How can Airbnb improve its website user experience for top-of-funnel users
Let’s go somewhere. Let’s travel to New York. We can visit all the art museums you have been dying to go. We can move to Tuscany. Imagine reading your favourite book in our backyard on a lazy Sunday afternoon. Don’t be afraid to dream. I am here to help make your dream come true.
As part of the rebrand, the redesigned Airbnb website talks in emotions. Communicating through inspiring photos and heartwarming words, the website is a spokesperson for Airbnb – compassionate, versatile, and most importantly, always there for you.
It is encouraging for us designers to see Airbnb focusing on the emotional aspect of its product. The effort Airbnb has put into its rebrand catches my heart. As a user, I embrace the humbleness and subtlety of the new website. It is exceptionally well designed and crafted. Unlike other travel sites, it never disrupts users with any conversion-driven messages.
This approach works in nurturing loyal audiences like myself. However, I question its efficiency in engaging new users, especially travellers who are in planning stage. Having the thought in mind, while my partner Jac and I were looking for places to stay in Europe next year, I observed how Jac interacted with the Airbnb desktop site and the findings were quite surprising.
Top-of-Funnel-User Exposure – Airbnb Desktop Site
Background: Jac and I were planning to stay with Airbnb in our trip to Berlin and London. We would also like to explore another city/town, which can be anywhere we find interesting on Airbnb. Dates were not confirmed yet. Our priority in this user exposure session was to get a cost estimate. Jac had stayed with Airbnb before, but it was her first time using the site. Jac didn’t have an Airbnb account.
The ‘Persona Jac’
I will be doing a lot of user exposure with Jac as I think she has the perfect personalities and behaviours for any customer-facing digital product testing. Jac is one of those who are stuck between the Gen X and the Millennials: curious and willing to try new technology, 50/50 computer/mobile user, uses apps to be informed and improve efficiency, moderate online shopper, writes notes in notebook, reads heavily on printed books and magazines, critical, impatient. Probably one of those hard-to-please customers but any smart, small detail could make her embrace a product.
Type: first-time engagement (top of funnel)
Goal: get an idea of travelling budget and potential places to stay. Images, prices and locations are the main information Jac will focus on.
Tasks: 1. Compare prices from quality one-bedroom apartments to share-accommodations, in London and Berlin (‘Search’)
2. Find an interesting place to stay outside of London and Berlin (‘Browse’)
Discoveries in Jac’s User Journey
Major Visual Changes in the Past Month
1. Website Layout
The past month saw a swap in the positions of the main content and the map. The main content now occupies the left column, which accommodate the F-shaped reading pattern, while the map takes up 2/5 of the browser width on the right. Collapsible search filter stays above the main content.
An observation: in the previous layout, although the main content was placed on the right column, Jac quickly adapted to the layout; map was very rarely used, but when it was used, due to unfamiliarity with the places, she showed frustration and confusion and soon completely ignored the map.
Question: does the map need to be visible at all times? What is the percentage of users who use the map? Moving the map to the right shows that users don’t primary consume the map contents. But would it improve the user experience if the map could be hidden so that it gives more real estate to more useful contents?
2. ‘Size’ Filter
A frustration Jac experienced was the inablity to narrow down results to one-bedroom properties. As she slid up to mid-ranged prices, results started showing multiple-bedroom properties. Jac felt a waste of time and effort eye-filtering contents that weren’t useful to her.
This must be a problem for many users as now the ‘Size’ filter has been added and it is meticulous. It definitely helps to tailor results to meet user requirements – not only for customers such as Jac, but also people who need 7.5 bathrooms – perhaps a family of seven and a dog? (I mean, seriously!?)
Overall the new ‘Size’ filter performs fine, but error handling is poor when there is no result that matches filter settings. Instead of informing user, the system comes up with random results.
