Written by: Gabriel Solberg Read time 7 minutes

ETL (Extract Transform and Load) solutions are designed to centralize data across products and software into a single data warehouse to create one source of truth. Reverse ETL is the inverse process. With reverse ETL, the goal is to send customizable data models and identifiers from your source of truth back to systems that can action these key data points.

Why should performance marketers care about reverse ETL?

Running advertising campaigns for a B2B SaaS product or Enterprise sales-assist funnel looks vastly different from e-commerce. For e-commerce the primary funnel is pretty standard: Product Page View > Add To Cart > Initiate Checkout > Purchase. Once you configure your e-commerce store with the ad platform, you will have most of the conversion signals you need for optimizing campaigns. 

For B2B, SaaS and lead gen marketers,  each unique customer behavior flow and each sales-assist funnel have different key measurement points that signal a high propensity to purchase. This complexity combined with the uniqueness of each product makes it harder to capture, track and leverage key conversion signals to optimize paid campaigns. Having an API integrated reverse ELT solution creates a real-time conversion feedback loop that is future proof for a privacy-first web.

Before we dive deeper into the reverse ETL process for performance marketers let’s quickly review how a privacy-first web is impacting the digital advertising ecosystem.

The Wild West of the Web is Getting Regulated

Privacy and data protection practices are becoming a dominant topic globally across the web. This impacts how we collect, store and use first-party data and impacts ad platform targeting and measurement capabilities. Last year’s iOS14 update and Safari’s cookie restrictions make it harder for platforms to target people at the specific time of intent. In addition, limiting the ability for cross-channel and device attribution. In addition, Google is phasing out third-party cookies, making it harder to track activity across the web. Falling back on the simple last-click UTM tracking model is important to identify the first measurable entry point. However, it misses the full journey across devices and channels. To solve this, ad platforms are pushing offline conversions or their conversion API and in our view, browser-based pixel tracking will become a legacy solution.

How can first-party data solve the new challenges facing performance marketers in this new world?

As this trend continues, the power lies in collecting first-party data as early in the prospecting journey as possible. Once you have a unique identifier, you can model your owned first-party data with a common join key to measure the CAC, payback period, and LTV of a customer. 

You can also marry your first-party data with the walled gardens, (for example Google and Facebook), for match backs in a privacy-safe way by sharing hashed data only. 

This is where reverse ETL comes into play. 

How Ad platforms are Adapting to the New Privacy-First Web

Ironically, as ad platforms are adapting to a privacy-first web, you as an advertiser get less control and rely more on the power of automation, machine learning, and the algorithm.

We have seen this across Facebook by forcing you into audience expansions and the continued removal of audience segments. Google’s newer campaign type called Performance Max creates a black box of data where you feed it copy, creative, landing pages, and intent signals and Google finds people that will take the key conversion action.

This leaves performance marketers with two areas they can control: Ad Creative and Conversion Signals. As we talk about the reverse ETL process we are going to be focusing on conversion signals.

The Emerging Trend of Reverse ETL Solutions

ETL solutions like Stitch are built so that you can centralize and join your data sets across your tech stack for advanced analysis. As a result of the “big data” trend, the industry has matured and democratized the cost and access to data warehouse solutions. Along with it, there has been an explosion of tools to support ETL data flows. A common example of this is Google. Big Query is a low-cost data warehouse that can be used for small and large data sets. Google’s new version of Google Analytics called GA4 supports Big Query streaming exports so that you can stream your Google Analytics 4 data to BigQuery. This was only available for Google Analytics Enterprise accounts in the past.

The Value of Centralized and Unified Data Sets

We have to keep in mind though that the value add of a centralized and unified data set falls into two categories:

• To perform data analysis and identify high-impact trends and patterns.
• To discover high-value data signals that can be actioned.

The discovery of actionable signals through analysis is an incredibly important first step. However, for us to action these, we go back to the ad platforms and make manual, directional, and inferred optimizations. What we should have is a reverse ETL solution. This solution would act as the real-time automation layer between our source of truth and the ad platform. This creates a conversion data feedback loop that trains the ad platform’s machine learning capabilities with high-intent signals from the source of truth in real-time.

The reverse ETL trend of sending events and signals that sit within existing data models back to operational systems is not just something that can benefit performance marketers. Rather I see this as a natural progression of “big data”.  We now have unified data sets that can be queried to answer complex questions. How do we action these in real-time within our existing technology stack? 

