Nykaa Fashion App | UX Portfolio 2025

Tired of returns and exchanges while shopping clothes online? 🤷‍♂️

💁‍♀️ Try Find my fit!️

A UX case study on how I added a personalised recommendation feature based on the individual user’s personality in the Nykaa fashion app within 48 hours.

Link to final prototype

(This prototype can be viewed on figma browser and mobile app)

About Nykaa fashion

Nykaa Fashion is an e-commerce platform in India that offers a wide range of clothing, footwear and accessories for men, women, and children from various Indian as well as international brands.
It is an extension of Nykaa, which originally started as an online beauty and cosmetics store.

About the project

Recently I participated in a Product design challenge organised by the UXM Community, where all participants received a problem statement and had to come up with a minimum viable product (MVP) within 48 hrs

Problem Statement

My goal as a product designer was to add a recommendation feature for products based on body type, size and skin tone in the Nykaa fashion app.

Design process : Moving from ambiguity to clarity

Initial research findings

Several studies and surveys have highlighted the significance of sizing and fitting problems in online clothing purchases. Here are a few findings :

Approximately 70% of online shoppers face sizing issues with not getting the right size and experiencing inconsistent sizing across different brands when purchasing clothes online.
- Body Labs, a body modelling company
63% of consumers worldwide have encountered sizing issues when buying clothes online.
- Nosto, an e-commerce personalization platform
36% of clothing returns in the UK are due to size-related problems.
- Fits.me, a virtual fitting room company
65% of American consumers expressed concerns about not being able to try on clothes before buying them online, particularly due to sizing and fit issues. - Qualtrics

Why did I specifically choose to work on this problem statement? 🤔

Potential for innovation and out of the box thinking.
Impact on customers : Although features like trial and exchange exist but they are time consuming. Customers cannot afford to waste some days just trying and buying the clothes.
Impact on business : Businesses also have to incur costs like packing, handling, delivery, etc. for providing such services. If the return or replacement rate is high then these costs have to be borne by the businesses resulting in decrease in revenue.

This feature can help Nykaa fashion by impacting the following business metrics :

💵 Increase in average order value :

This feature can help in increasing the average order value by recommending products that match the customer’s needs and preferences, as customer’s may be more likely to purchase additional products when they are recommended clothes based on their individual personality.

😀 Increase in customer satisfaction :

Personalised recommendations can enhance the overall shopping experience by making it easier for customers to find products they like and that also fits them perfectly well. Higher level of customer satisfaction can lead to repeat business and positive word-of-mouth advertising.

🚚 Decrease in returns and exchanges :

When customers receive products that do not fit their body type or match their skin tone, they are more likely to return or exchange the product. Personalised product recommendations can help reduce the number of returns and exchanges, resulting in lower operational costs.

Project guidelines

1. Context :

To provide customers with a personalised shopping experience by adding a recommendation feature that takes into account various factors such as size, body type, skin tone, and other personal preferences of customers in the Nykaa fashion app.

2. Core idea of the Feature :

The personalised product recommendation feature will use data analysis and machine learning algorithms to suggest products that are likely to suit a user’s body type, size, skin tone and other preferences, by asking them a few questions. The algorithm will also consider other user’s reviews and the customer’s purchase history to make recommendations that match the user’s style and fit.

3. Constraints :

The feature may be limited by the quality and quantity of data available for analysis. Additionally, there may be privacy concerns related to the collection and use of customer data.
This feature is being built for recommending ready made clothes and not custom made. It is important to note that it may not provide exact fit like a custom made or tailored fit clothing. But, the aim here is to provide the best match using AI keeping in mind the individual’s body type, size and skin tone.

4. Assumptions :

The feature will require a significant amount of data analysis, as well as resources to implement and maintain the feature.
Customers will be willing to share their personal information and preferences in order to receive personalized recommendations.
The recommendations generated by the algorithm will be accurate and relevant to the user’s needs and preferences.
User’s will not confuse this with sort and filter options. This feature is being built for providing recommendations based on an individual’s body type, size and skin tone. After getting personalised recommendations, user’s can further apply sort and filter options to narrow down the results.

