Shopping List for Flipkart Grocery

By connecting a user-created shopping list with Flipkart's Search API, I proposed, defined and successfully launched a new basket building journey that leveraged the power of existing search tools to uplift a prevalent user behavior.

Users who used a shopping list added to basket and converted nearly 200% better.

Introduction

Flipkart Grocery is a business unit via which we sold dry groceries - pulses, oils, flours, grains, etc - which are usually stocked in bulk in Indian households. The core value proposition for this business lies in the supply chain cost efficiency for large baskets, and ability to pass on the accrued savings to the user. As a result, Flipkart Grocery sees approximately 12-14 units per order, as compared to a Flipkart-wide average of 1.5.

Why build a shopping list?

1. We have limited time to convert users

27% returning customers just have 1 session a month and 75% of these users don’t do an add to basket. The grocery ‘Aha’ moment where users start to convert is 4 units, however few reach it.

2. Long list = more cognitive load

Users shop for a multi-product list by default and FK Grocery further incentivises this behaviour through tiered basket discounts. (units per order ~10-12)

3. Users are already using lists in the real world

Traditional grocery shoppers already create handwritten lists to keep track of their grocery shopping. We observed that since a household's grocery needs do not change drastically month on month, 80% of their cart repeats on a super-category level.

Design Considerations

Entry Point

The home page is an expensive place

The Grocery home page is one of the most viewed pages, as it often forms the start of the user's Grocery journey. We experimented with temporary entry points on the home page top fold, but this approach was fundamentally unviable in the long term. Why? Given the HP's high viewshare, it is an expensive and dynamic real estate, and cannot be blocked indefinitely by a particular feature.

A list maker already has search intent

Since the user intent behind list creation is "Search for multiple products", it seemed logical to include the entry point within the existing Search flow.

It works better the second time

Since a household's grocery needs are more or less consistent over time, we noticed that there exists an 80% overlap month on month in a user's cart, on a super-category level. Therefore, a user's list should be retained month on month, with an option to create a fresh list.

Gathering Input

My user is afraid to type

A user should be able to express themselves in a manner that is intuitive to them. Since Flipkart’s typical user base tend to be uncomfortable with typing in English, other tools are important to make text input easier - local language support, spelling prediction, popular suggestions, etc.

The list already exists

We've already talked about how most households this feature targets already make written lists as part of their grocery planning. Members in shared households often share these lists, either as physical paper, photos, text messages, or verbally. A user should be able to ingest their entire list at once - by pasting comma separated items, scanning a paper list, or speaking out the list.

Designing the List

Which direction should the list flow?

When you write a list on paper, the list flows from top to bottom. Does that need to be the case in a digital list? Consider Google Assistant’s native shopping list: The list input is at the top, and entries are ranked from newest to oldest. This is in stark contrast to their own note taking app, Google Keep, which orients the list in the reverse order. (As of writing this, Google has shut down their Assistant’s list functionality, directing you to use Keep instead)\

Results powered by Flipkart Search

Since the list is powered by Flipkart Search on the backend, it automatically inherits all of the magic that helps users find the products they love. Brands & SKUs a user buys regularly are ranked highest, followed by best prices, popular choices and more. Support for regional nomenclature and Indic text is baked in, and Search Synonyms take care of all the messy work.

A Flow Designed for Flexibility

A shopping list is often a mix of urgent and non urgent requirements, the newest items in the list tending to be the most important. Further, a user should be able to shop for individual items, or the entire list at once, in whatever order makes sense to them. The "Added" indicator keeps the system state visible regardless of which step the user is on.

Learning and Challenges

The challenges in this project were mainly around stakeholder management and in securing buy-in from engineering to make this a reality. Because i initiated this project as a designer and challenged the conventional product roadmapping approach at Flipkart, there was a significant learning curve when it came to estimating business impact, convincing stakeholders of the value of the project, etc

As a design initiated project, a lot of my responsibilities lay at the edges of my job description: Product Definition, Product Strategy, Writing a PRD, Coordinating various engineering, research and graphic design resources, and acting as a product owner.

Being the initiator of the project, I had to explore and solve problems far from the “typical” role of a designer - how do you write a PRD? how do you define and estimate success? How do you convince leadership to invest engineering bandwidth into your project? How do you resolve disputes between various stakeholders? While this learning was enormous, I was eventually supported in my efforts when a PM and engineering team was assigned. The PM and I collaborated extensively to refine the product vision, coordinate teams and align stakeholders. My design leadership too played a role by providing constant feedback and support when navigating bureaucracy.

Reflection

Looking back, this project taught me how to go beyond my role, and drive impact at a product strategy and roadmapping level. It taught me that as a designer, my ideas can make a real difference to the overall direction of the product.