Diving into the chaos: a design case study
Dive into the transformation of Casavo’s home estimate page, as design and content unite to elevate user experience. Navigate through the internal challenges, user interactions, and the shift from a cautious approach to a bold ‘let’s redesign it all’ stance.
Table of contents
- Who is Casavo?
- Why are our sellers unhappy?
- Internal storm
- “But we don’t have time for research”
- “We’ll check the copy later”
- In a nutshell, what has changed?
- Conclusions
1. Who is Casavo?
Casavo is the next-generation European platform that matches home sellers’ needs with homebuyers’ desires through an integrated and personalized experience.
Our mission is to change how people sell, live, and buy homes in Europe. To achieve this ambitious goal, we are redesigning the home selling and buying experience, piece after piece.
2. Why are our sellers unhappy?
Sellers can get a home value estimate on Casavo. They answer several questions and then get a fair market value for their house. Here is what they saw:
Following some shadowing with our advisor, we found out 4 main issues for sellers when seeing this page:
- Sellers don’t understand what the ‘Fair market value’ is, and always think it was too low.
- Sellers got an instant valuation that doesn’t take into account important features, such as garage and house photos
- Many sellers don’t trust algorithms, especially for something as personal as their homes.
- Sellers expected Casavo to buy their houses
3. Internal Storm
Fortunately, we were already at a point where we were supposed to update this page.
Here is our tribe Q3 2023 OKR:
- Objective: Educate, guide, and serve all customers starting their journey with Casavo in a more efficient way
- Key Result: Increase by 30% the number of users completing our home estimate flow move forward in the process (data gathering completed + buyer profiles)
Our work on the restitution page content (the internal term for the home estimate result page) was low-effort but highly impactful.
However, we didn’t know where to start. Some even suggested eliminating the instant valuation and keeping Fair Market Value only as an internal tool.
We started by examining competitors, engaging with other departments (Marketing, Operations, Legal), initiating brainstorming sessions, and doing what we do best: talking to users.
4. “But we don’t have time for research”
Working on such a crucial page takes a lot of time. And our developers were left with little to do.
The team initially claimed there was no time for research, but then happened to say: “This is the most important project of 2023”.
And we played our Reverse Uno card: “If it’s so important, how can we do it right without research?”
Here are our tips for research when time is short:
- Make research your only priority, and squeeze it all in a very short time (on our side, we had 7 interviews in 4 days)
- Pair up to ease stress and take turns talking to users
- Quickly tweak the prototype (both the copy and design), and note all feedback and next steps for each section
- Divide the prototype into smaller user stories, and assign the ones you’ve finalized to developers
5. “We’ll check the copy later”
Repeat if after me: testing a prototype with Lorem Ipsum (or low-fi copy) is useless.
Real data
Just as you can’t get genuine feedback from placeholder text, you also can’t get emotional feedback from fake data. There’s nothing more valuable and personal than our own house. Since the home value estimate is primarily composed of data explaining the worth of your home, reliable feedback — both emotionally and in terms of experience — can’t be obtained with fake information.
We used real numbers based on what the seller has already seen to check their reaction when talking about THEIR home.
Let them speak
Here are a few content-related questions:
- What happens when you click this button?
- Can you explain this number?
- Can you describe this page?
- What else do you want to see here?
We created a table where we put all user verbatims to remember exactly the words they use.
What about “no copy”?
We tested something. I am not sure it was a method by the book, but we did it anyway.
We presented some CTA with no microcopy on it. This approach aimed to understand what users logically expected to find behind a button and the words they used to talk about it.
One user, upon seeing the ‘?’ below the price, said clicking it would allow them to print the home value estimate. This reinforces that they thought the experience ended there.
We asked another person what would happen if they clicked the “?” below the “Garage, expenses, photos?”. They said, ‘I will land on a screen where I have to put information and photos’. When they clicked the button, their ‘yes indeed, just as I imagined’ made us understand that the title and subtitle gave them enough information to anticipate what would happen next.