Competitor analysis

I researched whether similar features are being offered by any competing or non-competing businesses. This included businesses like :

Zivame, which is a women’s innerwear brand, had a similar feature called My zivame perfect fit for finding a perfect bra.

My zivame perfect fit feature

Another brand known as Bombay shirt company which offers tailor made clothing for men had a size recommendation option based on men’s body type.

Personalised clothing feature by Bombay shirt company

So primarily I had kept these two brands for inspiration on my mood board and was also exploring other brands at the same time.

Even Amazon recommends whether user should select one size up/down or select same size by comparing user’s usual size with a particular brand’s size based on user inputs (here amazon uses data collected from user reviews to recommend the perfect size).

Research on various types of tools used for providing recommendations based on body types :

There are various types of size recommendation tools, which can be used to provide personalised recommendations based on size, body type and skin tone such as :

Quiz-based questions
Purchase history analysis
User review analysis
Photo/video based solutions
AR/VR based solutions

Source : 3dlook.me

Here my main constraint was time. I had to choose a way to figure out how to build my solution, build high fidelity wireframes and prototype the solution to conduct a usability test.

From the above five options I had already rejected the 4th and 5th solution as it was not possible to complete it in the remaining time, it would also have required a lot of in-depth research.

I thought quiz based questions were a good choice but too many questions can irritate the users, so to maintain a right balance was also very important.

Along with the option based questions, If purchase history analysis and user review analysis is done through AI and machine learning, then it can aid in building the recommendation feature more accurately.

Collecting the inputs required for building the feature :

After doing a thorough competitor research I got many valuable insights and had a starting point to begin with. Also the research on types of recommendation features turned out to be quite helpful.

I decided to design my recommendation feature based on a quiz in the form of questions/options which will also be powered by purchase history analysis and user review analysis using AI and machine learning. Using this way I would be able to provide most accurate personalised recommendations to user’s within the limited time that I had.

Initially I thought I should not include body weight and height in my feature as these things are more relevant incase of tailored fitting and here I was designing a solution for ready-made clothing recommendations (although I was proven wrong ahead). Then I started exploring these inputs one by one as per body type, size and skin tone.

1. Body type :

Male and female body types differ from each other. Infact, there are as many body types as people living on the planet, but all these can be broadly classified into 5 types. Based on the most common body types users need to select the most appropriate one that matches their actual body type. Separate options would be provided for men and women as both have different body types. Here it should be noted that the body type needs to be approximate and not a perfect match.

After studying various body types I decided to keep 5 body types each for male and female, which would help user’s to get better recommendations suiting their own body type.

(Source : How To Dress For Your Body Type)

5 Male body types are as follows :

Slim (Rectangle)
Regular (Trapezoid)
Muscular (V-shaped)
Slight tummy (Triangle)
Large tummy (Oval)

5 Female body types are as follows :

Slim (Hour-glass)
Regular (Rectangle)
Muscular (V-shaped)
Slight tummy (Triangle)
Large tummy (Oval)

2. Clothing size :

User’s usually know what size fits them the best. The problem arises when they wish to try some new brand and many brands these days have different sizes which breaks the standard sizing nomenclature.

There are brands which fit individuals properly almost like a tailored fit. For someone one brand may work and for some the other. Therefore it is important to ask user’s themselves, which brand fits them most appropriately?

This similar thing is already being done by amazon so I decided to adopt the same thing in my feature too. If users are not sure about the brand or fitting they can select I don’t know option. Then, the AI will suggest the most appropriate recommendations based on their body type without considering particular brands.

3. Skin tone :

The color of clothes we wear plays an important role in how we look. Generally contrasting colors look good on people. Suppose someone is fair, then dark colored clothes would suit them more. Similarly if someone is dark, light colored clothes would suit them. On the basis of skin tones Indians can be broadly classified into 3 tones as follows :

Users don’t need to select an exact skin tone but just an appropriate one. I decided to give only 3 options in skin tones as the purpose is to suggest clothes. If it was a beauty related product then might be, more skin tones would have been a better option.

UI wireframes :

After having gathered all the required data it was time to start making low fidelity wireframes to get an idea of how the feature would look and to come up with an MVP.

