Market Landscape
The concept of MarTech stack has gone through an incredible evolution in the last few years. While customer data platforms (CDP) are now part of most company’s vocabulary at least if not its technology roadmap. Today, most companies are chasing and looking to create their single source of truth: one single place that hosts the data that lives across different systems in order to overcome the data silos challenge and ensure companies can be data driven.
Nowadays everyone tends to agree that the best way to proceed with building their single source of truth is to move data towards data warehouses such as Google BigQuery, Amazon Redshift, Snowflake. Data warehouse is often and simply defined as a “data management system designed to enable and support business intelligence activities”. It is worth mentioning that over the last years, we have seen many companies spending tremendous amounts of effort building and investing in such projects.
Though another challenge subsists: being able to get data out of data warehouses. Indeed, while with data warehouses, companies have their most important data unified, cleaned and accessible they often lack the ability to activate it. Until recently, unless for data visualization tools such as Looker, more companies needed to write their own API connectors from the data warehouse to the SaaS products to pipe their data to technologies such as: ESP’s, CRM, Media platforms, Social media, etc.
This is where Reverse ETL intervenes. They simply describe the process of moving data from the data warehouse into third party systems to make their data operational. Simply said they are technology that enables you to copy data from your data warehouse to the tools you use for growth, marketing, sales and support. Reverse ETL offers out of the box connections with numerous systems so teams no longer need to create and maintain those connectors themselves.
ETL, ELT, Reverse ETL - Dafuq?
In order to understand what Reverse ETL stands for, we first need to talk about ETL. ETL stands for Extract, Transform and Load and describes the process of moving data from databases (cloud and/or on premises) to data warehouses.
Progressively the industry moved from ETL to ELT which stands for Extract, Load and Transform and consist of the exact same thing as ETL but with the important difference that the transformation happens after the data have been loaded into the data warehouse.
Now that we have clarified the concept of ETL and ELT we can go back to Reverse ETL. Simply said, Reverse ETL is the exact opposite of ETL. It is the process of moving data from the data warehouse to third party cloud apps.
Reverse ETL are relatively new on the market but are coming with clear added benefits. The first and maybe most important is the fact that it replaces the need to maintain manual connectors. From an implementation point of view this also means connecting your data warehouse with specific destinations only takes a few hours of work while custom connectors would take several weeks.
The typical Reverse ETL use cases
Well you might be asking yourself why moving data out of your data warehouse is important while you have just spent huge amounts of time and resources building this single source of truth? Unfortunately, while you tried to bridge your silos you ended up creating a new one.
Don’t get me wrong, data warehouses are key, they are a huge step forward as they allow you to build all this business intelligence: customer lifetime value, churn prediction, lead scoring, etc. Reverse ETL then allows you to do more than just adding a visualization layer on top of this intelligence. You can now pipe this intelligence in real time through all your third party activation tools in order to ensure your teams have a consistent view of the customer across all systems.
There are many use cases for Reverse ETL: (1) pushing data into your CRM means you can have an up-to-date list of high lifetime value customers. (2) pushing it into your customer support platform can ensure the interaction with the customers can be personalized and/or making sure the requests can be automatically prioritized as messages are coming in. (3) pushing data into your media platform (Google Ads, Facebook Ads) allows you to create audiences based on the enriched client’s attributes and intelligence you have created in your data warehouses. As a marketer, imagine being able to create on the flight an audience with “the top 10% of your customer in terms of lifetime value” or “the top 20% existing customers with the highest churn risk’. The list goes on.
The players on the market
Reverse ETL is still a very young category on the MarTech scene. At Human37 we consider Hightouch, Census and Weld as being serious players to consider and follow. From a features point of view, those players are relatively similar. Nevertheless, they differentiate in their pricing model: while Hightouch and Census both offer a free version of their solution Weld only offers a free trial period. Aside, Hightouch and Weld bases its pricing model on the number of connectors you activate while Census offers unlimited connectors and bases its pricing model on the number of fields you send.
Conclusion
While we can all agree Customer Data Platforms (CDPs) have been the hottest topic on the MarTech scene over the last few years, we are ready to put a ticket on the fact that Reverse ETL is next in line. Nevertheless, while they clearly look like promising technologies we also believe they still have a long way to go (especially in the European markets).
Their first challenge is for sure the need to educate the market. It took CDPs years before becoming a well known concept. And I do believe that there is still no clear and unique market definition that would ease some discussions. Reverse ETLs’ advantage here is probably that the scope it covers is smaller making it easier to understand. Also the operators of the tool is not the same.
Their second challenge is linked to their positioning. Reverse ETL bet on a data warehouse first model. Looking around, most companies don’t have yet a real single source of truth built on data warehouse technology. This means before even considering a Reverse ETL, companies will have to invest time and resources on building this single source of truth. Reverse ETL’s advantage here is probably that the fact they come up with such nice promises might speed up the data warehouse adoption and as a matter of consequence Reverse ETL’s adoption.
You want to know more about Reverse ETL? Reach out!
Comments