3. Invite Friends
Another change is the mustard yellow ‘Invite Friends’ button at the top, sitting next to the existing ‘List Your Space’ button. A business-goal-driven change but not as aggressive as many other referral marketing we’ve seen. The mustard yellow is quite passive. The primary action for ‘Invite Friends’ is to give Airbnb access to all your email contacts, which many could hesitate about. I would love to see how well this button has been performing and the abandon rate of the call-to-action page.
Keeping my eyes on further changes to these features.
Flexibility & Efficiency of Use
1. Search Bar & Error Handling
The Google suggestion came in handy when inputting destinations, especially for long city names. However, there are a few things the search bar fails in providing efficient user experience:
• Hitting the ‘Enter’ key to submit form, a habit of many users, is not accommodated, which seems to be intentional as it forces users to go through the form before searching.
• Do users have to fill up the form to see the results? Interestingly, user can search properties without entering the check-in/check-out dates, but it wasn’t indicated. A complaint I heard from Jac – ‘I don’t know what dates we are going. Why do I need to enter this information? It is annoying’. Because she was so impatient, she hit the ‘Search’ button without entering the dates and it worked! User problem self solved.
This is only the start of the journey. Would it help make the user journey started more smoothly by adding an option to not include the dates? A good example can be seen on Booking.com:
• Except some extreme criteria, the Airbnb search engine seems to be able to find ANYTHING – But fails sometimes, for example:
2. Search Filters
I think the search filter design is a winner. It is a combination of the top filter and left-side filter that is tailored for Airbnb contents. This expandable filter cleverly creates another layer of space for the large number of options, which are grouped into smaller collapsible categories. The flat design and typography also helps ease users’ cognitive strain when handling complex information.
What I have discovered in Jac’s user journey:
The most used filters were ‘Room Type’ and ‘Price Range’, which are always placed at the top of the page. Filters such as ‘Amenities’ and ‘Property Type’ were redundant as they are irrelevant to her priorities. She showed frustration when not being able to specify the number of bedroom (now added) and minimum stay. An assumption is that minimum stay is negotiable, yet it is still frustrating seeing a minimum stay of seven days when the user only looks for a stay of two days.
The triangle button for the drop-down filters such as ‘Neighbourhoods’ was completely missed when narrowing results to ‘Shoreditch’. When the filter was expanded, Jac was overwhelmed by the 100 alphabetically sorted neiboughhoods on the list. After ‘Shoreditch’ was found, Jac decided to add ‘Soho’ which was at the time placed at the top outside of the drop-down list. Unsurprisingly, ‘Soho’ wasn’t found easily.
Such long list is inefficient. First of all, the three top neiboughoods that are isolated from the alphabetical list could be easily missed. Secondly, scanning through a list of mostly-useless information line by line, adding the unfamiliarity (to travellers), could be frustrating. Perhaps it would improve the speed of scanning if the information is read by column? Or simply a search box with Google suggestion?
3. Result Sorting
Results seem to be displayed randomly without any sorting options, giving the primary focus on the photos of the properties. I see it as another subtle design that is determined by business goal – Mixing the results, it does not only gives the properties more exposures, but also increases the chance of selling the more expensive ones through comparison as the expensive ones usually have better photos. But is it user-friendly? Probably not as Jac still mentioned how inconvenient it was weeks after the user exposure session!
Content Findability and Discoverability
1. Neighbourhood Guide
While Jac was browsing properties in London, a question arose – Which areas should we stay? I suggested her to read about the Airbnb neighbourhood guide, which I have seen on its old website. It was a summary of a city’s neiboughoods and I found it very useful when I was planning my travel earlier. ‘But where is it’, Jac asked, failing to find it. I then had a look around and realised that it is hidden somewhere where Jac would probably never encounter if she didn’t know such guide existed.
But an interesting thing is that, Jac would probably open a new window, Google ‘London neighbourhood’, find the Airbnb neighbourhood guide at the top of 89,400,000 results and go back to the Airbnb site. It would be funny if the result was paid advertising!