Low fidelity wireframe version 1

Limitations of version 1

In absence of the option to choose gender, how will the users select the body types?
What if someone doesn’t want a recommendation for themselves but for someone else?

After making a few iterations of the above version, I came up with an MVP and decided to test it with the user’s.

High Fidelity Wireframe with basic prototype for usability testing

In the above version of the prototype I had already overcome the limitations of version 1.

Here, I’ve given users the option to select recommendations for themselves or for someone else.
If they select :

• Recommendations for myself : AI considers purchase history of user’s and reviews by other users, along with inputs entered for the 10 questions for providing accurate recommendations.

• Recommendations for someone else : AI considers recommendations based on user-entered inputs and reviews by other users but excluding their past purchase history.
Introduced an option to select gender. As the user’s can get recommendations for anyone else also.
Also, the body types should actually change as per the gender, but I was running short on time so, I decided to move ahead as it is for a usability test and make further iterations after taking user feedback.

Critical Decisions :

The most important decision to take was the placement (ingress point) and color for the Find my fit button. Types of buttons in the Nykaa fashion app are as below :

Color of the button : The app did not offer any specific/unique feature with some separate type of button. Considering the law of focal point, I decided to keep the Find my fit button in Nykaa pink color with white text as it was eye-catching and it was also the primary color of Nykaa fashion.
Placement of the button : After thinking thoroughly I decided to place the feature button on the home screen as well as under the individual store section of men and women.

The reason for these decisions were as follows :

Find my fit is a new feature to be introduced which is going to take a huge investment in tech. It is important that user’s notice the same and give it a try.
Keeping in mind the heuristic principle of flexibility and efficiency of use, I decided to place it on the home screen as well as in men and women’s stores individually too, so that users don’t have to come back to the home screen to try the feature.
On the home screen the button is placed between the store section and the discover section. Applying fitt’s law, the button is placed full length on screen and placed at such a distance that it is easy for the fingers to tap on it quickly.
I decided to place it in men's and women’s stores individually in proximity with the search bar so that users are able to find it easily.
I was also considering including the feature in the profile/account section but as I had already decided to keep it in the first fold of the home screen there was no point in keeping it in the profile section again. Maybe when people are used to the feature, then in the future the button can be removed from the home screen and placed in the profile section.

Understanding the userflow :

Usability Testing :

I tested my prototype with three people. Amongst them two were girls and one was a guy. I presented the prototype via a zoom call and asked them to observe the screen and what they felt after looking at the elements one by one. I was observing where they were getting stuck, feeling confused or unable to understand and at the same time I was noting down it.

Screens to be presented for usability testing

1-v-1 Zoom call with user’s for usability testing

Insights from usability testing with USER 1

Insights from usability testing with USER 2

Insights from usability testing with USER 3

Consolidated feedback :

The body types section does not seem scrollable. User’s can miss the remaining body types and may move forward without scrolling and checking all body types.
The top wear sizes section also doesn’t seem to be scrollable at the first glance. User’s may not check all the sizes available.
Female user’s found the color of the Find my fit button very attractive and eye grabbing, whereas the male user found it to be attractive but felt that it may be something relating to women, specifically due to its color. He also found the home page to be women centric.
Although this feature is being built for recommending ready-made clothes, it would be better if user’s are provided the option to select height and weight for more accurate recommendations.

Reiterating to make the solution better :

Changes to be made :

1. Separate detailed images for male and female body types to be used in the prototype.

Male and female body types during usability testing

Detailed images of male and female body types used to re-iterate the prototype

2. To give users the option to select height, as specifically different types of wear have different lengths. Giving the option to select height would give more accurate recommendations to users.

3. To give users the option to select weight, as using the body weight, the algorithm can analyse the body type more accurately.

4. To adopt skin tones matching more towards the Indian skin.

5. To change the color of Find my fit button on the home screen as well as in the store section (from pink to black, similar to button 8 mentioned above) to make it more gender neutral and to adhere to Nykaa’s guidelines.

Component library :

Final prototype :

Made a minimum viable product with high fidelity wireframes and a working prototype after incorporating user feedback