Follow Datanami:
November 2, 2021

Rockset Taps Reverse ETL for Last-Mile Delivery of Insight


Rockset has honed itself into a cloud-based database that generates real-time operational analytics from streaming data sources via SQL. Today it announced that it’s working with three of the early leaders in the nascent market for reverse ETL solutions to get those real-time observability metrics in front of the people who need them the most.

Reverse ETL is a relatively new data pattern that’s starting to gain traction in the market. As the name suggest, instead of extracting data from business applications and loading it into a data warehouse, as good old ETL works, reverse ETL reverses that workflow, and extracts insights from the data warehouse and loads them into business applications. The key difference is the business applications at either end of the pipe.

With today’s announcement, Rockset is now working with Census, Hightouch, and Omnata, three providers of reverse ETL solutions, to push Rockset-derived insights to be funneled direclty into SaaS applications, such as Salesforce, Hubspot, Zendesk, and Slack. This approach promises to get actionable insights in front of workers where it can have the greatest impact, primarily in sales, marketing, support, customer success, and operations, says Rockset Co-Founder and CEO Venkat Venkataramani.

“The raw data sources are all coming in real-time in Rockset. All the metrics are getting aggregated and reporting is all available in real-time,” Venkataramani tells Datanami. “But you really still have this last-mile issue of putting this in your business operations, whether its sales or marketing, trying to find which segment to target and metrics on what’s working, what’s not.”

For instance, traditional BI reporting may give a salesperson useful insight on a daily or weekly basis. But until now, it has been very difficult to instrument the data and applications to give the salesperson data reflecting insight that is just a few minutes (or a few seconds) old. That is what Rockset is doing with its reverse ETL partners, Venkataramani says.

“Now you can actually fan out and send your profit metrics for customers into Saleforce itself using these reverse ETL tools,” he says. “And so every sales rep can go to their account page or their opportunity page and actually track the usage, whether it is doing an eval for a big customer you’re looking to do an upsell on, or a big customer you’re trying to make sure doesn’t churn out. So you basically can have your cake and eat it too.”

This approach could also be useful in customer success. When traditional tools indicate that everything is working, but a customer insists that her service has been down for the past 30 minutes, having real-time metrics bubbled up from Rockset into the customer success agent’s Zendesk’s account could be the difference between ostracizing a valuable customer and getting to the bottom of a problem.

Any type of data that can be flowed into Rockset, where it’s joined, aggregated, and analyzed, can be turned around and broadcast to end-user tools via the reverse ETL vendors, Venkataramani says. That includes behavioral data, clickstreams, and customer usage data, he says.

“So now all this data is coming in from all of these different sources,” the former Facebook developer says. “You use SQL joins and other really advanced analytical functions in Rocket to bring in all of these disparate data sets together and then build the type of metrics you want to track for ever customer. Once that information is available, then you send that digested, processed information across a lot of different sources.”

This approach essentially is a new form of embedded BI. But because Rockset and its reverse ETL vendors have done the integration work ahead of time, it is easier to implement than a custom embedded BI solution would be, Venkataramani says.

“Because it’s full-featured SQL, we give customer full control. We also have role-based access control, so even when the syncs are all running, you can never access things that your roles don’t support,” he says. “It’s fully integrated. Customers don’t have to do anything ad hoc.”

Venkat Venkataramani is a co-founder and the CEO of Rockset

Reverse ETL is a fairly new concept and its usage is growing, says Eckerson Group Vice President of research Kevin Petrie.

“We’ve been hearing about it from our community of practitioners and consulting clients in the last year,” he says. “So it’s a relatively recent pattern and it’s I think a very natural outgrowth of the rise of data-driven applications by those applications that embed analytical functions within them.”

Organizations that have time-critical decisions to make when it comes to customer interactions are the most likely candidates for a reverse ETL solution, Petrie says.

“For some time now, organizations have been trying to figure out how to generate and expose data insights, analytical insight, within operational applications,” he says. “Reverse ETL has emerged as a pretty viable way to make the integration that much easier, especially for SaaS workloads.”

Census, Hightouch, and Omnata have jumped out to an early lead in the reverse ETL space, although it’s still early in the ball game, Petrie tells Datanami. “Those were the names we hear the most,” Petrie says. “Like a lot of spaces, it’s going to have a number of different players…It will take some time to develop, but I think that they’re starting in the right place with this group.”

Those three vendors are pleased to be working with Rockset, which has emerged as a leader in real-time analytics insight on operational data.

“We’ve seen first-hand how crucial it is for great companies to connect analytics directly into their business apps,” Census co-founder Boris Jabes says in a press release. “Whether it’s identifying new opportunities, preventing churn, or providing superior service, go-to-market teams need fresh data they can trust. With our native Rockset integration, Census makes it possible to bring true real-time data into any business workflow.”

That sentiment was echoed by Josh Curl, co-founder of Hightouch. “Our vision is to make data accessible and useful to everyone within the company,” Curl says in the press release. “One of the ways we achieve this is through our suite of native integrations with our customer’s data sources. We see Rockset as a leading real-time analytics database and we are excited about this integration bringing more actionable insights to our customers.”

The lack of additional middleware was highlighted by James Weakley, co-founder of Omnata “Omnata approaches reverse ETL uniquely by live-querying Rockset directly with no middleware, highlighting the real-time nature of Rockset,” Weakley says in the press release. “We’re excited to support Rockset as a live data source and harness its sub-second query power. It allows our customers to deliver large operational datasets to sales and service users at scale.”

Related Items:

Did Rockset Just Solve Real-Time Analytics?

Rockset Connects Kafka with SQL

Rockset, SQL Cloud Service, Emerges from Stealth

Tags: ,