8:42 AM - Dashboard open
I pour coffee and pull up the admin dashboard. It's 6:52 AM in Seattle, so the East Coast is already awake. Three signups came in overnight - two from Facebook outreach, one from a Google search. Total revenue showing for today: zero dollars yet. Week-to-date: $347. One of those is Carol Reyes from Reyes Family Landscaping, who signed up Thursday and already paid her first month. The other two are still on free trial, waiting for the actual software to land in their inbox.
Slack pinged me at 6:15 AM. The agent flagged five cold-outreach emails for review before sending. I scroll through. Four of them are solid. Professional tone, clean, mentions of specific pain points. The fifth one to Marcus Chen at Chen's Outdoor Design has a problem: the agent pulled his email from last year's licensing database and called him "owner of Chen's Outdoor Solutions." The company changed names in 2024. Sending that would tank the demo rate.
I delete that one and approve the other four.
9:15 AM - The agent's work
The approved emails go out in the next batch. I watch the Slack feed as the agent's system logs show them being delivered. Four cold outreach emails, each one personalized with something real about their service area or company size. The agent pulled these details from public records and their websites. It took me about ninety seconds to review what the system spent two minutes building and fact-checking.
This is the work. Not the sending. The filtering.
By 9:30, I'm into my email. One customer, Tom Wallace of Wallace Bros Landscaping, wrote back to yesterday's cold email. Not a demo request. He wanted to know if the software works with his existing CRM, which is some old proprietary system from 2008. I draft a response. Not the agent - me. This one matters. I tell him we work with most APIs but can't guarantee legacy systems. I ask what his actual pain point is. Maybe he doesn't need CRM integration. Maybe he needs something else entirely.
I hit send and move on.
10:25 AM - The conflict
Slack alert. The agent caught an issue in the database. One of our current customers - Carol Reyes again, the one who signed up Thursday - hasn't activated her account yet. She paid, but she's still in the free trial flow. It's been sitting there for three days. The agent flagged it and suggested an automated email asking if she needs help onboarding.
This is where I have to think. A bad automated email ruins the relationship. I look at her timeline. She signed up Thursday, paid Friday morning, and hasn't logged in. She might be busy. She might be overwhelmed. She might have changed her mind but forgot to cancel.
I send her a personal email instead. Short. I introduce myself by name. I tell her I noticed she hasn't logged in yet, and I'm here if she has questions before getting started. I don't assume anything. I ask.
12:15 PM - Metrics and money
Lunch is sitting at my desk. I pull up the Stripe dashboard. This month we have six active paying customers. Three are on the $120/month plan. Two are on a discounted $80/month (they're seasonal - they'll churn in August probably). One is on a custom $200/month because he's in a bigger market and needs priority support.
Total MRR right now: $640. That's about forty percent ahead of where the projections said I'd be after two months. The spreadsheet I built before buying this shows Year 1 ARR should hit around $108k if I keep conversion steady. At this rate, I might hit $105k. That's on pace.
But this is my second month. The real test is Q2. Landscape contractors are underwater right now with spring work. Demos are harder to book. I booked fourteen demos this month instead of the projected twenty. Only three closed. At least they're sticky - churn so far is zero.
I check the Google Analytics data that feeds into the dashboard. Someone from a landscaping company in Portland searched "lawn care management software" and landed on the pricing page. They spent two minutes there. No signup. I make a note to add a video to that page this week.
2:40 PM - The hard one
Email from Derek Powell at Powell Landscaping. He signed up six weeks ago. Paid two months. Didn't actually use the software, and now he's writing to cancel. He says he doesn't have time to learn new software when he's in the middle of the spring rush.
This one stings. He's right, too. I built the onboarding flow, but it assumes people have forty-five minutes to get started. In March and April, these guys have twenty minutes tops. They're in trucks, not offices.
I respond to Derek directly. No automated drip. I tell him I get it. I ask if he'd be willing to do a quick fifteen-minute call where I show him the two features that would actually save him time. I tell him I'll refund his second month if the call doesn't change his mind.
He writes back within an hour saying yes to the call. It's Thursday at 2 PM. I add it to my calendar.
3:45 PM - The fix
While waiting for the next batch of signups to roll in, I look at the bug Slack flagged yesterday. The export-to-PDF feature is breaking on mobile. It's throwing a JavaScript error when users try to download estimates. I don't code anymore - I pay someone for that - but I can see the issue. The agent described it correctly in Linear. I move it from "To Do" to "In Progress" and send it to Dev with a note saying this is a revenue issue, not cosmetic.
This is real work. Triage.
4:30 PM - Pipeline review
Carol Reyes wrote back. She said, "I wasn't sure I was worth the learning curve. Your email helped. I'm diving in tonight."
That's the good part.
I open the pipeline board. Sixteen prospects on it. Tom Wallace, who asked about CRM integration, is in the "Demo Scheduled" column now. I booked him for tomorrow at 10 AM. Two others are "Waiting on Email Response." The agent sent cold outreach two weeks ago. One of them, Sarah Kim at Kim Landscape Design, just opened the email an hour ago. Slack alerted me. I make a note to follow up Thursday if she doesn't reply.
The math is simple. I need four closes per month to stay on pace. I have two closes this month and three demos booked for next week. If I close one more this month, I'm solid.
5:50 PM - Wrapping
I close my email and Slack. The admin dashboard is still open. Today's revenue: $120. That's Carol's first month, which hit the account this morning. The system charged her automatically via Stripe. No manual work. No invoice chasing.
Week-to-date: $467. Month-to-date: $640 from recurring customers, plus $120 from Carol's new payment, plus the $200 I paid for the business itself in the first place.
Running Landscape AI isn't automation. It's a person with an AI assistant. I spend maybe two hours a day on this. I review emails. I handle the customers who need human judgment. I catch the edge cases. I watch the numbers. I book demos. I make one customer save their own time by sending them a fifteen-minute video instead of a full tutorial.
The agent does the work. I do the quality control and the strategy.
By 6:15 PM, the laptop closes. Tomorrow I'm chasing down that PDF export bug with Dev, onboarding Carol properly, and prepping for Tom's demo. The next close is probably him. The one after that is probably Derek, if that Thursday call goes well.
This is the day. Real work, real customers, real money. Not glamorous. Not passive. But honest.