At Funnel, we talk a lot about data connectors. They are an essential building block of a marketing data hub’s functionality, and without them, comprehensive data analysis is next to impossible.
Then, the other day while enjoying a bit of fika (it’s a Swedish coffee break with different confectionaries and good conversation), someone posed the question: “How would you explain what a data connector is to a lay person?” Then, someone else exclaimed in response, “That’s a great idea for a blog article!”
Seriously, though, it can be a bit tough to define. A data connector isn’t an API, it’s not a standalone piece of tech. So, what is a data connector, exactly?
We’ll break it all down for you, including the different kinds of data connectors and why they are so important to your data analysis.
What is a data connector?
A data connector is a bit of technology that allows for data to flow from one point to another. Say you want to pull data (like open rate) from an email marketing system to your customer relation management (CRM) platform. You’ll need to establish a data connection to do so. Similarly, to extract data from Facebook and share it to a dashboard, you’ll need to first establish a data connector.
To be clear, though, a data connector is not an API. The API is simply a metaphorical doorway through which a data connector can be established. The connector itself is an extra piece of software that establishes which data will be extracted from a given account, the place of the data’s origin, and the immediate destination – roughly speaking.
A data connector can be used to extract every ounce of data from a platform, a general set of data defaults, or it can be highly selective in the data it extracts. It all depends on your specific needs. In a marketing data hub like Funnel, they can come in two forms: core connectors and custom connectors.
Why are data connectors so important?
If you want to reach your audience, you need to meet them where they are. This often means that digital marketers need to run ads across different Meta properties, across search, a plethora of social media platforms, and more. In order to monitor the performance of your ads, you’ll need to review the data in an analytics dashboard or spreadsheet.
While nearly every modern advertising platform will provide you with some kind of a report, most marketers prefer to view overall performance in one place. This means you need to, first, extract the performance data from each platform.
To do so, you’ll need some sort of… connector! This is where the marketing data hub comes in. Of its many features and functions, this hub will make a connection to all of the platforms you ask it to, then extract all of the data you need for your analysis.
Without a data connector, you wouldn’t be able to extract your data to a single source of truth (you know, one dashboard to rule them all). That means you would need to go into each platform individually, review performance, take manual notes, and then try to make sense of any discrepancies or erroneous attribution in hindsight. And it can be hard to find the sorts of insights that will propel your business forward.
Custom versus core connectors
Let’s imagine you're starting out along your data maturity journey. You’re aiming to be more data-driven in your marketing decisions, but you still aren’t highly sophisticated in your analysis. Instead, you may want to scrape readily available data from the most common ad platforms (see: Google Ads, Facebook Ads, etc.).
In this case, you may want a core connector to extract your data.
Say you aren’t just any user, though. Instead, you may be after a very specific set of data, or you may want to extract data from a very niche platform. In this case, you need a custom connector.
As you can probably guess, these connections are a bit more bespoke. In a marketing data hub like Funnel, custom connectors are often built to spec in order to get you exactly what you need.
Can I build my own data connections?
Speaking of building connections, we often hear a common refrain: “Can’t I just go off and build my own data connectors to extract my performance data?”
In short, the answer is yes. If you have the programming expertise, you can certainly build your own data connectors. However, ask yourself if you really want to build them yourself.
Building your own connectors could, in theory, save you money up front (if you have the programming capacity freely available in-house). In the long run, though, self-made connections often lead to a lot of headaches.
It’s not that people are bad at building connections, either. The real culprit are those pesky APIs.
Maintaining data connectors
See, platforms are constantly iterating on and improving their capabilities. These improvements lead to changes in the code, which can often require changes to the APIs allowing external developers to connect to the platforms.
In the context of a data connector, a change to the API will often mean a broken connection. That means no data, no analysis, and no insights.
And these platforms can change their APIs. A lot. In fact, at Funnel, we have multiple developers dedicated solely to Facebook’s API changes, making sure the connector is up to date.
For those opting to build their own data connectors, it means you will need to employ your own experts to constantly monitor and maintain every API-enable connection. It’s the classic build-versus-buy argument, and one we have a clear preference on.
What kind of data connectors exist?
As the digital marketing industry continues its exponential growth, more and more advertising platforms emerge. As a result, there are already a nearly endless amount of data connectors.
Just take a quick look at our data connectors page, and you’ll see all of the famous platform names you know and love — plus many more that you may have never heard of.
Think of this, with Funnel, you can connect hundreds of platforms on day one. And that’s just the platforms. You can have multiple connections for each platform.
Data connectors are the lifeblood of great analysis
Without data connectors, great performance analysis just isn’t possible. Can you imagine trying to manually review the performance dashboards for 500 different platforms? No, thank you.
Those connectors need to be properly built and maintained, too. Otherwise you’re likely to spend countless hours (and money) fixing broken dashboards and suffering from the Sunday Scaries.
We much prefer a world where you can click a button and automate the whole process, but to each their own we guess.
