How I pick what to ship between Envoy and Studio
Contents
I'm the only PM at Voltade. We ship two 0-to-1 products: Envoy, the WhatsApp-first CRM for SMEs, and Studio, a no-code agent builder. They share infrastructure but compete for every hour of my week.
A few people have asked how I split time between them. Here is what I actually do, and the call I got wrong getting here.
#The Studio call I should have made earlier
Studio didn't exist as a product when we started. The original bet was Vobase, an open-source framework for AI coding agents. The thesis was simple: give SMEs the primitive, let their developers (or our integrators) build whatever agents they need.
What I underestimated: SMEs don't have developers, and the people running these businesses can't vibe code. They are not going to clone a framework. They want something that already does the thing.
Studio came out of that realisation. Same engine underneath, but packaged as templates an SME can adapt without writing code. The product is right. The mistake was that I treated Studio as a thin wrapper for too long and kept investing in Vobase as the customer-facing surface. I should have flipped Studio into the priority slot earlier than I did.
That is the prioritisation call I would replay if I could.
#What I actually do on Monday morning
The honest answer for how I pick what to ship is: whatever keeps Voltade moving forward this week.
In practice that's a short checklist I run through, mostly in my head:
- Is there a POC blocking a real lead?
- Is anything blocking a customer going live?
- Is there paperwork (a tender, a grant claim, a work order) blocking a deal from closing?
- Is Envoy stable enough that I can spend the week on Studio?
If any of the first three is yes, that wins. Studio's roadmap waits. This is uncomfortable, because Studio is the bigger long-term bet, but a deal that doesn't close this month is harder to close next month.
The list also has non-PM work on it. A website revamp, a tender response, a customer demo deck. As founding PM at a small company, refusing to do the unglamorous thing because it isn't "product" would be expensive theatre. So I do it.
#The Thursday rule
Envoy ships to production on Thursdays. The whole team knows it. Customer success blocks Thursday evenings for end-to-end checks.
The reason this exists is simple. Without a fixed cadence, every Envoy bug becomes a context switch from Studio. With a fixed cadence, only critical bugs interrupt that flow. Everything else waits until Thursday.
It is the simplest defence I have against being pulled into Envoy full time. The cadence buys me Studio days.
#The eval tax, mostly delegated
Both products are agent-native, which means every new capability comes with an eval and drift surface. In a traditional SaaS prioritisation discussion this would be a tax on every estimate. We move too fast for that to work.
Instead we have independent agents that monitor drift and write evals. The discipline is real, but it isn't humans manually adding test scaffolding to every feature. The work happens on cron, in the background, and I review what the eval agents propose.
So when I'm prioritising, I don't usually have to ask "how much eval work does this add." The system carries that load. Which means the PM-level choice is much closer to a traditional 0-to-1 prioritisation: customer value, deal pressure, strategic surface. I wrote more about the eval setup if that side is what you came for.
#What I actually do, in one sentence
Pick whatever moves the company forward this week, give Envoy a fixed Thursday cadence so it stops eating Studio time, and let the eval agents handle their own tax. Repeat. And try not to wait too long before noticing your first product packaging was the wrong one.