Shiyun, Tengwan, Jianwei, and Junhan are product designers at Zopim; responsible for designing and creating new features and products. While aesthetics and beauty are important elements of design, these four ensure that products are not only attractive but also functional for users.
Here, they share the design story behind our most recent launch: Zopim Premium.
When it comes to design, our overarching goal at Zopim is to create beautifully simple software. We want to build the tools that enable customer support professionals to do their best work.
So, when we first whiteboarded Zopim Premium, we focused our attention on providing supervisors with a more holistic understanding of their team’s productivity and performance. This meant building a reporting tool that was simple to use, and could provide actionable insights. Enter our newest Analytics product — Agent Reports.
While the design team may have its own pre-existing opinions , we always put a new project through a rigorous design process, starting from understanding our customers’ needs.
Step 1: Empathize — Stepping into the shoes of the user
Design doesn’t start from a void. It starts with an empathetic understanding of customer needs and wants. Thus, as tempting as it is to jump into development for features right away, it is important to pause and ask “who are we building this for, and why do they need it?”
But, as Steve Jobs famously proclaimed, users are often unaware of their true needs and wants. Thus, before designing Agent Reports, we ran both investigative surveys and interviews to identify the potential pain points of our users.
The preliminary survey revealed that users required more help managing their workforce, so we focused interviews on trying to find out the exact functions they needed. Filling in this gap would help us to generate hypotheses about the actual demands of our users, so we could develop a tool that actually helped admins fulfil the jobs they needed to be done.
The interviews were illuminating. For a long time, we had been reluctant to surface the length of time agents had been logged in to the Dashboard. We believed that this metric led to micromanagement and the stifling of agent autonomy. Through the interviews, however, we realised that we were mistaken. While there was potential for misuse, this was an important metric that allowed greater accountability and better scheduling.
Support managers had a number of key questions that they needed to answer and the agent logged in metric made their lives a lot easier. For example, how could admins uncover which agents needed more guidance, if they were not even sure who was actually online when chat performance dipped? How could admins schedule agents better, if they did not know how many agents were online during peaks and lulls?
Step 2: Iterate — Lather, rinse, repeat
Armed with the interview insights, we came up with several ideas for Agent Reports feature. The most promising were quickly prototyped.
We sent out the prototypes to users, and scheduled more interviews for validation. Flaws were quickly uncovered and ideas were rapidly discarded. If there were signs that the prototype was not actionable, comprehensible, or actually useful to the user, it raised an alarm about the design. It meant that our original decisions had to be reevaluated: the solution we provided, the assumptions we still had, or even the user needs that we had hypothesised from the interviews.
Step 3: Growing with our users, and guiding them along
There are many alternatives to designing a successful analytics product. A simple option would be to just present all possible metrics to our users and leave it to them to interpret. Such a tool could be incredibly powerful, but it would most certainly be irresponsible.
Zopim is used by a wide spectrum of users. Some are skilled at navigating data and have specific metrics they want to track (for these users, we decided that providing APIs would enable them to build their own custom dashboards). However, during the interviews we learnt that many of our users have an idea of the goals they want to track, but need a helping hand along the way. While such users are growing core business operations, they are also trying to scale their chat teams. Workforce management is hence the biggest issue many grapple with.
To provide the most value to these users, our Agents Reports product needed to provide insight into where their chat support was lacking, and allow admins to pinpoint the specific areas for improvement. Providing all available metrics, without any proper structure, would overwhelm our users and only make it harder for them to leverage on analytics to achieve their intended goals.
So, our greatest challenge was to identify the story that Agent Reports would tell about our users’ chat support teams.
These considerations led us to the design of Agent Reports — a tool to inform and help admins investigate exactly how they can improve their chat support team, with insight into individual agent performance, drill-ins to chat history, and agent/chat sessions.
Every designer begins the design process with an idea of how they want the final product to look. But before getting stuck in the weeds, it’s important to question our initial assumptions and figure out what our customers actually want.
This is why we ended up creating an analytics product that not only informs, but also helps users take action and improve their team’s’ performance.
Designers are human too, and it’s almost impossible for us to come up with a perfect prototype on our first try. So, it’s important to incorporate iteration and customer feedback into the design process. It may seem tedious, but the rewards of seeing our users reaping the tangible rewards from Agent Reports makes everything worth it.
Latest posts by Chow Li Ying (see all)
- Chat Routing: Help Your Agents Manage High Chat Volume and Increase Satisfaction - January 26, 2016
- Zopim Agent Essentials: An Introduction to Live Chat - November 27, 2015
- The Finest Chat Experience on Mobile Devices - November 24, 2015