The Builder-Seat Trap
I had a SaaS app ready to submit. 131 tests passing. Payments working. The thing was done. I'd built it alone using Claude Code over about three weeks, and by every measure I know how to check, it was ready for the App Store.
So I spent the next three days building cold email sequences for my consulting business instead.
I polished a confession about not shipping while continuing to not ship.
I don't mean I took a break. I mean I opened a new project, pulled up my prospect list, and started designing outbound workflows for staffing agency owners. Real work. Genuinely useful work. Work that needed to happen eventually. I just happened to start it the same week the app was sitting there waiting for me to press one button.
When I came back to the app, I ran another bug sweep. My own codebase told me the false positive rate was higher than the real bug rate. I told it to look deeper. It found 130 accessibility issues that changed nothing about how the app functioned. I fixed every single one of them. I created a file called "Final Audit." Then another one. Then another one. I ended up with five separate "Final Audit" files, each one a slightly different version of the same conclusion: the app was ready to ship and I was not ready to ship it.
Then I wrote a LinkedIn post about the experience instead of submitting.
Then I spent five rounds editing the LinkedIn post between two different AI sessions, polishing a confession about not shipping while continuing to not ship. My AI collaborator eventually said something like, "You have five Final Audit files for the app. You're about to have five final draft versions of a post about having five Final Audit files." That was accurate and I still didn't submit.
I run a consultancy that fixes sales systems. I diagnose exactly this problem in other people's businesses. I'm good at it. I sit down with a founder who's been building dashboards for six weeks instead of calling their first ten prospects, and I can see it immediately, the whole pattern, the way productive work becomes a hiding place. I can name it in their business in about fifteen minutes.
I apparently cannot name it in my own business while it's happening.
I've started calling it the Builder-Seat Trap, because I needed a name for it and because "productive procrastination" doesn't capture what's actually going on. Procrastination feels bad. You know you're avoiding something. You feel guilty. You watch TV or reorganize your desk and there's a low-grade hum of shame underneath it. The Builder-Seat Trap doesn't feel like that at all. It feels like working. It feels like progress. You're producing real output, things are getting checked off, files are shipping, and the feedback loop is telling you that you're doing great. The problem is you're solving the wrong problems and the reward signal is identical.
01Fifteen automated jobs for a one-time cleanup
A client asked me to help deduplicate their contacts. One-time cleanup. They had about 3,400 records with maybe 600 duplicates, the kind of thing you can do in a spreadsheet in an afternoon if you're careful. I built 15 automated deduplication jobs. Fifteen. For a one-time cleanup on a database that hadn't changed in four months and probably wasn't going to change significantly in the next four. I automated a problem that was going to happen once. I could feel how productive I was being the entire time. Every job I configured felt like progress. The client got a clean database, which is what they wanted, and I had built a system designed to solve a problem that didn't recur, which is not what anyone asked for.
I also built a system to connect two Claude instances to each other. A bridge that would let one session pass context to another. Technically interesting. I spent maybe a day and a half on it. It produced nothing usable for any client or for Outblox. I couldn't tell you today what the intended use case was. I think I just wanted to see if it would work, which it sort of did, and that was enough to keep me in the chair for the duration.
02Building downstream of nothing
The pattern is always the same. Something hard is sitting there waiting to be done. Not hard in a technical sense, hard in an exposure sense. Submitting the app means finding out if anyone wants it. Sending the outbound sequence means finding out if the messaging works. Following up with the prospect means finding out if they're actually interested or just being polite. Those are the moments where you learn something real about your business, and they're the moments I will build literally anything to avoid.
I built an entire prospecting playbook for Outblox. Identified the right staffing agencies. Designed the sequence logic. Refined the email templates. Organized it all in Apollo. The sequences sat there unsent for weeks while I refined them further. At one point someone I work with told me "content without a running outbound pipeline is theater" and I remember agreeing with that and then going back to adjusting the templates.
I built analytics infrastructure for traffic that didn't exist.
I built a visitor behavior tracking system for outblox.com. Smart CTAs, referral source detection, high-intent pattern webhooks. The site was getting maybe twelve visits a week. I built analytics infrastructure for traffic that didn't exist. The pushback I got was something like "when you're getting 50 visits a week, deploy this, that's probably four to six weeks out if you start actually sending outbound." Which was correct. The tracker still isn't deployed because the traffic still isn't there because the outbound sequences still weren't running when I built it. I was building downstream of a problem I hadn't solved upstream, and it felt exactly like solving the upstream problem because I was working hard and producing real things.
