Divide and Conquer - A Startup that Rents Space
A redesign for a faster more flexible home page with simplified yet well thought out categories to improve the user experience.
This project was conducted over a two-week sprint using Agile. It primarily ran from a kanban within Trello as we worked on site with the client and our team base. Visibility was a minimum given the rapid timeframe of the project. We began this case study at the first meeting with the client. We were tasked with improving the categories that define the spaces within the platform and to help improve the onboarding process when it comes to listing a space.
To understand the problem, we required an understanding of the user. During the empathetic stage of the design process we required some existing data from the client to help us map insights and put ourselves into the shoes of the potential guest and host.
I focussed primarily on heuristics and desktop research.
There was an absolute magnitude of personal stories involving online booking engines. The primary source for these were Airbnb, Splacer, RealEstate and Gumtree.
Unsurprisingly, many of these platforms had very similar user interfaces and seemed to all follow a distinct methodology when booking online.
We leveraged all the insights we had gathered through our desktop research. The method behind this was to identify objectives which address the overall scope of the client. I worked with a colleague to generate this script using a very structured and
task-oriented approach. The intent was to methodically extract the participant’s feelings and needs at every key point in the experience. The primary objectives were to determine how the participant viewed the categories and the meaning behind it. We also aimed to see how they felt about the listings page.
The follow on from this was the actual user testing, how would you use the platform. The tasks finding a desk space, picking the space that best suited them, and exploring the specifications that matched that space.
With limited participants provided by the client (who were also only available off site) when it came to card sorting we had to figure out how to reach more participants in order to increase the credibility of research data to analyse. Our solution was creating an online unmoderated card sort.
This allowed us to have more participants and spend less time on traveling offsite, setting up the card sorting activity and recruiting users. We also save time on the analysis of the gathered data. Online card sorting tools do a really good job of outputting results to a spreadsheet helping us to come to a conclusion more efficiently.
As a collective, we affinity mapped the test responses to start deep-diving into insights. The affinity mapping itself was quite good. We segregated it into user testing and card sorting. This helped us obtain the understanding of the overall problem rather than mix the two sets of research information.
Undoubtedly the most difficult experience of the sprint. Absolute collaboration required with diplomatic and fearless communication. Coming from a background in design I felt the need in this instance to push everyone a little further. Generic responses were being pushed up that weren’t particularly addressing the base need that our notes were telling us. I was consistently asking, ‘but why, why do they say that, why do they think that’. I feel it got to the point that it was resulting in frustration however this is the crux of research. Identifying the correct insight and not taking shortcuts.
In terms of the actual usability testing, there was an overwhelming desire for discoverability. The participants had no context of the application and how it would work, however, it was clear what the participants were in dire need of. Due to an NDA signed at the start of this project, those insights discovered can not be open to the public.
For ideation we decided to invite our clients to the team base for a co ideation session. Once again we presented our insights and asked them to base their ideation on a specific insight they wanted us to focus on. This proved to be very successful as it changed the path of our second sprint to the “host”, rather than the “guest” as we had originally planned. Our clients then did a ideation activity called “6 up, 1up” where they narrowed down the design they pictured for their platform based off the research and insights we had presented. We used this to our advantage and decided to ideate upon what we had received from them. I loved this part of the process as it provided us with a lot more clarity around the direction they wanted us to go for sprint 2.
Due to absentees in our team, I decided to tackle wireframing, information architecture, leveraging existing platforms and desktop research to come up with a baseline wireframe. This then led to the initial phase of ideation of the user flow. What might we expect the host to do when listing a space online. What might they expect to see and where might they expect to see it?
Before wireframing, we needed to revisit our objectives and see if there had been any changes since conducting our research. We looked through our affinity mapping and decided that these would become our 3 new objectives for our wireframing:
1. How do the 4 categories resonate with people
2. Is the current platform what you would expect when listing a property
3. Does the process of listing specifications match your expectations
The process of wireframing involved identifying the key elements we wanted the users to see to impact their relevant emotional responses around the insights we found.
When considering the wireframing, the emphasis was placed on the logistics, empathetics, objectives and introducing a mechanism of interaction. The three primary elements included social media trends, a community of like minded professionals and your personal profile.
The idea I alluded to was that the functionality for the website needed to be based on the “push and pull” methodology. The user needed to have more control on the avenue they went down rather than be lead down a path that might not be for them. This was done by offering the host a feature where they could easily categorise their listing in their own words, rather than being forced by the preset options.
Prototyping was a collaborative effort. It culminated in a medium fidelity prototype which was then used for user testing. A lot of the smaller quick wins such as button hierarchy, headings and copy were ironed out.
The main effort involving the prototype was the script forming for the user testing.
Same format as initially, objectives followed by the route however this time; specifically by each screen.
Developing a prototype is an intimate process which isn’t always best worked with 4 people aiding it, rather everyone working on the same individual thing. By having a singular person design the prototype, the remaining members of the team went into the user testing with no real exposure to the prototype or its core functionality. Big mistake.
We gained many useful and interesting insights from the testing, which in term presented many key problems that our low fidelity prototype had occurred. Although the initial thoughts of problems being found may sound disheartening, it was quite encouraging as it just helped iron out our product before presenting back to the client.
User testing and Insights
When conducting our user testing, we had our participants to place sticky notes on print outs of the prototype so they could have a more hands on approach on where the pain points lay. We did this by using a colour coding system; yellow for tangible components, and blue for feelings on a certain component. This turned out to be a really successful method for us.
We then used those insights and went back to our prototype to make the necessary adjustments. There were some clear patterns coming out of our testing. We discovered that users found the categories “Office” and “Creative” too confusing and wanted more context around this. As well as this, the drop downs made people unsure of the relevance the sub categories had. This was due to the main headings not being properly clarified.