8:15 AM - Overnight triage
I'm on my second coffee by the time I settle into the desk. My Slack notification count is twelve, but most of it is noise - platform alerts and digest summaries. I open my inbox and scan for anything urgent. There's a message from Marcus Chen, VP Sales at Pioneer Technologies, sent at 11 PM last night: "Got the sample module your system generated for our reps. This is actually good. Can we jump on a call today." That's the kind of message that makes the early mornings feel worth it. I flag it and move on.
I pull up the admin dashboard. Yesterday's numbers are in: four new signups, two of them from LinkedIn outreach sequences I ran three weeks ago that are finally converting. The revenue dashboard shows 1,400 dollars hit Stripe overnight - that's Carol Reyes at Reyes Family Practice, who we'd quoted last Tuesday. She signed up for the annual plan, which means she's locked in at the lower rate. I subtract my payment processor fee and do the quick math. Twelve thousand five hundred per month now, running rate. We're tracking toward sixty thousand dollars in annual recurring revenue by end of year if the close rate holds. Still below the eighty-five thousand mark I need to hit, but it's early.
I skim through the onboarding emails. Slack has flagged three messages marked "needs action." The first is a delivery problem from our email service provider - some of our outbound sequences hit spam filters for clients in the financial services vertical. I add it to my Linear backlog. Second is a customer question that came in through the support form: "Can we customize the training modules for our specific sales process." Third is a notification that someone upgraded from the monthly plan to annual midway through their billing cycle. Stripe handled the proration automatically, but I need to confirm the customer got the right invoice.
I should spend more time with the Stripe dashboard this week. The churn rate last month ticked up to three percent. That's not catastrophic, but it's something to investigate.
10:45 AM - Agent review and approval
The system has generated five recommended training modules overnight, based on onboarding data from three new accounts. This is the work I can't automate yet. I open each one in the editor.
The first two are solid. One is a basic objection-handling module for a SaaS company selling project management software. The scenario is realistic - a VP Engineering pushing back on implementation cost. The framework walks reps through reframing cost as productivity gain. The language is natural, not robotic. I approve it.
The third module is a mess. It's about commission structures, and the system has generated an example that references a specific company's pay scale that the rep mentioned in an intake form. I can't send that to their team - it's way too specific and frankly a privacy violation. I flag it and manually rewrite three paragraphs, stripping out the company name and restructuring the scenario as generic. This is the annoying part of the job, the part that requires judgment. I save the revision and make a note to add a content policy check to the generation logic in Linear.
The fourth module I skim and reject. It's on pipeline management, and the system has made it way too long. I leave a note for the system: "The customer is a twelve-person sales team. Keep all modules under three minutes of video. Condense." I delete it and move on. The agent will try again tomorrow.
12:30 PM - The conflict and the manual fix
Lunch can wait. I get a direct message from Sarah at Apex Sales Solutions. They signed up six weeks ago and have been the most engaged customer I have. They've been sending me weekly feedback, which is gold for product development. This message is different. She says one of their reps is complaining that the modules are too generic for their specific vertical - commercial real estate. He's right, and she's right to flag it. They're paying 1,400 a year, which means I don't have much margin on custom work, but I also don't want to lose them.
I pick up the phone and call her. Five minutes later, we have a plan. I'm going to spend two hours this evening recording a custom module for commercial real estate closing dynamics. She's going to have one of their reps send me their most recent deal recording so I understand the specific pain points. In exchange, I'm extending their contract through the end of next year at the same rate. It's not a brilliant business move, but it's the right move. These customers are my feedback loop. Lose them and I lose the insights that make the product better.
I write it up in Linear and set a reminder for 6 PM. Custom module, Apex Sales, due by end of week.
1:15 PM - Metrics and the churn reality
I'm eating a sandwich at my desk. The week-to-date pipeline sits at eight demos scheduled - that's on pace for the fifteen-demo monthly target. I scan through the deal stages. Two of them are getting cold, which means either I didn't follow up or the prospect lost interest. I make a note to send a breakup email to one of them. If they don't respond in three days, I move them to lost.
But there's something harder in my inbox. An email from Derek Shaw at Shaw & Associates came through this morning. They signed up two months ago, around the time I launched. They've used the platform maybe three times. The email is polite: "Not sure this is the right fit for our team at this stage. We're going in a different direction. Appreciate the help." This is what I hate about this business. I spent time getting them live, spent time onboarding them, and they're gone. I do the math on that lost revenue. Fourteen hundred a year. Not huge, but it hurts when you're running at the margin I'm running at. I'll follow up with him personally next week, but I know how this ends.
I check the churn rate again. Three percent this month means I'm losing forty-some dollars per hundred dollars of revenue. If I want to hit eighty-five thousand in year one, I need to get churn under two percent and increase close rate.
3:45 PM - Customer email and quick win
There's a reply in my inbox from Jennifer Voss at Crescent Financial Services. She's been a customer for three months, and last week I sent her a custom module on compliance-aware sales pitches. The email is short: "This might be the most useful thing we've done for our team this year. Three deals have moved because of this conversation piece. How much do we owe you for this."
I don't owe her anything - it's part of her subscription. But the question tells me something. I write back: "Nothing extra. I'm glad it's working. Can I ask you a favor. Would you be willing to spend twenty minutes on a call next month talking about how your team is using the modules. I'm trying to understand what's actually moving deals so I can keep building things like that." Nine times out of ten, this request gets a yes. And those calls are where I hear the stories that make the product better.
5:00 PM - Pipeline review and demo prep
I open my calendar. Tomorrow at 2 PM, I have a demo with Marcus Chen at Pioneer Technologies. That's the one from this morning's message about the sample module. I spend forty minutes building a custom deck for his team. He mentioned they're dealing with a lot of objections around feature parity with a competitor. I pull together three modules that speak directly to that and build them into a demo flow that shows how the system learns from their reps.
I also respond to Marcus: "Tomorrow at 2. I'm building this around the competitor parity objection. Tell me if there's anything else burning that I should address." That's another thing I do that I can't fully automate. Everyone needs to feel like they got a custom experience, because they kind of did.
I glance at the pipeline. Eight opportunities. Three are solid. Three are maybes. Two are probably lost. If I close the three solid ones, I'm at forty-two hundred in new ARR this month. That puts me at fourteen thousand five hundred dollars annual recurring revenue. Still not enough. But it's progress.
6:15 PM - Wrap and reflection
I close the laptop around six-fifteen. There's still the custom module for Apex Sales waiting in Linear. I'll do that tonight after dinner, record the real estate scenario for their team. It's not how I imagined running a SaaS company when I bought the keys to this engine. I thought the AI would do more of the heavy lifting.
In reality, the AI generates three things that have real value: it creates draft content and modules, it flags which leads are ready to move forward, and it surfaces patterns in what's working. But everything it touches still needs a human to verify it, adapt it, push it in the right direction. The system can write a training module but not understand if it's actually going to change a rep's behavior. The system can identify a hot prospect but not know if the timing is right or if there's a competing priority we don't see.
What worked today: the phone call with Sarah, the reply from Jennifer, the overnight signup from Carol Reyes. Those are the threads that matter. What's hard: the Derek Shaw churn, the spam filter issue, the fact that I'm still doing more manual work than I'd like.
Second month in and I still don't know if I've got a real business yet. But twelve thousand five hundred a month is more than zero. And Marcus Chen is ready to talk tomorrow.