07: Brian Leonard: Be friends with engineering with open source Martech
There's a lot lost when we think of marketers and engineers as separate things and not the organization as a whole. The right thing to do is engage with the engineers that power your marketing tech stack. And meet them where they are. Open source martech can help with that.
So I’ve gotten down some rabbit holes prepping for this episode so pumped to dive in with you today.
Why don’t we start with the big differences between martech and open source martech.
I know that normal software does not include the source code while open source does and modifications and customizations are encouraged, but what does that mean in a martech context?
For me, open source is about control and trust.
You have the ability to control how your customer data is handled and where it is stored. We’ve seen this lead to people taking advantage of more data in their marketing efforts.
So then, you can see the code. You can control the data. This leads to trust. Only give the external tools what you want to share. Privacy and compliance get easier. We are talking with lots of medical companies, for example.
What’s the advantage of this business model, like why make Grouparoo Open source vs. the tried and tested SaaS model?
I don’t think the world needs another marktech SaaS solution. There are already thousands of those and yet these problems (integrating tools) persist. So we wanted to do something different. We think that working closer to the engineering level (and making it super easy) will disrupt how these tools get implemented.
Because there are so many tools to integrate with, open source will also help us build up those connections. We will actively engage with the more popular ones, but it’s exciting to see others interested in contributing connections to the long tail of tools.
Finally, there is cost. These SaaS solutions tend to be quite expensive and the incentive structure doesn’t line up between the company and the SaaS tool. For example, with Segment you send it a lot of events and more or less get charged per event. Then you pair back what you send. But then, later it turns out that you needed that. Doing all of this and owning that data is great for not only for privacy, flexibility, but also for cost.
Martech today has an awesome article on open source tools, I’ll add it in the show notes, but in there, they make the case that the Open source model is not ideal for Martech.
The most successful open source projects tend to be developer oriented—developers building tools for other developers, but in this case, the end user is often a marketer. I’m guessing you disagree?
When you want to integrate with Marketo, it’s the engineers that do that. I’ve talked with companies with millions of users that have been paying for Braze for a year and haven’t automated anything. I’ve met marketers that come in as CMO and demand tool X because they like it. A few quarters later, it’s more like “I just want to send a cart abandonment email! VPE, whatever you want to use is fine.”
I think there is a lot lost when we think of marketers and engineers as separate things and not the organization as a whole. The right thing to do is engage with the engineers that power your marketing tech stack. And meet them where they are. Open source helps with that.
If we can get the engineers excited about setting up the right architecture in an open way, then it will be easier to get more data later to existing and new tools.
Couple years ago Acquia acquired Mautic. They said in their press release that it was the start of a new generation of open source communities and projects to reinvent the martech stack.
Do you see an evolution of open source tools in the automation space?
Mautic was an ambitious project to do everything - to replace the tools you are using now. The evolution is about more target solutions.
A notable one is that there are even more marketing tools and they specialize, so Mautic would have to do everything. And maybe it didn’t do drip campaigns or push better than Iterable.
There’s a similar trend in the engineering world, especially on data teams. Data teams are growing and getting their own budgets. They are getting their own set of specialized tools. One example is Fivetran. That will store everything from Hubspot in your data warehouse. The missing piece, as we see it, is to make that actionable in the best tools for use cases in an efficient way.
I want to finish with integration of data in between platforms, and I know you guys solve this problem. Adobe Microsoft and SAP launched the Open Data Initiative that aims to standardize data across platforms.
The problem still isn’t fixed though. If you’re using 14 martech tools, chances are several aren’t Adobe products.
What’s the solution?
(open source community, and a standard by which data can flow between any set of systems, not just into one. You need the community to build all the connectors and adaptors between those tools, so you don’t have to custom build and code everything.”
I’m focused on building out a community around this tool. We won’t live or by whether the code works. That’s not a problem. The main thing is to make sure we get in front of the people that would benefit from it.
We need to get the word out so that when there is this need, Grouparoo is the obvious solution - both for now (easy to get going) and later through self-serve and lots of integrations.
To do that we’re doing podcasts, blogging, talking with people that are interested in using it, and building out the team.
Alright, Brian, what’s Grouparoo?
It solves an organizational problem we saw at TaskRabbit. We saw challenges between the product team and other teams that needed our help getting data for them to be successful, for example marketing and customer support. In general, these things don’t get prioritized and engineering becomes the bottleneck.
We’ve talked to others and they saw something similar. So we made Grouparoo to sync data from your product database or data warehouse to the tools you use like Salesforce or Zendesk. But we also made it open source and targeted at engineers to get it installed and data flowing. And then we added in ways for those non-engineers to help themselves to the data they need to be successful.
This help to solve the organizational problem through empowerment hat kind of autonomy for, say, a marketing operations team.
How does it work?
Grouparoo is open source and up on Github. There are examples of how to get it running on Docker or Heroku or any way that you run a Node app. You run that and point it at your primary data source.
What it will do is create a profile for each of our users, starting with their user_id or email or something like that. Then you can keep building out that profile from that or other sources until you have a centralized profile of who that person is with many properties.
With those properties, you can do segmentation. We can create dynamic groups of users based on their property values. For example, “High Value Bay Area” customers or “About to Churn” customers. These will always stay updated automatically with the real data.
Now we know these properties and group membership about each user. We can sync this data to destinations like Marketo, Salesforce, Hubspot, Zendesk, etc. You choose what you want to sync and it happens. And it keeps happening as the data changes.
Brian Leonard on LinkedIn.
Intro music by Wowa via Unminus