tldr

In this project, I worked closely with a PM and Analysts to identify opportunities to improve site-wide engagement. From looking at relevant data, we identified opportunities to improve engagement with photos on the browse photo and view photo pages for both desktop and mobile web. After this, we worked in sprints to improve each of these opportunities.

My role

As the lead designer, I worked cross functionally with 1 Product Manager, 3 Developers, and 1 Product Marketing Manager in the same Palo Alto office. I collaborated with the PM to identify opportunities, set up a roadmap, and worked with Engineers to roll out the work in sprints.

Results

We managed to boost time on site across all the projects we identified. On mweb, after redesigning the browse photo page, the overall time on site went up 11%. After redesigning the view photo page, time on site went up 7%. For dweb, the browse photo page redesign led to a 6% increase in time on site.

opportunity Identification

For this project, we had a very broad goal of improving engagement metrics across Houzz. Given such a broad scope, we decided to hone in on the photo journey because it’s the most commonly trodden path on Houzz.

First, we looked at the data to understand how people navigate the photo pages across different platforms. We noticed that users spent a lot less time on mobile web and mobile web users browsed x% fewer photos than users on desktop or app.

One of the shortcomings of mobile web was that users could not tap or swipe to the next photo on the view photo page. However, these actions were the most frequently used actions on other devices in order to view more photos.

We also conducted competitive research and found that competitors had large sections of similar recommendations at the bottom of their main photo pages. So when you view one piece of content, for example on Facebook, there would be similar content underneath so that the user could continue scrolling and seeing similar content to stay engaged. And on Pinterest and Instagram, photo sizes are smaller enabling users to see more photos in one scroll. In the discovery phase, this is important, because it allows users to view more photos more quickly to find what they’re looking for.

hypothesis

From these findings, we derived the following hypotheses:

• Ability to swipe to the next photo on the view photo page will lead to more engagement on mobile web

• Smaller photos on the browse photo page will lead to more engagement on mobile web

• Similar photo recommendation at the bottom of the view photo page will lead to more engagement on mobile web and desktop

• Infinite scrolling + sticky nav will see more engagement on browse page on mobile web

Gathering Design Requirements

Before I started the design process, I gathered all the additional design requirements that helped me to understand all the moving pieces that I needed to design for. These included:

• View Product Mode. Previous photo page allowed users to see products in the photo. The new view photo page must allow this to function to exist as getting rid of it will hurt monetization

• Ads. Both browse and photo page had ads and we needed to keep these ads. I worked with the ads team to understand how the ads ecosystem works, and designed new ads to fit the new browse page and view photo page design

• SEO. We provided a different experience for both signed in and signed out pages to ensure strong SEO performance. I worked with the SEO team to understand what impacts SEO and designed around this

• International. New design had to be localized, so I worked with the international team to understand if there were any constraints

Testing

After coming up with all these requirements, I worked with an engineering lead and the PM to test in phases. For each test we wanted to see if we could increase engagement, while maintaining SEO traffic for the signed out version:

Results

We managed to boost time on site across all the projects we identified. On mweb, after redesigning the browse photo page, the overall time on site went up 11%. After redesigning the view photo page, time on site went up 7%. For dweb, the browse photo page redesign led to a 6% increase in time on site.