How Zillow Plans to Scale Data-Driven Product Development Along with its Expansion of Products and Services

Zillow is the most visited real estate website in the United States, with more than 200 million people visiting their websites and mobile apps every month. Over the last few years, Zillow has been expanding its services to help customers across their real estate journey. This expansion in services requires a solution that would enable Zillow to efficiently scale its data-driven rigor across its solutions and services while maintaining the partnership between product and marketing. Luckily, they found it all in one place.

Zillow’s 100+ data science and analytics team supports the buyer, seller, renter and real estate agent experiences. But as their business grows and scope expands, they can’t simply add headcount to service these growing needs.

When Senior Product Manager Vickie Chang joined Zillow in October 2021, she started her journey by interviewing data scientists, product managers, and marketers. “About 90% of the most frequently asked analytics questions were along the lines of, ‘How many customers performed X action on Y day, filtered by Z?’” says Vickie. She knew standardizing data science operations into a single tool could save time and allow the business to scale.

They had a page-based digital analytics solution in place, but it wasn’t cutting it. Nearly 200 people at Zillow had access to the tool, but most didn’t use it. The app’s interface was confusing to navigate and wasn’t user-friendly. Instead of self-serving, they came to the data science team with questions, adding to their workload and creating bottlenecks.

Vickie and the team knew they needed a new analytics solution to measure across all products and services, provide better data visualizations, and make inter-team collaboration easier. They started with a pool of eight to 12 options, but that quickly whittled down.

In addition to their data science team, their large AI/ML team builds countless models and personalizations for the Zillow website. To empower that team, they needed to own their clickstream data, not simply build a vendor’s app into their website.

Zillow ultimately decided to work with Amplitude for two reasons. Amplitude provided a versatile set of analytics capabilities, which would allow Zillow to arrive at directional insights and explore from there. Second, Amplitude’s partnership approach during the proof of concept phase.

The Amplitude team took the time to understand Zillow’s business, and put thought and effort into the demo to help them see the platform’s full capabilities. It was clear that Amplitude was the right partner for Zillow.

While only in the initial stages of implementation, the Zillow team is excited by what’s ahead. Vickie expects the product team will use Amplitude for discovery and opportunity sizing, and product managers will prioritize feature testing with self-serve data —and this new workflow will increase efficiency. The increased decision-making velocity will also apply to feature monitoring and identifying and prioritizing bugs.

Zillow’s marketing team will also benefit from Amplitude. That team has traditionally focused on driving website traffic and engagement but didn’t have insight on what happens after consumers leave Zillow’s website and app. “Having Amplitude will allow marketing to measure if their personalized campaigns are helping customers find the product or service they were looking for,” says Vickie.

Amplitude’s easy-to-use interface and collaboration features will help to eliminate silos at Zillow by enabling consistent customer and metric definitions between data science, product, and marketing. With Amplitude, product and marketing team members at Zillow will have the ability to ask and answer critical questions without always writing SQL or needing a data scientist to provide an answer.

Employee interest at Zillow increased from 10 to 50 pilot users over the course of the evaluation period. The reporting and analysis platform team at Zillow have planned roadshows and trainings through next year in order to grow adoption by the end of the year.