary Product
Primary Product
The product, user intends to purchase
Example -
Primary Product (P) : Shoes 👟
Secondary Product
A complementary product shown to the user that can be used alongside the primary product to achieve better results.
Secondary Product (S) :
Socks 🧦 (recommended based on the selected shoes)
Business Opportunity
At the time of checkout, we can suggest product bundles to farmers within the BharatAgri application.
Pilot (Q4 2023)
We implemented this feature using a third-party tool called Releasit on our mobile web platform.
Results
The Pilot was successful 🎉
We were able to sell bundles, and the average order value (AOV) increased.
Our next goal was to evolve this experiment successfully into a long-term driver of AOV.
🥷🏻
I stepped in as the product owner for this initiative
— and that’s where the journey truly began.
Year
2023-24
Role
Product Owner
Skills
Product Management, Product Design
Table of Content
Version 1 : Mobile Web
1.1 Goal
1.2 Derived Goals
1.3 Challange 1
1.4 Release Plan
1.5 Bundle Creation Process
1.6 UI Implementation
1.7 Measuring Results
Version 2 : Android App
2.1 UI Implementation
2.2 Measuring Results
Version 3 : Bundling Automation
3.1 Automation Logic
3.2 AOV Data
1.1 Goal
Increase average order value(AOV) using bundling
1.2 Derived Goals
1.2.1 Increase Bundle Coverage
Introduce bundles for top-selling products
Expand bundles to cover the full product catalog
1.2.2 Increase Bundle Sales Conversion
Improve secondary products relevancy
Optimize pricing of secondary products
1.3 Challenge 1 : Replacing Releasit
The third-party app limitations -
❌ Scale limitations
❌ Cost inefficiency
❌ No bulk upload capability
❌ Design constraints
❌ Slow iteration cycles
❌ Limited automation
Releasit
The third-party app
Identifying Releasit data points
1.4 Release Plan (Mobile Web)
Version 1 : Manual sheet upload
Frontend :
Replicate the same functionality as releasit, i,e show the assigned secondary product alongside the primary product.
Secondary product properties :
Checkbox (CTA)
Name
Image
Selling Price
Discounted Price
Discount %
Backend :
Map secondary product with primary product using product IDs using sheet upload or directly from the CMS.
Sync the database in order to reflect the changes.
WHY the manual sheet upload?
WHY can't system create bundles automatically ?
Creating a bundle is a complex process where we need to consider multiple aspects from agronomy to delivering products from same warehouse.
Here, Let me take you to the bundle creation process to understand it better…
1.5 Bundle Creation Process
To map secondary product we need to understand the category & type of primary product so that we can match the compatibility of both. Let's understand this process using a primary product "coragen" which is a insecticide and used in spray form. Here there are 2 secondary product which can be bundled with it.
Example :
Primary Product :
Coragen
(Insecticide)
Spray
Secondary Product Options :
IFC Super Sticker
Anand Wet Gold
LEVEL 1 : Agronomy
✅
Both the products are available in warehouse and with seller
IFC Super Sticker
Anand Wet Gold
LEVEL 2 : Stock Availability
✅
Condition : Primary product weight bracket should not exceed
IFC Super Sticker
Anand Wet Gold
LEVEL 3 : Weight Bracket
Condition 1 : Bundle margin should not drop
Condition 2 : Increase margins till secondary product selling price is Rs99
✅
IFC Super Sticker
❌
Anand Wet Gold
LEVEL 4 : Margins
1.6 UI Implementation
For the V1, we've replicated the same UI of the third party app (releasit) so that user experience does not changes.
1.7 Measuring Results
Here's a 10day sale data example :
Column A : Primary product names
Column C : Total primary product sold
Column D : Total bundle sold (primary+secondary)
Column F : Bundle sale contribution %
Low performing bundles (Sale < 10%) gets replaced with a different secondary products
My Bandwidth Allocation
Low performing bundles (Sale < 10%) gets replaced with different secondary products
Top selling 30 products gets updated in 10 day interval
Category-wise 10 top selling products gets updated in 10 day interval
UI & Microanimation experiments
Check for top selling & least selling bundles
Goals Achieved
Introduce bundles for top-selling products - ✅ Achieved
Optimize pricing of secondary products - ↗️ Ongoing Process
Improve secondary products relevancy - ↗️ Ongoing Process
Expand bundles to cover the full product catalog - ❌ Not Achieved Yet
Version 2 (Mobile App)
Mobile web was performing good!
Time to replicate the feature on application
2.1 UI Implementation
For the V1, we've replicated the same UI of the mobile web so that user experience does not changes.
1.7 Measuring Results
We analyzed just the first 10 days of data after introducing bundling—and even within this short window, the results were significant.
⭐ GMV increased by 25%
⭐ GMV increased by 36%
21-29 May
01-09 June
Goals Achieved
Introduce bundles for top-selling products - ✅ Achieved
Optimize pricing of secondary products - ✅ Achieved
Improve secondary products relevancy - ✅ Achieved
Expand bundles to cover the full product catalog - ❌ Not Achieved Yet
Version 3 (Bundling Automation)
Last goal : expand bundles to cover full product catalog
After nearly a year, it was finally time to automate the entire process. Bundling had consistently delivered strong results, and it became clear that we needed to prioritize this feature and build a proper backend to support it.
3.1 Automation Logic
Here's the ideation for automation logic
3.2 AOV Data
Few more google sheets
Projects
Home