Yem Jam (Vol. 11)
We moved past two big milestones this week:
We were able to bring our customer's Substack data into our database without any human intervention.
We sent our first notification without any human intervention.
A round of applause (👏 🙌 🙏) for our development team. The first bullet is foundational to any data-driven product we want to build. The second bullet is a proof of concept for the initial product offering we want to launch (let’s call this our “MVP”).
The piping is set up — we now have a pathway to launch a "product" that can push us towards 50+ paid customers. Our focus now turns to scoping, testing, and building an MVP that will attract 50+ paid customers.
We're going to sprint during the next 4 to 6 weeks to launch an MVP. We'll see how it goes, but we're excited. If the Yem Jam becomes a bit spotty, it's because we're laser focused on building.
Let's dive into a bit more detail on our first notification without any human intervention.
Current Growth Efforts
Below is a selection of growth tests we’ve run over the past few months. Each row represents a growth test and each section represents a customer.
We've sent these growth tests from Substack or the customer's personal gmail account. When we run tests through Substack, we face a few challenges:
issues with deliverability (spam filters, Promotions tab, etc.)
difficult to target with precision
impossible to automate anything that works
Sending from a customer’s personal gmail sidesteps deliverability issues and targeting. But it’s still impossible to automate and it’s a pain for our customers.
These challenges pushed us towards developing our own communication stack. (We recently wrote about how this also enables us to solve harder problems a bit further down the road.)
Proof of Concept
This past week, we sent our first notification without our customers or our team having to press send. There was a lot of work that went into reaching this point:
Automate data extraction from Stripe into Snowflake
Create logic in Snowflake that identifies the distribution list
Set up email template in SendGrid
Integrate with SendGrid API and schedule when to send the email
We started with the payment failure notification. The email itself requires little customization and has had a nice impact on saving paid subscribers.
We saw solid performance: we saved 22 paid subs, driving $860 in incremental cash for our customer. All the performance metrics were better than our Substack sends.
We had 0 recipients mark us as Spam (🥳). We also received valuable deliverability data that we otherwise wouldn’t receive. For example, one of the emails bounced because it was a work email that was no longer valid. We’re hoping this helps us clean up our email lists and improve deliverability over time.
It’s an exciting milestone. We've laid the groundwork to automate a lot of the work we've been doing, which will let us bring on more customers. We're now able to target with more precision, allowing us to do things that were impossible before. We're positioned to build this into a machine that's always running and improving.
That’s it for this week — let us know what you think. As always, thank you for the support!