Marketing ETL: the futureproof method to connect marketing tools and centralize data
Marketing data has been booming, but there’s been very little reflection on how to organize that data efficiently. How many marketing tools is your organization using? Probably at least a few, from a CRM to email marketing software, automation tool or analytics. To connect all these tools, you either need to stick with one vendor, or use connectors. But we have a better solution: marketing ETL. With marketing ETL (extract-transform-load) you can centralize all your tools. Get the full scope in our latest Tricks of the Trade.
We've embedded content from Vimeo here. As Vimeo may collect personal data and track your viewing behaviour, we'll only load the video after consent to their use of cookies and similar technologies.
What is marketing ETL?
ETL is nothing new: this term has been a mainstay in IT for years. But the concept has almost never been applied to marketing. And with the huge boom of marketing data, it makes sense to look at best practices from big data IT, right? Enter marketing ETL, which stands for extract-transform-load.
It’s not a platform or tool in itself, but a method to process data. You extract data from all your different sources, transform it into a uniform data format, and load it onto a central platform. This platform is often linked to your customer data platform (CDP).
Why use marketing ETL?
Well, if you’re using multiple marketing tools, it’s the best way to organize your data. There are only two alternatives, and they both have their drawbacks. Either you use tools from the same vendor, in which case they are automatically connected. But you become very vendor-dependant, making it hard to change tools or even change vendor entirely.
If you use tools from different vendors, you can use connectors. Usually, vendors release these connectors to be able to exchange data with other tools. But this is nothing more than a makeshift solution, that rarely works as well as it should. Only certain functionalities are covered with the connector, and when one of the tools updates their API, the connector requires an update too. If you have a failed data stream during this downtime, you wouldn’t even know, since there are no reports of it.
In short: connectors work fine to easily connect tools without handling too much data. But as soon as you get serious about marketing data, you really need a better solution.
What are the advantages of marketing ETL?
Marketing ETL really is a best practice to handle big datasets. There are many advantages to using this method for your marketing data:
Loosely coupled architecture
Your applications and tools are not directly linked, but they are each linked to the ETL individually. So if you replace one of them, the whole system doesn’t collapse. This gives you a lot more flexibility and freedom.Polymorphism
The ETL sends data in a standardized way to every application. If one of them is replaced, you don’t need to reconfigure the interface of existing tools.Better insights
Since you’re working with a single source of truth, you can appoint someone to monitor your data, create up-to-date reports and gain much deeper insights into your marketing data.
Retry failed data streams
With a connector, the data stream that fails to deliver is just cancelled altogether. You would probably never know the data collection failed. An ETL will keep trying to send the same data until it is successful.Historical data
The history of data streams is temporarily stored within a period you choose. This means that you can do replays of your data streams from a set amount of time. If something goes wrong, you can simply reset all your apps, and continue without losing any data. Moreover, if you implement a new AI-based tool, you don’t need to train the model from scratch. You can just play the historical data streams and feed it with real data immediately.
Want the full story on marketing ETL?
Quinten and Niels dig deeper on our youtube channel.
How to implement marketing ETL?
There are a few options. If you’re working with one of the big vendors, they usually have an ETL solution in their services. Google Cloud has Data Fusion and Data Flows, AWS from Amazon has AWS Glue, and Azure has Azure Data Factory. These services are very user-friendly and basically point-and-click.
If you’re not with any of these vendors or don’t want to use these solutions, there are several SaaS tools that you can use, the most common ones are Talend and Stitch. These are just as easy to use but require a bit more technical knowledge to set up.
