8:42 AM - Inbox triage
I open my laptop at the kitchen table and reach for cold coffee from yesterday. The Slack notification badge shows 12 unread messages. Most are system alerts from the Lead Router dashboard: three new signups came through overnight, two from LinkedIn profiles that matched our ICP filters, one from a cold email reply. The revenue column in my head jumps from what it was yesterday. I pull up the Stripe dashboard in a new tab.
Today's new customers: Marcus Chen, RevOps manager at a 180-person martech firm. Sarah Torres, Sales Ops lead at a mid-market healthcare tech company. Neither signed up for a demo yet. They just activated free trials, which means they got curious enough to enter a credit card. Week-to-date revenue sits at $1,247. That's good. That's tracking toward the monthly target.
I check Gmail next. The AI agent, which I built to handle prospecting and outbound, has prepared six email drafts to send this morning. They're sitting in a shared folder I access from my email client. I skim the subjects: one to a prospect named Carol Reyes at Reyes Family Practice, one to David Kim at a 250-person SaaS company, four others. I need to review each one before they go out.
10:15 AM - The flagged email and a tricky decision
The third draft, to a prospect named Lisa Wong at a Chicago-based fintech, has a red flag in the AI's summary. The agent found Lisa on two different LinkedIn searches: once as a Sales Ops manager, once as an Operations Director. Same person, same company, same email address. The AI flagged it because it's unsure whether to mark this as one prospect or two. If we send two emails, Lisa will notice the duplicate, and we look sloppy. If we send one, which message do we use.
I open the draft marked "Sales Ops angle" and the one marked "Operations angle." The Sales Ops version emphasizes lead routing efficiency and HubSpot integration. The Operations angle talks about reducing manual data entry and scaling processes. They're different pitches for the same person. I need to decide.
I think about what I know about Lisa. Her LinkedIn profile says Sales Ops is her primary title. But the Operations draft isn't wrong - it's just a different hook. I click into her profile directly from our system. She's been at the company for three years, moved up from Sales to RevOps to her current role. She probably cares about both angles, but Sales Ops is her identity.
I edit the Sales Ops draft in Gmail, making it more specific: I reference the fact that she's scaling a team from 80 to 120 people in the next six months. That's a signal I found in an archived company blog post. Now it doesn't feel generic. I approve it, delete the Operations version from the queue, and move on. Two minutes well spent.
Carol Reyes' email looks good. It references her practice's recent expansion and mentions a specific feature of Lead Router that would solve a problem I know she has. David Kim's email is solid too. I approve five of the six. The sixth, to a prospect at a company that just laid off half their sales team, I reject. Bad timing. We'll revisit in two months when they've stabilized.
12:30 PM - Lunch and the metrics check
I step away for lunch, come back, and open the Lead Router dashboard on a second monitor. It's the admin UI I built and continue to maintain. I check the metrics module. Today's numbers so far:
3 new signups. 2 of those converted to paid trials already. Week to date: 12 signups, 7 paid conversions, $1,247 in MRR from new customers. Pipeline for next 7 days: 5 demos scheduled. Active customer count: 38.
That's solid. We're running slightly ahead of our monthly close target. If I get four more closes before the end of the month, I'll have six for May. That moves us forward.
I check my Linear instance for any bugs logged. There's one new issue from Sarah Torres actually, the healthcare tech Sales Ops person who signed up this morning. She says the API integration with her Salesforce instance is returning empty records for a specific field mapping. I add it to my queue. Second bug this week. That's normal.
2:08 PM - An unpleasant conversation with a churning customer
Slack notifies me that a customer, James Murphy at Murphy Digital, has sent a cancellation request. James has been with us for four months. He was paying $200 a month. I immediately open an email and write to him directly, not a template. I ask what's going wrong.
He replies within minutes: "Your product works, but we integrated HubSpot Workflows natively instead. We're doing the same thing you do, and it costs us nothing now."
That stings a bit. I write back acknowledging that he made the right choice for his situation. I don't push to keep him. I cancel his account, process a refund for the month he won't use, and add a note to revisit him in six months. Sometimes the honest move is the cleanest one.
I lose $200 in monthly revenue, but I gain a reference customer who will respect the decision. I document it in a Slack channel where I track churn: "HubSpot native integration made us redundant. Monitor: other customers asking about native HubSpot integration."
4:30 PM - The bug fix and a win
I pull up the Sarah Torres issue again. I dig into the Salesforce API logs and see the problem: a field mapping I set up uses a deprecated Salesforce field name. I push a fix, test it against my own Salesforce sandbox environment, and send Sarah an email telling her it's resolved. She replies within an hour. "That was fast. Thanks."
It's a small interaction, but it matters. She's day-two of being a customer, and we've just proven we respond and fix things quickly.
I spend twenty minutes reviewing the agent's outbound performance. Of the twenty emails it sent yesterday, three got replies. One is a warm intro. One is a "not now, ask in Q3." One is a rejection. The conversion rate is holding steady at about 15 percent. That tracks with what I need.
5:45 PM - Pipeline and a moment of clarity
I pull up my pipeline view. Next week has five demos scheduled. That's two more than the week before. If even two of those convert, we're on track for eight closes this month. That would put us at a $2,400 MRR increase, moving total MRR closer to $11,000. That gets us to a $132,000 ARR run rate.
I think about what I've done today. I rejected one bad email. I solved one customer's tricky decision. I fixed one bug. I lost one customer but kept our integrity. I approved five outbound emails that the AI drafted. I reviewed metrics three times. I read Slack.
This is the job. It's not a robot running while I sleep. It's me, amplified by an AI agent that handles volume and pattern matching, but still me making judgment calls. The agent finds the prospects and drafts the emails. I decide which ones feel right. The dashboard alerts me to problems. I decide which ones to solve. The customers email when something breaks. I respond.
It's real work, but it's sustainable work. I'm not manually prospecting or typing every email. I'm not managing a sales team or taking calls all day. I'm orchestrating, reviewing, and fixing at scale. It's why I bought this thing.
6:15 PM - Closing the laptop
I close the dashboard, close Slack, close Gmail. The MacBook goes dark. I've hit the target. New customers are signing up. Pipeline is building. Revenue is moving. One customer left, which was the right call. One bug got fixed same-day. Tomorrow is Wednesday and I'll do this again.
What would I change. Maybe I'd spend more time on the long-tail customers, the ones who aren't scaling fast, who are happy but quiet. I could probably find more expansion revenue there. Maybe I'd automate the daily metrics review instead of opening the dashboard three times. The numbers are good enough. I don't need to check them that often.
But mostly it works. I built a machine that finds people who need what we make, and I get to decide whether we actually reach out to them. That's a good business. For two hundred dollars, I'm not sure I'd find better.