When customers shop for new homes in person, it’s easy for sellers to get to know them and create personalized recommendations based on an individual’s budget, interests, and unique home-buying needs. But when home buyers take their search online, building connections can require a little reinvention. For one new-build home buying platform, a customized Azure-based recommendation engine uses machine learning to deepen customer engagement.
Getting site traffic to this platform wasn’t an issue. The real problem was getting to know customers once they arrived; site visitors were largely anonymous, making it difficult to determine if the homes the platform displayed in its top search results were uniquely tailored to each visitor. The site’s existing search engine couldn’t intelligently sort results based on impersonal click-stream data gathered during a user’s search, for the most part returning random recommendations to shoppers within their geography of interest. Key sales measures like lead quality, conversions, and acquisition costs were unacceptable.
To provide the modern shopping experience its online visitors expected, the company sought to improve its search system by enhancing visitor data and using machine learning to supply personalized home recommendations.
The company partnered with BlueGranite and a third-party identity management provider to create a new recommendation engine powered by machine learning. Identification data can be used to ethically augment web traffic data from providers like Google Analytics with demographic (but not personally identifiable) attributes, such as gender and ranges for income, age, and family size.
Using the managed Spark platform Databricks on Microsoft Azure, BlueGranite was able to integrate large, diverse datasets like the enhanced clickstream, CRM, and product data to move beyond anonymous visitor activity to develop a more holistic view of customer behaviors. Attributes like demographics, lifestyle, purchase intent, search patterns, and other browsing behavior can be matched to other, similar visitors. The solution enabled not only business insights into previously unknown shopper patterns, but development of a sophisticated recommendation engine that produces personalized results for a customer based on what users “like them” have also found relevant.
While the company was already seeing impressive site traffic before the new recommendation model, this added touch of personalization helps the company:
Explore our Machine Learning and AI page to discover the many ways BlueGranite helps organizations incorporating transformative technology. If you’d like to learn more about how machine learning can revolutionize your shoppers’ experience, contact us today!