HubSpot is Live
We shipped a big one—HubSpot integration is officially live. It’s working, it’s out in the wild, and we’ve already had some great early feedback from customers. That’s always a satisfying moment. The build took months, but shipping it is just the start. We’re already making tweaks and improvements based on real-world use.
This kind of launch is a good reminder of why we do what we do. There’s something gratifying about delivering something customers actually care about. But… this week wasn’t exactly filled with that same kind of payoff.
A Week of Proving (Not Improving)
After the launch, Tyler dove into SOC 2. If you’ve been through it, you know: it’s not about becoming secure—it’s about proving you already are. For a small company like ours, where we built things right from the start, SOC 2 doesn’t change how we operate. It just forces us to show our work.
So instead of building new features or improving UX, Tyler spent four days restructuring AWS networking just so we can check a box. It’s necessary work, but not exactly energizing.
The takeaway: sometimes the job isn’t about value delivery. Sometimes, it’s about grinding through the unglamorous stuff that ensures we’re set up for future wins.
Vibe Coding and the AI Orchestration Analogy
We got talking about “vibe coding”—the trend of letting AI generate entire chunks of application code with little human oversight. It’s a tempting shortcut, but one we’re skeptical of. Not because AI isn’t useful (we use it daily), but because someone still needs to know what the code actually does.
Tyler shared a favorite analogy: using AI today is like conducting an orchestra. If you don’t know how the instruments work, you won’t know when something’s out of tune. That’s the risk: blindly shipping AI-generated code might get something working, but it could open the door to serious security flaws.
Pen Testing: Good, Bad, and Useless
On that note, we’ve done penetration testing recently. And we’ve learned there’s a huge gap between good testers and box-checkers. The good ones find things you didn’t expect—and that’s exactly what you want. The bad ones run automated tools and call it a day. If you’re shopping for a pen tester, choose carefully.
The Future of Junior Roles
We also talked about a broader trend that worries us: AI is really good at doing. That makes life harder for entry-level folks, whether in engineering, content, or marketing. It’s not clear what the new entry point is going to look like, but we hope to be part of shaping it—mentorship, internships, whatever we can do to help.
It’s on founders like us to figure that out.
Workload, Profitability, and a Healthier Growth Model
We ended the week with a conversation about work culture. AI should reduce our workload, but the fear is that in many companies, it will just raise expectations instead. That’s the American way: do more with less.
We’re trying to take a different approach. The trend toward profitability over “growth at all costs” feels like a move in the right direction. It forces better products, better use of resources, and maybe—just maybe—more sustainable work for everyone.
The Other Half of Onboarding
On Ben’s side, this week was all about onboarding. Not the sexy kind. The hard kind. Getting admins trained is one thing—we’ve made that fast. But rolling out Masset across an entire org? That’s where the work really begins.
We’re building more resources, running trainings, and working closely with our customers to make sure the rollout sticks. Because without adoption, even the best product won’t matter.
That said, when it works, it really works. One customer did a rebrand and we saw 65% of their team log in to Masset in a single day. Those are the moments we love.
If you made it this far, DM us the word “spaghetti” and we’ll send you something fun.
Until next week,
Tyler and Ben