I saw a competitor launch an AI training course. Four hundred signups, six sessions, a whole roadmap. For about two hours I was seriously considering building a course through Outblox. An entire curriculum. Lesson plans. A community. This was, I should mention, before I had sent my first Apollo sequence to a single prospect. I was going to build a six-session educational product before I had verified that any human being wanted to buy the thing I was supposedly going to teach them about.
03The same trap from the outside
The reason I'm telling you all of this is because I watch the same thing happen in every small sales team I work with, and it looks different from the outside but the mechanics are identical.
A founder I was watching, this was maybe six months ago, spent three weeks integrating Clay with HubSpot. Building the data pipeline. Mapping fields. Setting up enrichment triggers. Beautiful system. I asked him what his best customer looked like and he didn't have an answer. Not a vague answer, no answer. He had built the system that would operationalize his ICP before he'd figured out what his ICP was. He was in the builder seat. The integration was real work. It would've been genuinely useful eventually. But he was doing it because connecting APIs feels like building a company and calling strangers to ask what they need feels like begging.
The output was real. The commits were real. It just wasn't connecting to the thing that would've told them whether the business was viable.
I watched a different founder sort of fail in slow motion over a few months. Kept building. Kept shipping features. Kept refining the go-to-market materials. Never quite got to the part where someone outside the building had to say yes or no. Everybody around the situation could see it and nobody could figure out how to say it because from every normal angle the person was working incredibly hard. The output was real. The commits were real. The Notion boards were full. It just wasn't connecting to the thing that would've told them whether the business was viable.
I see this in sales teams constantly, too. A 15-person staffing agency where the reps spend Monday morning building call lists instead of calling. Where the manager builds a new dashboard instead of sitting with a rep on a live call. Where someone spends an afternoon setting up a new sequence template instead of sending the 47 emails already drafted in the old one. It's all real work. The list needed building. The dashboard will get used. The template is probably better than the old one. But the ranking of those tasks against "just call the prospects you already have" is wrong, and nobody catches it because everyone's busy.
04Whether knowing helps
The thing I still can't figure out, and I've been thinking about this for a while, is whether knowing about the pattern actually helps. I know about it. I named it. I wrote a whole piece about it. I still do it. I did it last month. I'll probably do it again before this piece goes up on the site. The awareness doesn't seem to create immunity. It just means I can describe what's happening to me while it's happening, which is a weird kind of progress that I'm not sure counts.
I think part of the problem is that the builder seat is where I'm best. I'm good at building systems. I'm good at configuring tools. I'm good at designing workflows and writing automations and setting up integrations that actually work. That stuff comes naturally and it's satisfying and I'm fast at it. The other stuff, the selling, the following up, the putting your name on something and waiting to see if people care, that's where I'm average at best. And the rational thing would be to spend more time on the thing I'm worse at because that's where the growth is, but nobody actually does the rational thing when the irrational thing feels like winning.
Nobody actually does the rational thing when the irrational thing feels like winning.
I keep a running list in my head of the things I've built that I didn't need to build yet. The tracker for a site with no traffic. The automation for a one-time task. The course outline for a service nobody had purchased. The bridge between two AI instances for no reason. The five Final Audit files for an app that was ready on audit one. It's a long list and it gets longer most months. Some of those things eventually became useful. The prospecting playbook did get used once I finally started sending. A few of the automations saved time later. But the timing was wrong on almost all of them, and the timing being wrong is the whole point. I built them when I did because building them was easier than doing the thing I was supposed to do next.
There's a version of this piece where I tell you I've figured it out. Where I give you the framework I use to catch myself, or the three questions I ask before starting a new project, or the accountability system that keeps me honest. I don't have any of that. I have a consultancy that diagnoses this in other people's businesses and a personal track record of doing it repeatedly in my own. I have an AI collaborator that flags it sometimes, which helps in the moment and apparently doesn't help two weeks later when the pattern reasserts itself. I have the name, the Builder-Seat Trap, which at least means I can say "I'm doing it again" out loud instead of just feeling vaguely like something's off.
05How most of these end
The app did eventually get submitted, by the way. I don't remember what finally pushed me to do it. It wasn't some breakthrough moment where I confronted my fear of judgment or whatever. I think I just ran out of things to audit.
That's probably how most of these end. Not with a decision to be brave. Just with the builder seat running out of things to build and the real work being the only thing left.
I have a client call on Thursday and I'm pretty sure I'm going to spend Wednesday evening redesigning how I present audit findings instead of reviewing the actual audit data. I already know this about myself and I'll probably do it anyway. The slides don't need redesigning. The data is ready. I'm going to do it anyway.
If your team is building instead of selling: demo.outblox.com