Case Study: Customer-Centric Data Integration at Bloomreach
Bloomreach powers personalized experiences for 25% of eCommerce websites in the US & UK, including CVS and Staples. Powered by advanced AI, Bloomreach excels in conversational commerce, recommendation systems, and search enhancement. Leveraging Nexla, the company streamlines its data integration processes, expedites data provision for AI applications, and broadens its platform offerings.
As Bloomreach’s customer base grew, so did the demands on the data team to orchestrate large-scale customer data integrations. At the time, onboarding a new customer was a 3-month process that involved extensive engineering effort, which delayed their time-to-value. Therefore, Bloomreach needed a solution to reduce customer onboarding time and increase its scope of connectivity to various platforms.
Nexla enables Bloomreach to efficiently build data pipelines and automate data exchange with all its data sources. Through Nexla, Bloomreach extracts, transforms, and loads (ETL) customer data, so that high-quality, ready-to-use data is automatically fed into Bloomreach’s system.
Customer data integration is the prerequisite for Bloomreach’s AI-powered services. To provide customized solutions for each customer, Bloomreach put in extensive engineering effort and resources to connect to data sources and manage hundreds of data flows per customer.
“We are constantly ingesting a massive amount of data in all different types of formats,” said Xun Wang, CTO at Bloomreach. “If it takes a long time to configure things, the customers realize the benefits later.” As Bloomreach expanded its customer base, building pipelines one-by-one and putting them into production were proven to be costly and unreliable. Customer satisfaction and deal success were at risk during the onboarding phase. In search of a scalable solution, Bloomreach’s data team turned to onboard customer data through Nexla.
We were able to take a lengthy 3-month engagement and onboarding period for data integrations and cut it down to 1.5 months with Nexla.
— Xun Wang, CTO @ Bloomreach
Nexla’s universal connector architecture allows Bloomreach to build pipelines to efficiently integrate customers’ data from any source into its AI models, reducing the overall integration time by 50%. Nexla’s error detection and quarantine feature ensures uninterrupted data streaming. At the same time, the data team maintains full control over the data pipelines and has immediate insight into the pipelines’ health. All this translates into “a huge actual improvement in terms of our ability to bring customers live,” as stated by Wang.
Smoother experience, happier customers
In order to ensure the quality of data that enters the retrieval augmentation system, Bloomreach strives to keep customers’ data up to date at all times. For example, if a product goes out of stock, the related product SKU (stock-keeping unit) and catalog information must be updated. The data engineering team then had to communicate with the customer back and forth to make the change. Onerous processes like this would negatively impact customers’ user experience and cost additional resources.
The (Nexla) system is highly scalable, so maintaining data no longer costs our customers thousands of dollars. This makes our customers a lot happier.
Nexla’s intuitive interface and clear data ownership flow make it convenient for Bloomreach’s data team to collaborate with customers and ultimately delegate data updating tasks to business experts on the customer side. A change can be made in a few clicks without involving Bloomreach’s data team. Customers can easily navigate their data and update catalog information as often as needed without extra resource input. This way, Bloomreach is able to “maintain a higher uptime and less failure,” said Wang, and customers may stay nimble.
Unlimited connectors out-of-the-box
One of Bloomreach’s core goals is to deliver seamless experiences to all customers across all e-commerce platforms. Nevertheless, handling complex data generated by diverse platforms with various customizations was a roadblock to expanding sales opportunities. Without a centralized data exchange platform and capabilities to scale and maintain the pipelines, Bloomreach initially had to rely on building one connector at a time, which resulted in extremely low ROI.
Now with Nexla, we get a whole host of connectors that can be plugged into any kind of commerce platform. On top of that, we also get the excellent usability, maintainability, and supportability of the connected pipelines.
The variety of connectors enables Bloomreach to quickly execute proof-of-concept by instantly connecting to prospective customers’ data sources. This accelerates deal progress, allowing customers to experience Bloomreach’s product features instantly.
As Bloomreach moves forward in becoming a unique complementary feature offering at platforms such as Shopify, Nexla equips Bloomreach with flexible and reliable connections to every platform. With Nexla, Bloomreach can focus its resources on its core competency in AI and analytics to further perfect its powerful product without wasting data engineering effort. After all, as Wang said, “we have to focus our energy to churn around high business value rather than trying to engineer everything — we don’t want to reinvent the wheel.”