2. More Hidden Contents
The only thing I like about this ghosty ‘Browse’ button is that it leads you to the most wonderful place in the entire Airbnb site. Its well-curated, themed lists always excite me; properties featured on the landing page are often surprising and different to what we normally see in ‘Search’.
But why do we have to treasure hunt such interesting contents (on the Airbnb mobile app, these contents are showed on the first page)? Now let’s look at how boring the homepage ‘Browse’ section is, ironically.
3. Propose to Redesign the Homepage ‘Browse’ Section
This section seems to miss the opportunity to engage users with interesting and relevant contents:
• Beautiful images of cities such as Byron Bay (location-based), New York and Paris serve as quick links to the lists of properties. For these obvious-top-destination cities, while the search bar is placed above the fold, does everyone need them as quick links?
• Contents on this section is deadly static – They are not very often updated. Rather than ‘Browse’, ‘Quick Links’ seems to be a more appropriate label for this section.
I assume that this part of the site would be more useful if contents are dynamic, based on user preferences and behaviours. Contents can be a mix of popular destination quick links and curated lists from the ghost ‘Browse’ button mentioned above. This will also give the opportunity to advertise wish-listed or viewed properties.
Sign Up & Wish List
In this user journey Jac didn’t end up signing up nor creating any wish list. The uncertainty about this trip was a reason. But Airbnb’s passive engagement with users should also be factored in.
Booking.com’s wish list is a good example for short-term users. The wish lists can be created by users without signing in. The cached wish lists then serve as a reminder in the users’ future visits. This approach lowers the barrier for users to remember favoured properties, but also for the site to track users behaviours.
But don’t get me wrong. Airbnb has the best aesthetics among all travel sites and the service it provides has been highly praised by many hosts and travellers. It only shows its weaknesses in customer engagement, when compared to other travel websites. What we haven’t seen from Airbnb is real-time engagement with users, for example, onboarding users with messages that take them through the website interface; instant chat for host to answer traveller’s questions; if new user spend more than a certain amount of time on the website, push them to create account and save properties to wish list.
Such features could easily clutter and destroy Airbnb’s minimal design. However, if they are implemented consistently and thoughtfully within the brand aesthetics, Airbnb could possibly expect to see more satisfying customer experiences and bigger sales.
The Loop of Digital Customer Journey
Jac’s customer journey ended as soon she put her hands off my computer and probably not going to continue in any time soon. I was then expecting Airbnb’s behavioral targeting to roll out, while I was searching for places to go in Europe, looking at friends’ photos on Facebook, reading news on Fast Company. Surprisingly barely anything. Airbnb doesn’t seem to invest much in online advertising, at least in Sydney if location is related. There was a couple of text-based Google AdSense ads in Chinese popping up randomly, which I am sure that it would fail to communicate with anyone. But how do they know that I can read Chinese? And why do they only show ads in Chinese? It is just bizarre.
What about Email marketing? Unfortunately I opted out from Airbnb’s newsletter, so none of its email marketing has reached me.
As its new feature ‘Invite Friends’ suggests, Airbnb needs new customers. While these new customers land on its website, it needs to provide human-centred user experience as its new branding promises. More importantly, it should inspire customers to endorse its service. It all comes down to who the new customers are and their personas, user behaviours and priorities etc. Jac is one of the millions whom Airbnb tries to attract. Her persona and behaviour could present either 10 people or 10% of the new customers at a certain time. What I have recorded and suggested is simply based on the desktop site activities of ‘Persona Jac’, without factoring in many other elements. With mobile users on rapid rise, although it has a beautifully designed mobile app, Airbnb still has to improve its mobile site user experience. I am looking forward to seeing how the brand and its products evolve.
Now over to you …
How was your experience on Airbnb website? Let’s continue the discussion on Google+.
Note: All images above are screenshots of Airbnb website, www.airbnb.com.au. All rights reversed © Airbnb.