MARTIN MATĚJKA explains how Keboola uses Keboola and other third-party tools to enhance their data pipelines.
Oh boy, it’s been more than a year again since my last HKBFUKC article (yep, that’s a new standard abbreviation).
This is the fourth article in the series. You can always check out the first, second and third on our blog.
Again, loads of stuff has happened since the last time. I made the top 16 at the 4 Seasons MTG Legacy tournament in Bologna, I visited Lego House in Billund and I got married!
During all these events, I’ve been accompanied by my friends — my wife in Italy and Denmark plus a bunch of others at our wedding. Honestly, (almost) everything's better with friends. It's more fun and you can help each other if needed. That’s why I want to talk about how we are using third-party friends to do cool stuff with Keboola.
We’re using different tools to help run our company for either web analytics, CRM or email. We’re just like everybody else in that matter. If you’ve read my previous articles, you know we automate our data pipelines and processes mostly with Keboola. This setup often makes sense for us (considering we know the platform inside-out — we created it!). We can use Keboola for any interaction between the tools we use. But, sometimes we don’t because the platform would be overkill for the task at hand. In such scenarios, we’re turning to Make (formerly Integromat).
So far, the most common use case for this tool is to process data provided via some kind of form. In general, we need to acquire the data upon submission, process it and distribute it to a couple of different places. I’d like to show you some examples of how we’re combining the strengths of both Make and Keboola.
We have a bunch of forms on our website serving different purposes. I’m going to describe what’s going on behind the curtains of the Contact us form.
It’s pretty obvious what this form is for: to contact the Keboola team. The most simple way to manage data from this form would be to send the response to some generic email and let a dedicated person handle the rest.
But, we want to do more with it:
If we’d like to accomplish these goals strictly in Keboola, we’d write a configuration of Generic Extractor, which would regularly receive data from Webflow (the tool we use for our website). Then, we would process the data via Transformation so it matches the desired output for Slack. After that, we’d write Generic Writer configuration to send data to the right Slack channel.
I’m not saying these steps are super complicated to do, but the entire process is too much hassle and takes longer than needed. It’s full of unnecessary steps and then you’re stumped with what to do after the data has been sent to Slack.
Hence why we decided to use a third-party tool Make-s sense (hilarious joke). Let’s check how Make fits into this use case. Below is the Make workflow for processing form submissions from our website:
To describe the steps:
This whole process in Make takes 15 seconds. That’s the time between form submission and the Slack channel notification.
In contrast, these steps would definitely take minutes in Keboola, as the platform isn’t built for these smaller tasks. But, Keboola is super great for the lead processing part, where you need to combine different data sources, add more info to a lead, decide if you want to store it as a contact in CRM instead and eventually trigger any other subsequent data pipelines. Just two good friends working together.
You can also check the Flow in Keboola, which brings the whole picture together. The flow is somewhat big, considering it’s amalgamating leads from multiple sources at the same time:
For a different type of scenario, I’d like to show an example where Keboola is completely omitted in the initial process (whaaaat?!). This process is actually something I’m currently working on. However, with pre-production tests behind us, I’m quite confident sharing it with you.
Our sales team is utilizing one external partner for acquiring cold leads and setting up meetings where our team can take over and present our awesome product. But, the “data process” around the use case is just terrible:
I’ve been wondering how to make this process easier for them. Below is the Make scenario I’ve prepared.
Of course, the leads are then synced to our CRM and grabbed by Keboola later on for all the other shenanigans we do with such data, so Keboola isn’t completely left out :)
As you know by now, the ideas in my blog posts are by no means final solutions. We’re always building on them. In the future, I’m pretty sure we’ll use Make or other third-party tools to enhance our current data pipelines and come up with more cool solutions.
I urge you to always try to find the right tool for the task at hand (and if you think Keboola is the one, you can try it for free, today).
And, never forget that with good friends there’s always a good party!
