In this article, you'll learn:
- What is reverse ETL?
- What’s in a name: ETL vs reverse ETL
- Modern data stack 2.0: The era of operational analytics
- Why you need reverse ETL
- Reverse ETL vs point-to-point solutions
- Reverse ETL use cases
- What to look for in a reverse ETL tool
- How reverse ETL makes your data more effective
Do user-centric companies dream of data stacks? You betcha.
For business users, this dream looks like fresh, actionable data available in frontline tools like Salesforce, Braze, and Marketo whenever they need it.
For data teams, this dream looks like a world where their data--enriched with tools like Fivetran, Snowplow, and dbt--is used to its full potential to fuel operational excellence for every business team that consumes it.
However, the reality of data stacks past hasn’t lived up to these dreams, but they’ve gotten pretty close. Data teams leverage ETL tools like Fivetran to load customer data from mobile and web apps into a central data warehouse. This lets them do deep analysis, build advanced predictive models, and feed BI tools and dashboards to help business teams make decisions. But there’s still a gap--what we call “the last mile”--between the warehouse and frontline tools.
The bridge that lets you cross the gap over this last mile of your modern data stack--and from your current data reality to your data dreams--is reverse ETL. Reverse ETL is the difference between making decisions based on your data and finally being able to take action to realize your data dreams.
Let’s dive in.
What is reverse ETL?
Reverse ETL syncs data from a system of records like a warehouse to a system of actions like CRM, MAP, and other SaaS apps to operationalize data. That’s basically just a fancy way of saying reverse ETL lets you move data about your users from your warehouse and makes it available for frontline business teams to use in their favorite tools.
However, to really understand the power of reverse ETL (and why it’s not just another data pipeline), we first need to take a quick look at what traditional ETL pipelines made possible for business and data teams.
What’s in a name: ETL vs reverse ETL
The traditional extract, transform, load (ETL) data pipeline has remained largely unchanged since the 1970s: extract the data from the source, convert it to a usable format (or transformation), then load it into your data warehouse.
The advent of flexible data pipeline tools like Fivetran has also made it possible to load your data into the warehouse and then use your storage target to transform it (referred to as ELT). These ETL/ELT enabled companies to combine data from multiple sources into a single source of truth to inform business intelligence decisions.
This version of the modern data stack worked well when data sources were more limited (i.e. there was less data volume) and the data engineers who supported these stacks had ample bandwidth to process and answer questions about data. As you’ve probably experienced, that’s no longer the case and teams need more sophisticated tools to achieve the dream of operational analytics.
This reverse journey à la reverse ETL makes operational analytics possible. Reverse ETL tools flip the Fivetran role, extracting data from the warehouse, transforming it so it plays nice with the target destination’s API (whether Salesforce, HubSpot, Marketo, Zendesk, or others), and loading it into the desired target app.
Modern data stack 2.0: The era of operational analytics
The reverse ETL-inclusive modern data stack is the modern data stack 2.0. The growth in popularity of this new generation of data stack is emblematic of an important trend: Companies need to move data capabilities out of centralized silos and embed them within teams across business functions.
Reverse ETL equips these teams with detailed data inside the tools they're already using like Salesforce or Hubspot, empowering them to be more effective in their day-to-day work. The reverse ETL process effectively aligns your organization and applications around your source of truth. From there, business teams can build a shared, deep understanding of customers like never before.
The continuous flow of data--from raw data being pulled into apps to data being modeled to data being deployed into each app--creates a virtuous loop of operational analytics. And it’s only possible with reverse ETL.
This new genre of data tools closes the feedback loop that separated DataOps from DevOps and makes it possible for teams to deploy relatively real-time data and insights to core apps and services. - Boris Jabes, CEO at Census, The Operational Analytics Loop: From Raw Data to Models to Apps, and Back Again
The modern data stack 2.0 generally consists of the following tools performing four key functions to close the operational analytics loop:
- Data integration: Also referred to as collection, this is an ETL tool like Fivetran or Snowplow that integrates your data sources into your warehouse.
- Data storage: A data warehouse that can store structured and unstructured data in one place like Google BigQuery, Snowflake, or Amazon Redshift.
- Data modeling: A modeling tool like dbt comes pre-configured with a massive library of data models to make your data usable in different situations.
- Data operationalization: A reverse ETL tool like Census will pull data out of your warehouse, validate it, and load it into applications that need it like Salesforce or Zendesk.
As more teams within an organization require data to drive their daily operations, reverse ETL will become necessary to support democratizing data at scale.
Why you need reverse ETL
Without a reverse ETL tool, your data, and the insights from it, are locked within your BI tools and dashboards. This won’t fly in the era of product-led growth, which pushes companies across the B2B and B2C spectrum to improve customer experiences with personal, data-informed strategies.
As we touched on above, the key to this personal customer connection lies in operationalizing our data. Before reverse ETL, data pipelines were built for analytics alone (which meant data efforts were primarily focused on understanding past behavior). Now, companies can architect their data stacks to fuel future action, as well as understand past events (aka operational analytics).
At its core, operational analytics is about putting an organization’s data to work so everyone can make smart decisions about your business. - Boris Jabes, Census CEO
Reverse ETL lives at the heart of operational analytics at scale, constantly pumping real-time customer data into third-party applications to ensure when it comes time to make a decision, the right person has the right data to do it.
