This video introduces the series of extensive user tests we recently conducted, studying a set of 14, mostly larger, central London, estate agent websites. We did this with 10 real buyers, sellers, landlords and tenants, all of whom had changed hands on property within the last 3 months. Taken together, the user testing series provides a wealth of valuable insight, identifying both some big themes and guidelines, but also dozens of minor-usability “gotchas” or useful bits and bobs.
Before we dive into the tasks, which include design first impressions, views on site navigation, finding phone numbers, searching for property, or deciding who you would trust to sell your house, it’s worth understanding the principles of the user testing first. So, pop the kettle on and pull up a chair…
This first task looks at how people respond to the design of the agents websites, from a purely visual appeal perspective.
Each user is given 20 – 30 seconds looking at each site completely afresh. We rotated the order in which we exposed the sites to each user.
And we then asked them for snap reactions: What they liked? What they disliked? And for their scores out of 10, which we average into a “design ranking”. At one level design is purely subjective, but at another level it’s science, and in this video analysis, we see some really interesting themes emerge.
Next up, in the second video, we ask the same set of users to spend a bit longer, navigating around the websites.
We explain what we mean by navigation systems to them, before they start, asking them to explore the menus, the drop downs, the sections of the website. Not looking at the content they discover, but exploring how the “navigation system” itself works. Do they like it? Is it easy to get around? Do they feel lost? Could they find what they were looking for.
We chose the sites quite carefully, to include some with single tier navigation, others with drop downs, some with split-navigation systems, and yet others with the new “mega-menus” or panel based navigation systems.
Following this we turn to the main navigation search pathway through the websites.
With this task, we we gave a looser brief to our users. We asked them to search for a house, either for sale, or to rent, depending on whether they had recently rented or bought. We gave them the town to search in, and set their budget (as it’s hard to compare the search experience across different agents, who occupy different market segments in different areas, without providing this guidance).
But beyond that, we gave the users wide scope. We asked them to dig around a bit, find a house they liked, perhaps save it to their account, certainly look at the photos, floorplans and map. Then, when they had chosen the property they liked, we asked them to make an enquiry to the agent to see how simple that was. The results are less structured, but they turf up dozens of “gotchas” and great little tips and ideas. Pure gold.
In this fourth test, we simply ask the users to find us a branch phone number to call. We tell them to imagine they were driving through “Richmond” (we give them the town name for each agent, to make the test fair) and we ask them to find the contact number as quickly as they can. What we liked most about this task was how some of our users completely failed to find the phone numbers, on some of the agents site. Yes, they actually failed!
Fabulous value could be released by fixing these points, and the testing gives several really nice examples of quick simple methods that users found easy, to get to specific office branch phone numbers of even really big agency groups.
Most of our user testing was done on desktop (despite the importance of mobile). This was simply due to the time available, and the screen recording software we used.
However, we repeated the fourth key task of finding a branch phone number on a mobile phone, and on an iPad. We did so, as to ignore the mobile use case would have been sinful. And we suspect that the “find a branch phone number” task, is even more primary a use case on mobile phone, than it is on desktop. The variation in quality of solutions was far wider between agents, once you start looking on a mobile phone. Some were great. Others, not so much!
And of course, as agents, we’re principally worried about trapping more vendors using our websites. So, in this 5th and final task, we ask users to choose the agency they would most trust to sell, or let out, their house.
We gave each of them a smaller set, 4 or 5 agents, and we tried to roughly match the agent market position to their genuine selling position. We also tried, where possible, to pick agents outside of their actual market area, so they could leave as many pre-conceptions at the front door as possible.
It’s hard to be truly scientific clearly, and in the real world the website first impressions for selling, will only be part of the wider brand impression and story, clearly. However, your website has it’s role to play in “selling” to that most essential of customers, the potential vendor / landlord instruction. And here’s what our users thought of each agent.