Reverse ETL Bringing the Modern Data Stack Full Circle

Reverse ETL is an emerging category in the modern data stack space. Learn how you can eliminate data silos, operationalize your data, and automate your data integration processes.

By Josh Hall on

Ops

The modern data stack has democratized access to data for everyone within an organization. At the core of the modern data stack lies the data warehouse. The benefits of a data warehouse are vast, with the goal of getting all data from across your business into a single place to remove data silos. Ironically, this means that the data warehouse itself has become a data silo: anyone who is not a technical user is limited to the dashboards and reporting that have been created for them and often don’t get used. So how do we solve this issue and unlock the data warehouse from being a data silo?

An emerging category in the modern data stack space is Reverse ETL. Reverse ETL solves the problems outlined above by copying data from your source of truth, the data warehouse, to your operational systems. This in effect operationalizes your data, enabling all users to have the data and insight they need in the tools they use most often, without ever having to use or check a dashboard. Hightouch, the leader in Reverse ETL defines it this way:

Reverse ETL is the process of copying data from a central data warehouse to operational systems of record, including but not limited to SaaS tools used for growth, marketing, sales, and support.

But before we dig into how Reverse ETL can really benefit your organization, let’s examine how Reverse ETL is different from ETL.

What Is ETL and Reverse ETL?

ETL, or extract-transform-load, is the process of copying data from a data source into a destination, most commonly a data warehouse. Fivetran is a great example of an MDS tool that does this really well.

Reverse ETL on the other hand is concerned with copying data from the data warehouse into systems used within an organization. So instead of getting data from systems (ETL) and siloing it in a data warehouse, Reverse ETL redistributes data back into the necessary systems within the organization. This is the core value of Hightouch.

Hightouch is the premier Reverse ETL tool. Instead of building complex data pipelines and managing changing APIs, Hightouch puts the power of Reverse ETL into the hands of anyone who knows basic SQL. Hightouch has built a robust platform that only loads data that has changed since the last load, and also provides logging and observability around all the records processed. Additionally, Hightouch integrates with many modern data stack tools, most notably dbt, allowing you to pull existing dbt models into Hightouch.

If you’re still not convinced that Reverse ETL and Hightouch are the right choice for your data stack, let’s examine the benefits of using Reverse ETL.

Eliminating data warehouse silos

In the past, the focus has been on moving data from disparate systems into a centralized destination. However, the planning around how to leverage this data has often been underdeveloped. After all, just having data in a data warehouse doesn’t provide any inherent value. By leveraging Reverse ETL, we can now move data that has been aggregated, modeled, and enriched, back into the systems used by teams on a daily basis.

Instead of having to build custom integrations to send data between your various systems, you can use an off-the-shelf solution that only requires basic SQL skills. You don’t need knowledge of APIs, you don’t need to know programming languages in order to create a custom integration, and you don’t need to maintain what is likely a brittle integration. Instead, data can be redistributed back into the ecosystem of your business systems allowing insight to be obtained at the point of user interaction, all while using a robust tool like Hightouch. No custom integration management, no dashboards, just insight. But how is moving data back into business systems valuable? After all, didn’t we just ETL data from those systems into our data warehouse?

Operationalizing your data

With the ability to move data from a data warehouse back into business systems, we now unlock the power of operational analytics. Operational analytics enables teams across your organization to have the insight they need within the business systems they use on a regular basis. This means that time isn’t wasted building dashboards that won’t be used, or worse yet, building a dashboard and attempting to get teams to adopt it. And best of all, it’s just as easy to enable operational analytics as it is to build a dashboard.

The teams across your organization work with specific tools day in and day out. By feeding data that has been transformed and enriched back into those systems, those same teams can now make highly informed decisions directly from the systems they are used to working in. Here’s an example:

Let’s say that an organization has a model in dbt that calculates a lead score based on certain criteria. While the organization could build a dashboard on top of this model, we can only hope that sales reps will look at the dashboard, not to mention those who may be confused by the dashboard and never adopt it. As a solution to this, we could create a custom field in Salesforce called “lead score”. From there, we could use Hightouch to Reverse ETL the lead score calculation from our dbt model, into our new custom field in Salesforce.

We’ve now operationalized our data by giving a lead score to all of the leads in our Salesforce account, all while giving sales reps that information in the platform they use on a daily basis and are comfortable with.

Automating data processes

Many times, even when a dashboard has been created, users still ask for an export or CSV of the data represented in the dashboard. This often results in the manual process of writing a SQL query to obtain the data requested, exporting the data as a CSV, and sending it to the relevant parties. Because we’re already using SQL, we can use Hightouch to automate this process.

For example, at the end of each month, the finance team asks for a CSV containing total sales by each day of the month. The finance team, who love their spreadsheets, prefer not to use a dashboard because they like to crunch the numbers in different ways that a dashboard doesn’t always allow them to accomplish.

An analyst at the organization is responsible for obtaining this data for the finance team at the end of the month and has a SQL script saved to quickly obtain the data needed. The analyst then exports the results of the script to CSV, and sends a Slack message to the finance team with the file attached.

While this process may only take a few minutes, the analyst has to remember to perform this task, or the finance team has to remember to request the data. Instead of this manual process, we can use Hightouch to send updated sales data directly to the spreadsheet of the finance team’s choosing, at whatever cadence they prefer. This process has now been completely automated without introducing any additional work. Not only that, but the analyst no longer has to remember to perform this task - the finance team has data in the spreadsheet that they want, and the data is updated on a regular basis, not just once a month. A win for everyone.

Case Studies

Here are some ways companies are activating their data using Reverse ETL solutions:

Frequent roadblocks in Reverse ETL adoption

At Untitled, we routinely see the same struggles in understanding the value of adopting Reverse ETL as a part of a broader data stack. Below, we’ve outlined the three most common roadblocks and why they don’t need to be a stumbling point.

Isn’t Reverse ETL just ETL?

No, it’s not. Reverse ETL copies data from your data warehouse into business systems, enabling operational analytics. ETL focuses on copying data from a source into a data warehouse. So Reverse ETL is literally doing the opposite of ETL, hence Reverse ETL.

Why would I move data back out of my data warehouse?

Yes, you just used an ETL process to move data into your data warehouse, but by using Reverse ETL, we can operationalize data and enable teams to effectively take advantage of the data in the systems that they routinely work in. We don’t have to push the adoption of a dashboard, or train people to be comfortable using dashboards. We’re able to feed those insights directly into the tools used every day.

Why do I need ETL and Reverse ETL?

ETL and Reverse ETL are not the same. You need both to effectively take advantage of your data. ETL is copying data into your data warehouse while Reverse ETL is copying data into business systems in order to operationalize that data.

Without ETL, we wouldn’t have data to work with. Without Reverse ETL, we can’t operationalize data, which means reverting to pushing the adoption of dashboards and hoping that people can glean insight from them

How to get started with Reverse ETL

Want to get started with Reverse ETL? You can use Hightouch for free at https://app.hightouch.io/signup: you can send unlimited data to your first tool (ex: Salesforce or 70+ tools), alongside our forever free integrations (Slack and Asana). Or you can book a demo here.

Are you still setting up your warehouse or modern data stack? If so, Untitled implements and deploys best practice modern data stacks in days as opposed to weeks or months. Start making data driven decisions today by leveraging your data. To learn more, visit https://untitledfirm.com/.

Sign up for more articles like this