Gift Giver is a shopping assistant which provides gifts suggestions based on a person’s previous purchases. Think of it as an advanced wish list which is also capable of sending alerts and reminders of upcoming important dates. Earlier this year I had the privilege of working on this project.
As the UX designer I worked on all stages of development from user research to prototyping and usability testing. Come with me as I take you on a journey of my process throughout this design project.
Prior to research and development, I preliminary hypothesised that gift giving is a personal activity. Most will dread the gift search but will persist to reach the ultimate goal of pleasing the gift receiver at the moment of handover.
Going forward the strategy was to focus research on understanding user gift shopping habits, pain points and the emotions attached at each step of the shopping process. Development aimed to resolve pain points and further refine existing processes.
To test my hypothesis, the initial aims or user research were to understand if the app would be of use and to interrogate current gift shopping habits. To answer the two questions, contextual inquiries were carried out on potential users within the age range of 20–35. Key focus was placed on participants goals, points of enjoyment, pain points, and how they currently worked around them.
Findings and observations were gathered consolidated into an affinity diagrams. The four insights gathered from these diagrams were:
· A general enjoyment of the gift-giving process because of the ultimate happiness of the gift receiver.
· A dislike of how time consuming the process can be.
· A reliance on reviews when ultimately deciding on purchasing.
· Either forgetting about upcoming dates or buying a gift so late they risk not receiving it in time for the important date.
Participants were then grouped by similarity in needs and goals. Among the two groups one came out as the primary persona group as it had additional goals and needs unique of those of the other group. The ‘Shopper’s additional needs included: video and written reviews, deals and calendar alerts.
Understanding that human shopping habits have been established for years and it was unlikely I could entirely change them, my first step was to look into what the market currently offers and how it works within rather than against existing habits.
I carried out a competitive analysis of apps interview participants had mentioned when describing their current gift shopping process. ASOS, Amazon and Ebay where the few constant mentions. I focused on what features each app provided to enable a shopper to go from product search to checkout.
A number of features outside the essential were found to help in resolving user pain points, this included:
· Extensive sort and filter parameters allowed curated product search results
· Ability to add/edit address and payment info within the checkout page
· Customer Questions & Answers covered information missing from customer reviews
In my continuous aim of narrowing down features, I used red routes as a method of prioritising the exhaustive list.
Following feature consolidation and prioritising features came developing an information architecture. The first step, card sorting. The initial set of cards were used to show relationships between cards and collect user language friendly group headings.
During the initial round of the card sort, users showed aversion to some of the terminology used as a result a number of adaptations were made before the second round:
· ‘Account Creation’ removed as it doubled in function with ‘Manage User’
· ‘Favourite’ replaced VIP
· ‘Options’ was labelled ‘Manage Notifications’ for clarity in purpose
Second set of cards had predetermined group headings and focused on cards to group relationships.
Ultimately card sorting data was consolidated into a draft information architecture:
Knowing the draft was just that, a draft, I decided to formative test the flow using a paper prototype and the persona’s user story.
“Your name is Eli Gill, you have a best friend Vanessa Hudgens whose birthday party is coming up soon. You want something amazing to take to the party. You know she loves technology and that she’s getting bored of her old gift an iPhone 8. As it’s your first-time using Gift Giver to find Vanessa a gift, set her up as a new favourite person and buy her a gift that is highly rated and reviewed.”
Following testing I summarised the key problem points as:
· Lack of clarity around flow of new favourite person creation.
· Awkward transition from person creation to product results page.
· Irregular order of features on menu bar.
Consolidating feedback on the successes and failures of the primary task flow, I continued to iterate the information architecture until reaching this final iteration.
Designing and testing
The time finally came for me to put my creative chops to the challenge. I created wire frames to fully explore the other task flows that would take place within the app. I carried out a number of formative testing sessions to allow for multiple iterations that would further refine the task flows.
The key issues found following testing:
· Confusion with how the ‘Previous Purchases’ feature functions
· Unclear language used in instructions
· Clarity around which delivery address is used at checkout
· Order Summary not providing enough details
In further iterating I upgraded to high-fidelity mock-ups to replicate a more to true to final product experience. Once again, I had participants take part in formative testing to guarantee there weren’t any glaring outliers in the design.
The key insights following testing:
· Previous Purchases step often felt like a never-ending task that would benefit from a progress tracker.
· Gift results page felt heavy with information bloat and inconsistent use of colour.
Testing against metrics
In the final stages of design, I came to an understanding that to really test the success of the current iteration of the design, I would need quantitative irrefutable data. To do this I carried out summative testing with 5 participants comparing the at the time current iteration vs the wire frame iteration.
The key focus was getting data that showed evident improvement of the app’s work flow in areas of effective, efficiency and satisfaction.
Ultimately the percent decrease across time taken, number of taps and requests for assistance show an overall improvement in effectiveness and efficiency between iterations. Continued high levels of satisfaction was also maintained proving a continuous improvement across iterations.
The process was a tough one but there were a number of key takeaways when designing around shopping habits that I learned from the process:
Ease the checkout process: Shopping habits of the people interviews showed that the buying stage in the gift giving process is either the most rushed or least paid attention to. Around this stage, the user is bored or exhausted and wants to complete the process and move on often leaving them open to making errors. Design with the aim of easing this process and closing the margin for error. Strip the page of bloat and leave strictly the information and instructions necessary for task completion.
Buying is the ultimate goal: People bear the often-tedious gift-giving process because of the end result. The point of purchase is the most pivotal part and as such should be made available to the user on as many pages possible — of course where relevant. Making the ‘Add to Bag’ button evident and continuously available is great way of achieving this.
Presenting ‘Gift Giver’
Try the prototype at: https://adobe.ly/2ra1qMu
Thank you for reading! I would love to hear your comments, feedback, or if you just want to say hi.
Designing a personal shopping assistant app — a UX case study was originally published in UX Collective on Medium, where people are continuing the conversation by highlighting and responding to this story.