A figure that gets quoted often: companies actually use and activate only 30% of their data.
So where is the rest of it? Usually in the company's data lake or data warehouse, where marketers have very limited access because getting at it takes at least some SQL or analytics engineering.
Everyone talks about being data-driven, so it is fair to ask why 70% of a company's data goes underused. Data can lift marketing performance and grow revenue by winning new customers or keeping existing ones. You just have to be able to use it.
To use that stored data and get it out of its enclosure, you have to make it accessible — and that is where reverse ETL comes in.
Contents:
- What is reverse ETL?
- What is the difference between ETL and reverse ETL?
- How reverse ETL actually works
- Why your business needs reverse ETL
- The reverse ETL tools on the market
1. What is reverse ETL?
Reverse ETL is the process of sending data stored in a warehouse out to business tools and applications — CRMs, marketing automation, analytics or media activation platforms.
Reverse ETL keeps data fully in sync across the various tools and applications a company uses day to day. Keeping those flows alive is the key part.
2. What is the difference between ETL and reverse ETL?
Here is the acronym again: ETL stands for Extract, Transform and Load. ETL is the process of collecting data from various sources, then cleaning and transforming it so destinations can use it, and finally loading it into a data store — usually the data warehouse.
In short: ETL pushes structured data into a data warehouse, and reverse ETL takes that same data back out of the warehouse to put it to work in your marketing tools.
3. How reverse ETL actually works
So how does the reverse ETL logic work in practice?
There are several steps:
- Extract the data : query the tables in your data warehouse (BigQuery, Snowflake and so on) to pull out the data you want.
- Transform the data : you will sometimes need to break out of the specific, heavily structured format of warehouse tables. Plan on transforming the data so it matches what the target destinations expect.
- Load the data : load the data, transformed or not, into your destinations. That can happen by API, manual integration, in batches or in real time.
Once those first steps are done, it is time to activate. The data now sits in your marketing, CRM and BI tools, and your teams can use it day to day. One word of caution though: monitoring. This is a data flow with transformations in it, so you need to check reliability regularly. Some reverse ETL tools will flag failed syncs for you.
Four things matter a great deal in reverse ETL:
- The source: the data source, usually a warehouse, can affect the format of the data.
- The model: the specific query (usually SQL) you use to sync data from your warehouse into the business tools.
- The destinations: the tools and applications that receive the data coming out of the warehouse.
- The mapping: matching the data in your warehouse to specific fields in your target destinations.
4. Why your business needs reverse ETL
Reverse ETL gives companies access to enriched data and lets them use it to improve decisions, personalize customer-centric campaigns and so on.
Putting your data to work :
The data currently sitting in your warehouse holds only vague potential value for your business. Your analysts use it for analysis, but the moment you put it into applications and business tools, you have the chance to make it central to your marketing campaigns, your product development and more. Reverse ETL is best seen as an opportunity.
Avoiding data silos:
By its nature, reverse ETL stops data being locked away. It gives internal teams — marketing especially — access to holistic datasets. Teams are no longer limited to the data they can reach inside their own tools: product teams working only from Mixpanel, say, or media teams working only from the ad platforms (Google Ads, Meta, Snapchat, TikTok and so on).
Reverse ETL makes data far more accessible across the organization — data stops being the preserve of the engineers and analysts who know how to extract it.
Asking the right analytics questions:
Answering some questions requires data from several channels and business units.
> What do my highest-LTV users have in common?
> How does my welcome program affect customer retention?
> and so on.
As your business grows, you want to keep asking and answering data-driven questions without building ever more complex flows between your tools. Reverse ETL gives you access to your centralized database.
5. The reverse ETL tools on the market
Whether reverse ETL is feasible in your company is another important factor. Think about the impact the reverse ETL process could have on your current architecture.
5.1 Segment — reverse ETL:
As a packaged CDP solution that has made a serious name for itself in the analytics ecosystem, Segment offers reverse ETL too.
One of the main advantages is that Segment provides every type of data pipeline in a single platform (event streaming, ETL, reverse ETL). With a packaged CDP you get everything in one place, which can be very convenient — but it has a price.
We like the product at Smart Bees. If you need help implementing Segment and getting value out of it, come and talk to us. We have used reverse ETL for plenty of use cases: sending BigQuery data (a specific segmentation) to a Braze CRM, sending BigQuery data into Segment to enrich user profiles (through Segment Profiles), syncing Snowflake data through to Salesforce so sales teams can prioritize their calls, and of course the classic one — sending offline conversions to ad platforms to lift ROAS.
5.2 DinMo — reverse ETL:
We have already built an architecture involving DinMo, where reverse ETL carried first-party data — attributes, characteristics, custom predictive metrics — to enrich customer knowledge in a CRM (Braze in that particular case).
The verdict on this one: perfect for a quick fix, less so for the long run. 😉
5.3 Hightouch — reverse ETL:
Hightouch is Segment's offspring — the packaged CDP we discussed above. Segment's reverse ETL block did not exist back then (it is fairly recent), and Hightouch was founded around that idea by a Segment alumnus.
That makes it the most popular reverse ETL platform. It supports over 200 destinations and integrates with modern data tools such as dbt, Fivetran and Looker. It offers version control, a live debugger, alerting support, and even a no-code audience builder so non-technical users can build audiences themselves through a visual interface. We talk to this tool regularly — it is a genuine alternative to the options above.
In closing
Reverse ETL platforms come with plenty of nuance. If you want best-in-class tools and a fully managed solution running within minutes, we would point you at the ones above.
Depending on your maturity, one specific option will suit you better than the others. If you have questions on the subject, get in touch with us at Smart Bees.
