Problem statement
Support Centre, the go-to resource for new and existing SumUp business owners. While the company evolved, the support centre has not quite scaled to meet the growing needs.
According to the feedback we received in a customer support survey, merchants found it difficult to get relevant contents to their questions in the support centre. They contacted customer support and ended up spending more time to have their enquires solved. From a business perspective, this issue is overstretching already busy customer support channels.
Methodologies
At the beginning of the research, we carried out usability tests with interviews to learn about issues that occur to merchants when using the support centre for specific tasks. We observe their behaviours while they perform predefined tasks including, account management, setting up card machine, and troubleshooting, etc.
Participants Criteria
Berlin merchants who contacted customer support in the last 90 days. We aim to have participants from different business type.
Duration
60 mins
Participants
👦Fashion store owner
👱‍♂️CTO of event tickets platform
👱‍♀️Irish pub owner
🧔Meal kit business owner
Measurement
⏱Time on task
âś… Success rate
đź’™Subjective satisfaction - ease of use
Current design
Tasks
Find out how to change bank details
Find out how to order a card machine replacement
Find an article to troubleshoot your blocked card machine
Main insights
We’ve noticed the current design performs significantly poor at helping merchants solve hardware and product issues compare to other areas.
Weak info scent - participants were struggled to understand what support materials are included in each category.
Promote recall over recognition - when users were reading an article, they need to rely on their working memory to remember which category the article belongs.
Inconsistent search functionality - the feature is sometimes visible and sometimes not or move around.
The term “search” primes users into thinking the search field will bring them to google search result instead of searching inside the support centre.
Only 1 out of 5 participants were able to find support materials about hardware issues.
Avg satisfaction rate - ease of use: 2
Change statement
Based on the insights, the current system performs relatively poor at helping merchants to solve hardware and product issues compare to other areas. We proposed a redesign to improve overall information scent to reduce the amount of learning.
Hypotheses
- Having a dedicated "fixing an issues" category upfront will emphasis the context and speeds up users' comprehension
- Rearrange materials and rename some of the categories or incorporate visual cues will prime user in the right direction
- Quick access to support articles from landing page will reduce users' effort to find the correct info they need
🪠Design objective
To make sure merchants can find relevant contents to their questions with the least effort. The design needs to build on users' existing mental models, give users clear directions throughout the experience and ensure contents are easily accessible.
🎯 Business objective:
The more time support agents spend to resolve merchants' enquiries, the higher it costs for operation. We will know the redesign has an impact on business when we see a decrease in support touchpoint. We expect to reduce that by *2X% by mid 2021.
Proposed design
The redesign should build upon the mental model users already have based on their past experiences using other services' support channel.
Utilise familiarity - reuse visual clues merchants encounter in other SumUp products to reduce the amount of learning and make the interaction intuitive.
Emphasis content to speeds up comprehension - the design emphasis the support materials to help users find the answers they need
Improve overall information scent - inform users the context of each category to primes them in the right direction
Revise category titles and visual clues - renamed categories and rearranged representative items based on the result of card sorting sessions
Quick access to support material - help users to form an interpretation for each category without inspecting in full detail
Evaluate our hypotheses
Methodologies
5-second test to gauge the merchant’s first impressions of the redesign
Qualitative usability testing - we ask participants to perform three predefined tasks using the redesign. We observe the participant’s behaviour and listens for feedback after they finish
Quantitative usability testing - we collecting metrics that describe participants’ experience so we can compare the new design to the old one
*We ask participants to perform the same tasks we used to test the old design
Participants Criteria
Berlin based, english speaking merchants
Contacted customer support in the last 90 days for card machine and account issues
Duration
60 mins each session
Who did we interviewed
🧔Staff in Italian restaurant
👨‍🦰German pub owner
👱‍♀️Owner of art gallery
👨‍🦳Restaurant owner
👩‍🦱Bakery owner
Measurement
⏱ Time on task
âś… Success rate
đź’™ Perceived ease of use
Testing insights - compare the proposed design to the current (old) design
Participants found the redesign perceived as trustworthy in general.
The amount of content in each article discouraged Participants from reading.
We see time on task reduced when participants try to find instruction for changing bank details.
We also see time on task reduced when participants try to find support article for hardware troubleshooting.
75% of participants went to "Fixing an issue" when they are asked to look for information for card reader issues
75% of participants attempted to search for contact info in the search field. They prefer to have them on the front page
90% of participants prefer to reach to support team then look up answers themselves when experience issues with their card reader.
Support channel preference - Apart from hardware issues, support chat is the go-to channel if available.
What happened next?
The research and findings were passed on to the project team and content team as rationales for further design decisions. The uncovered mental models were taken as design principles throughout the process. Here's the final design as of @July 31, 2020