2nd place
3000+ teams
100 teams shortlisted
5 finalist teams
2nd place
3000+ teams
100 teams shortlisted
5 finalist teams
We participated in IIM Ahmedabad’s Product Wizards Case competition
and received the topic of AI Personal Shopping Companions.
We were an incredibly diverse team, with a tech, UX design and economics student working together. We bought so much to the table and created a submission we were proud of.
We developed Talon through our research in this competition. As the UX designer on team, I made prototype for the competition, but immediately knew that this concept deserved a full UX project.
So here it is :)
The idea
The idea
Global Sales in Fashion
Global Sales in Fashion
Global Sales in Footwear
Global Sales in Footwear
of footwear shopping shifted online due to the pandemic
of footwear shopping shifted online due to the pandemic
And yet, shopping for shoes online feels really unreliable. Foot wear as a category actually has one of the highest return rates.
And yet, shopping for shoes online feels really unreliable. Foot wear as a category actually has one of the highest return rates.
$ 2.5T
$ 2.5T
$ 460B
$ 460B
30%
30%
Footwear is an afterthought?
Footwear is an afterthought?
Most shopping sites are built for clothing- with footwear being added just for completion of the wardrobe. Eg: Myntra has 6 filters for sleeves alone, but only a few basic choices for the entire category of footwear.
So we decided to use the strengths of the current AI landscape to solve some of the most common online shoe-shopping pain points:
Most shopping sites are built for clothing- with footwear being added just for completion of the wardrobe. Eg: Myntra has 6 filters for sleeves alone, but only a few basic choices for the entire category of footwear.
So we decided to use the strengths of the current AI landscape to solve some of the most common online shoe-shopping pain points:
Inconsistent Sizing: There is no universal standard for shoe sizes. A size 9 in one brand may feel like an 8.5 or 9.5 in another.
Inconsistent Sizing: There is no universal standard for shoe sizes. A size 9 in one brand may feel like an 8.5 or 9.5 in another.
Lack of Comfort data: Shoes are complex 3D objects, and fit depends not just on length but also width, arch support, toe box, and instep height.
Lack of Comfort data: Shoes are complex 3D objects, and fit depends not just on length but also width, arch support, toe box, and instep height.
Other features
Other features
of shoppers expect personalization and precise results, in place of endless scrolling. (AI search impact)
of shoppers expect personalization and precise results, in place of endless scrolling. (AI search impact)
of shoppers want to be more sustainable, and are looking for easier ways to build those practices into their life.
of shoppers want to be more sustainable, and are looking for easier ways to build those practices into their life.
71%
71%
88%
88%
I knew I wanted to include SDG 12 (Responsible Consumption and Production) by addressing the product lifecycle and consumer behaviour. The app reduces returns/waste by improving comfort precision, educates buyers about their shoe’s impact with a sustainability score, and enables circularity through Care, Repair, and Sell features.
I knew I wanted to include SDG 12 (Responsible Consumption and Production) by addressing the product lifecycle and consumer behaviour. The app reduces returns/waste by improving comfort precision, educates buyers about their shoe’s impact with a sustainability score, and enables circularity through Care, Repair, and Sell features.
Branding
Branding
in real use cases, pain points, measurement technology and shopping tag data. Not AI magic.
in real use cases, pain points, measurement technology and shopping tag data. Not AI magic.
harnessing of AI and shopping behaviour data, while knowing its strengths and weaknesses.
harnessing of AI and shopping behaviour data, while knowing its strengths and weaknesses.
and sharp execution for a common and expensive purchase problem.
and sharp execution for a common and expensive purchase problem.
Innovative
Innovative
Technological
Technological
Grounded
Grounded
TA
TA
ON
ON
PHASE 1
PHASE 1
ONBOARDING
ONBOARDING
Collecting measurements and preferences with compliant AI privacy notice, and local handling.
Collecting measurements and preferences with compliant AI privacy notice, and local handling.
= Contour map of the feet, later used for fit & comfort analysis.
= Contour map of the feet, later used for fit & comfort analysis.
= Record of past purchases that are later used for understanding style and preference.
= Record of past purchases that are later used for understanding style and preference.
PHASE 2
SEARCH
State: AI listening
State: User Speaking
Data point: The average search term pre-AI was 3-4 words.
AI natural language search is an average 23 words.
The insight is clear: Users are ready for a conversational experience, and willing to put in a greater effort upfront, for precise results and reduced scrolling later.
Source: Writesonic (2024)
Moz, Statista, & SparkToro (2019–2024)
Lets harness that with
Visual Search currently accounts for 6.4% of total e-commerce revenue.
2-STEP SEARCH
In this additional step, the user is asked clarifying questions to better understand their request, and deliver fewer but more personalized results.
0
2019
2020
2021
2022
2023
2024
5
10
15
20
25
ChatGPT launches.
Users enter keywords for the search engine.
AI search explodes in length as users provide context, persona, and format instructions. Prompt engineering becomes a real skill.
PHASE 2
SEARCH
State: AI listening
State: User Speaking
Data point: The average search term pre-AI was 3-4 words.
AI natural language search is an average 23 words.
The insight is clear: Users are ready for a conversational experience, and willing to put in a greater effort upfront, for precise results and reduced scrolling later.
Source: Writesonic (2024)
Moz, Statista, & SparkToro (2019–2024)
Lets harness that with
Visual Search currently accounts for 6.4% of total e-commerce revenue.
2-STEP SEARCH
In this additional step, the user is asked clarifying questions to better understand their request, and deliver fewer but more personalized results.
0
2019
2020
2021
2022
2023
2024
5
10
15
20
25
ChatGPT launches.
Users enter keywords for the search engine.
AI search explodes in length as users provide context, persona, and format instructions. Prompt engineering becomes a real skill.
in the age of ethical boundaries being rampantly crossed by AI companies.
in the age of ethical boundaries being rampantly crossed by AI companies.
All the data is displayed along with the relevant, compliant privacy notice.
All the data is displayed along with the relevant, compliant privacy notice.
It can all be viewed, edited, exported or deleted when the user wants.
It can all be viewed, edited, exported or deleted when the user wants.
TRANSPARENCY
TRANSPARENCY
and complete data control
and complete data control
PROFILE
PAGE
PROFILE
PAGE
in the age of ethical boundaries being rampantly crossed by AI companies.
in the age of ethical boundaries being rampantly crossed by AI companies.
All the data is displayed along with the relevant, compliant privacy notice.
All the data is displayed along with the relevant, compliant privacy notice.
It can all be viewed, edited, exported or deleted when the user wants.
It can all be viewed, edited, exported or deleted when the user wants.
TRANSPARENCY
TRANSPARENCY
and complete data control
and complete data control
PROFILE
PAGE
PROFILE
PAGE