The Intersection of Reverse ETL and Performance Marketing

BI and product teams do amazing work. If you are joining a business with a decent data set it won’t take long to figure out key conversion signals. In addition, these conversion signals will continue to change over time.

However, as a performance marketer:

1. We don’t always have the SQL experience and the engineering knowledge/resources to build complex data streams.
2. The BI Analytics and Engineering teams often do not have the bandwidth to support performance marketing-specific requests.
3. We are not close enough to the BI, Mops, and Engineering teams to facilitate a real-time Reverse ETL data transfer.

What B2B Performance Marketers need is a self-serve, low code, reverse ETL solution to integrate the source of truth, be that the sales-assist CRM or Product Analytics tool. This will enable us to select high-intent signals through a user-friendly interface and send them back to their respective ad platforms as conversion signals in real-time.

Identify the Right Conversion Signals

If you are a performance marketer in today’s age you know that sending high-intent conversion signals is a key component for running successful paid advertising campaigns.

For B2B and SaaS lead generation, identifying high-intent or valuable lead enrichment data signals within a CRM or Product Analytics platform is relatively easy. Below are some example signals:

Sales-Assist Funnel – CRM:

• Leads with a predefined lead score
• MQLs for priority business segments
• MQLs that match a target account list

SaaS Funnel – Product Analytics:

• Trial sign up + Completed Onboarding
• New user + 3+ actions
• New Users that completed action x

The key to selecting the most effective event is to identify high-intent conversion signals that have recency. Ideally, ones that occur in a 7-day window. So what we are looking for are signals that have:

1. High-Intent/Propensity to Convert
2. (Ideally) Occur within a 7-day window
3. A healthy conversion volume

Note: I won’t get into the importance of working closely with the product and sales teams to properly identify and map these signals. I will save that for another article and link it here once posted. 

Once the high-intent signals are identified what can you do with them?

Leverage a Reverse ETL Solution

Ok, you have identified a high-intent action that occurs within a 7-day period. You did this by analyzing data sets that live in your data warehouse and identifying patterns of user journeys that drive revenue. Thanks to the ETL systems in place you were able to look at ad platforms, CRM, and product data in one place. Next, let’s look at what today’s potential conversion data flows look like without a reverse ETL solution in place.

Server-Side Events

You can set up server-side events. However, for SaaS products that is often not trivial. If you have a tool like Segment in place and Segment is synced with your CRM you could potentially create a reverse ETL process that could push server-side conversion events back to an ad platform. However, to achieve this you would need to unlock dev and BI engineering favors. These are often hard to get and create bottlenecks, especially in larger organizations.

Pixel Based Events

If you are working as a consultant or within an agency, this becomes even harder. Mainly because as an outside partner you do not have direct access to the BI Analytics, Engineering, or Sales Ops team.

You can use pixel events and try to fire pixel events on key action steps. However, the challenge here is the following:

1. Your action does not always live in a place where you can fire a pixel event. If your actions are in a CRM that is manually updated, a pixel event will not be possible.
2. If your conversion action takes longer than 24 hours, data loss becomes a concern.

Alternatively, you could leverage conversion APIs or offline conversions. Both of which require a custom configuration or a manual CSV upload process. 

So what is the ideal solution? Performance marketers need a way to directly connect advertising platforms to their CRM or Product Analytics solution. This platform can then leverage ad platform conversion APIs to send high-intent conversion signals back to ad platforms in real time. This reverse ETL process creates a real-time automated conversion feedback loop that does not require any additional BI or engineering work.

Does This Solution Exist?

Yes! ConversionStream, recently released out of Beta, offers a reverse ETL solution for performance marketers. As performance marketers, our job is to drive revenue and there are factors that play into making this possible. For example, product expansion, sales workflows, data analytics, and data quality. However, it is crucially important that we identify and use quality data signals within this complexity to make a meaningful impact and drive ROI-positive performance.

Our reverse ETL solution, ConversionStream, requires no manual uploads and no additional engineering work (you can create your first data stream in under 15 minutes) to send CRM data as conversion events and audiences to ad platforms in real-time. ConversionStream is specifically designed and developed for SaaS and B2B performance marketers so that they can drive impactful results today and in a future privacy-first web.

Learn more about ConversionStream by watching this video, creating your free account here, or contacting us for a demo today.

AUTHOR BIO:

With over a decade of digital marketing experience, Gabriel excels in search, display, and social advertising across various industries. He specializes in performance media and data-driven storytelling, building custom digital ecosystems that enhance user experiences across both paid and owned channels. Connect with Gabriel on LinkedIn or explore more of his articles here.