Redesigned the gift personalization experience into a single-screen progressive layout, increasing conversion on new multi-input products by up to 14% and cutting completion time by up to 22%.

• The Problem:
  The legacy flow forced users through a separate step and button click for every single input. On complex, multi-input items, this created heavy friction and caused high user drop-off.
• The Approach:
  Analyzed technical constraints across 8 product categories, focused on high-friction multi-input products, and replaced the multi-step sequence with a single-screen progressive input UI featuring live previews and streamlined copy.
• The Outcome:
  Raised conversion on new multi-input products by up to 14%, boosted multi-gift conversion by +8.7% overall, and reduced median completion time by up to 22%.

IGP is an Indian gifting platform where people buy cakes, flowers, and personalized gifts for birthdays, anniversaries, and celebrations. Aside from personal occasions, a large chunk of orders comes from seasonal events like Rakshabandhan, where users buy rakhis and gift hampers.

The company wanted to sell more personalized gifts, but our existing flow wasn't built to handle complex items or high volume smoothly. The original personalization backend was slow, expensive to run, and very rigid on the frontend. Because the frontend was tightly bound to this legacy system, users had to click "Proceed" after entering every single text field or photo. If someone was buying an item with four custom fields, a multi-product basket, or multiple personalized gifts, they had to go through a long, repetitive sequence of screens.

We decided to upgrade the backend to handle personalization better. While the product manager worked on the backend and internal panel, I was tasked with updating the frontend user experience.

Before jumping into designs, I spent time understanding how the new backend worked and what technical limitations I needed to design around. The main constraint was that we couldn’t rebuild everything from scratch or migrate all products overnight. While the new backend was live, almost all existing products on the site still ran on the legacy backend. The catalogue team was updating products manually, but with new items added daily, both backends were going to coexist for a while.

I had to design a new UI that felt unified to the user across the site, even though the underlying backend differed depending on the product. To account for all variations, we mapped out eight product categories across the platform:

• Single product : Single input
• Single product : Multiple inputs
• Multiple products : Single input
• Multiple products : Multiple inputs (Basket)
• Single product : Single input : Multiple personalizations
• Single product : Multiple inputs : Multiple personalizations
• Multiple products : Single input : Multiple personalizations
• Multiple products : Multiple inputs : Multiple personalizations

To decide where to focus our effort, we first pulled data on how the existing flow was performing:

• What was the overall personalization conversion rate?
• How did conversion differ between single-input and multi-input products?
• How were multi-product baskets and multiple personalizations performing?
• How long did users take on average to complete personalization across these different types?

The data made one thing clear: single-input products were already fairly simple, so updating their UI wouldn't change conversion or completion time much. The real bottleneck was on products with multiple inputs and multiple personalizations, where users were forced through tedious multi-step screens.

So, we focused our design efforts on making multi-input and multi-personalization products faster and easier to complete. We also looked at how competitor sites and personalization-focused platforms handled complex inputs before drafting our UI variations.

We replaced the rigid multi-step flow with a cleaner, progressive interface. Here are the main changes:

• Single-screen progressive input
  Users can now fill in all inputs on one screen and see live updates on the photo preview as they go. Previously, every single input required moving to a new step to generate a separate preview.
• Trimmed unnecessary text
  The old screen had accumulated redundant titles, helper text, and legacy elements over the years. I stripped away the noise and kept only the copy that was actually useful to the user.
• Better image editing tools
  Previously, users could only upload a photo as-is. We added simple cropping and rotation tools directly into the upload flow.
• Clear progress indicators
  Added a straightforward indicator so users always know how many inputs remain.
• Structured multi-product flows
  We separated multi-product inputs clearly. Inputs for one product stay on a single screen, and moving to the next product moves to the next screen. This replaced the old flow where every single field felt like an identical, generic step.
• Natively integrated multiple personalizations
  In the old design, multiple personalization was placed at the top of the screen like an afterthought. In the new UI, it's built directly into the main input block.
• Easier editing
  Because all fields stay on one screen, users can review or edit any input instantly without navigating back and forth through multiple screens.

We ran an A/B test comparing the legacy flow (Default) with the updated flow (Variant).

• Conversion improved across every segment.
  Single personalizations saw steady gains (+0.8% to +6.6% overall), reaching a +16.8% lift on combined inputs for new products (Types 4 & 6).
• Multiple personalizations showed the biggest impact.
  Overall conversion jumped from 64.06% to 69.65% (+8.7% lift). For new products with multiple inputs, text lift reached +14.2% and image lift reached +22.2%.
• Text and Name personalizations converted best in absolute terms.
  Name inputs improved +2.57 pp under single personalization, and Text improved +6.85 pp overall under multiple personalizations.
• Users converted faster in almost all flows.
  Overall median time dropped from 1.4m to 1.2m for single personalizations and 46s to 38s for multiple personalizations. Image personalizations under multiple gifts saw the biggest speedup (-30%).