6. In a nutshell, what has changed?
In June, we conducted 4 interviews with sellers who had never used Casavo. Then, in July, we did 7 interviews and usability tests with 7 Casavo users who got a home value estimate in the last 2 weeks.
Our goal was to understand what pushed seller to sell their house, what emotions were connected to it, and what influenced their decision-making process.
Price and guidance
- We kept a price range but also included an average, as we found out that many users were already calculating it themselves.
I prefer this version it also includes the average data that I calculated — User #7
- We give more guidance about the selling process and timeline. We added a whole page to explain that the number and value of offers depend a lot on how long the house is on sale.
- Next iteration: Add a time indication based on similar houses.
Algorithm and emotions
- Selling a house is often a challenging decision influenced by external factors. Yet, sellers opt for an algorithm — a computer assessing their home.
I had to sell quickly for personal/emotional reasons. My sister would cry every time she entered that house. — User #4
- We decided to reshape the perception of the algorithm by prioritizing the user, their emotions, and their home. We bring back the human touch. It’s not about how perfect our algorithm is. It’s about them.
Some criteria can’t be found on an online simulator (e.g., very quiet street). — User #2
My house is unique, it deviates from the standard; perhaps it’s me who couldn’t make it understood — User #3
Comparison
- We made Casavo more trustworthy by adding what users usually look for outside, right inside the app. We included similar houses to give them a helpful comparison and to let them see why we suggested a specific range.
- In the old version, there was no “similar houses” section. In Iteration #1, we had them on the page “understand the values”, but they were quite hidden. As users really liked the idea, we created a specific page accessible directly on the homepage.
- Next iteration: make the listings link clickable
Casavo fees
- Several users wanted to know our fees before engaging with us. No one likes to waste time without knowing how much a service costs.
“We haven’t talked about your fee yet; I think it’s included in the selling price. I’d like to know this at the beginning for clear information. Otherwise, I wonder, ‘Why should I choose you?’ “ — User #6
- We spoke with the sales teams to get the pricing grid. As it varies from country to country, city to city, and depends on the home value estimate, we are working with data on how to display exactly the fee that the seller will pay. In the meantime, we added it to our FAQ page.
- Next iteration: add Casavo fees on the home estimate page
A shifted business model
- But we knew there was a key question we weren’t addressing in the previous version: ‘Why doesn’t Casavo buy my house anymore?’”
- Since 2016, our users have known us as an iBuyer, meaning a company that buys your house. It was our differentiating factor. In 2022, we started transitioning to a new model: the marketplace. Instead of buying houses directly, we mainly help sellers find potential buyers in the market.
- Now, we try to explain better how we found potential buyers and how the process works. In the Q&A on potential buyers, we address users’ doubts head-on. Internally, we started some discussions with the brand team on how and when to communicate such a change.
Conclusions
So, we’ve revamped Casavo’s home estimate page. After dealing with internal requests and listening to user complaints, we did our best to create a clearer, user-focused page rather than treating our algorithm as all-powerful.
We shared our research results with the whole team in short (500 words) and long versions (2300 words), so everyone could choose how deep they wanted to dive. We’ve also outlined the next steps.
User research helped a lot, but we also realized we can’t be 100% sure of our solution. That’s why our work is not done yet.
How is it performing?
- We managed to double the conversion rate of the CTAs to refine the home estimate, with an even stronger increase for sellers who had potential buyers
- Seller advisors are no longer receiving feedback about home estimates which upset sellers. We don’t judge their house anymore based on an algorithm, but we go further and let them see how selling with us could potentially look like. Still, we have to deep dive into the presentation of our Marketplace value proposition. We just planned some shadowing sessions!
So here it is. Hope you enjoyed reading this article!
If you have further questions, just send me a LinkedIn message ❤️
You might be interested in:
- Post-mortem: how a sorry email had the best open rate ever reached
- Content design and data: a love story
- 9 actionable tips to improve your UX Writer/Product Designer collaboration
Or check other Casavo stories by my colleagues ✨