When teams across an organization work with synced data, traditionally difficult to automate tasks become much more straightforward. For example, reverse ETL makes it possible to intervene in the customer journey at just the right time by connecting your CRM and email platform to your data warehouse. This means more successful outreach campaigns and more delighted customers.
Reverse ETL use cases
Connecting teams throughout your organization to the warehouse using reverse ETL empowers them with data enriched with valuable context about what your customers are doing in real-time. As we discussed, an operational analytics approach puts data into the hands of people to inform their day-to-day operations. Let’s look at some examples:
- Customer success and operations teams can identify at-risk customers by surfacing product usage data in a CRM or other tools. They can automate notifications to other teams when usage trends downward so that they can immediately take action to prevent churn. Read more about how Loom did just this.
- Sales teams can set up rules around user behavior and other signals to identify potential sales among freemium users and automatically sync these into a CRM as new qualified leads. Called lead scoring, these rules can help sales teams prioritize leads based on the likelihood to convert and be more productive in their strategies. Read more about how Figma did just this.
- Marketing teams can build hyper-personalized marketing campaigns by merging product, support, and sales data to power customer segmentation. They can target users based on who should hear about new features or an upcoming offer or automatically send the latest blog post to those most likely to read it. Read more about how Canva did just this.
- Data teams can reduce the amount of time they spend building custom connectors for the teams above, instead of investing their skills and time into advanced modeling and more interesting data science work to empower business teams. Read more about how Clearbit did just this.
Without using some kind of reverse ETL tool, cross-discipline activities (like customer success managers reaching out to customers using your product less frequently) are difficult. You need to create the triggers and audiences within the platform sending the emails. Trying to run this campaign in Marketo is nearly impossible without access to your product usage data. Reverse ETL tools like Census make the relevant product usage data from Amplitude available in Marketo when the customer success team needs it most.
Reverse ETL vs point-to-point integrations
The no-code, plug-and-play nature of a point-to-point platform like Workato, Zapier, or Mulesoft often entices teams without dedicated technical or data resources to set up any necessary integrations. But relying too heavily on these quick fixes can quickly get messy as your data stack grows.
Fully integrating point-to-point solutions with your data stack requires exponentially more connections as your stack grows. The number of connections grows by the square of the number of applications, meaning eight apps could require as many as 64 distinct connections to keep your entire stack in sync.
Things can get messy quickly when you’re trying to manage too many integrations.
But with all of the customer data you already have sitting in your warehouse, there’s a better way. Instead of a messy, spaghetti pile of point-to-point integrations, you can use reverse ETL to architect your data infrastructure as a series of orderly spokes around a central hub (data warehouse). This creates a single source of truth informing each application and workflow within your stack to make you truly data-informed.
What to look for in a reverse ETL tool
As is the case with most software, when looking for a reverse ETL tool you’ll have to decide whether to buy an established product or attempt to build a bespoke solution with your resources on hand.
Building a custom reverse ETL pipeline may seem attractive, but it comes with the added complexity of not only engineering each individual connector but maintaining them against ever-changing destination APIs.
If you want to save your business teams from endless ticket filing (and save your engineers from having to address all those tickets), it’s time to consider a managed reverse ETL solution from an expert vendor. Here is a high-level overview of the seven key features to look for in a potential reverse ETL tool:
- Connector quality: A reverse ETL tool is only as useful as the applications it connects to. Look for the connections you need today and the specific features of each.
- Sync robustness: Syncing is arguably the most important feature and should be fast, be reliable, sync only data that’s changed, and be automatable.
- Observability: Your reverse ETL should offer alerting, integrations with monitoring tools, detailed logs, and the ability to rollback syncs, if necessary.
- Security and regulatory compliance: Vendors should have security credentials like SOC Type I or II, encrypt data in transit and at rest, and use best-in-class security for APIs.
- SQL fluency and ease of use: To be as user-friendly as possible, your reverse ETL tool should be SQL friendly, allow for easy modeling, and have an intuitive user interface.
- Community and vendor support: Make sure your reverse ETL vendor has a high commitment to SLAs, readily available support and in-app support, and good documentation.
- Transparent pricing: When buying a reverse ETL tool, make sure you know if the vendor charges by consumption, number of connectors, or fields per sync.
If you do your due diligence when selecting reverse ETL vendors, you’ll have the ultimate tool in your toolbox to ensure you get the most of your data today and as you scale in the future.
Reverse ETL makes your data (and the teams that use it) more efficient
When front-line teams can self-serve highly detailed customer data, translated, validated, and formatted for their favorite tools, data teams can spend less time crunching numbers and running reports and more time using their insights to inform business strategy.
The traditional role of data or analytics teams was, first and foremost, to report on how a product or campaign performs over time and serve the requests of the business teams they support.
This type of reporting and support was useful for monitoring the long-term health of your user base or high-level budget planning, but it couldn’t power automation or help customer success managers triage incoming support requests.
Today, data teams have embraced a whole new set of sophisticated analytics engineering skills. Unblock them and let them use these skills (you’ll be amazed at what they can do, we promise).
With reverse ETL in place, modern data teams turn data warehouses into the central nervous system of an organization, fueling email marketing, customer support tools, sales tools, or even financial models. This means more successful business teams who can self-serve deep, useful data, and more efficient DataOps overall.
Want to see how a reverse ETL system can change your day-to-day work and, your whole world of data, from day one? Schedule a demo with Census and we’ll show you what we can do. Or want to check it out today?👇Start Census for free