8:42 AM - Dashboard check
I pull up Moving AI on my second monitor while I'm still on my first coffee. The numbers hit me first: 4 new leads came in overnight through the Facebook group outreach we set up last month. The system has already drafted personalized quotes for three of them based on their inquiries. The fourth one is flagged - incomplete address data - so there's a note waiting for me to review it manually. $850 in revenue showing for this month so far, which means we're on pace for solid numbers.
My inbox shows 12 new emails. I know from the Slack alert that came through at 6 AM that three of these are from the Moving AI system itself - approval requests for outbound follow-ups it wants to send to leads who haven't responded in 48 hours. The other nine are a mix of customer questions, payment notifications from Stripe, and one message from Mike Torres at Torres Moving Solutions. That one makes me pause. I click it.
He's asking about a feature request: can the dispatch system alert his crews when a job assignment comes in, instead of just updating the dashboard. Right now Mike's team has to check the app, and he says they're missing notifications. I make a mental note. That's a real pain point, and I've heard it from two other customers already. I flag it for the product team's feedback queue.
10:15 AM - A conflict I have to untangle
I get into the review queue for the outbound emails. The first one is solid: a follow-up to Carol Reyes at Reyes Family Practice about a quote we sent three days ago. The draft is clean, personalized, references the specific estimate amount, includes a clear next step. I click approve. Five seconds of my time, probably closes into a $1,200 moving job in the next week.
The second one is trickier. The system drafted a follow-up email to Robert Chen at Chen & Associates, but I notice something in the timeline: our initial quote went out at 8 PM on a Thursday. Now it's Tuesday. The follow-up is worded as if we're checking in after a few days, but I can see from the email log that the system actually received a reply from Robert on Friday saying "very interested, want to call this afternoon."
There was no call scheduled in the system. Robert replied but nobody tracked the promised callback. Now the system wants to send a generic follow-up email that completely misses the fact that Robert's waiting on me. I delete the draft. I open Gmail and write Robert a new email from scratch: apologize for missing the call, ask for his availability this week, keep it short. This is the kind of thing that makes or breaks trust with a customer.
I send it and make a note in Slack to set up a proper integration between our calling system and the lead database. That's a gap that matters.
12:30 PM - Lunch and the metrics check
I pull the weekly dashboard while eating. Week-to-date: 14 new leads, 5 have moved to quoted status, 1 has closed. Revenue this week is $2,100, which includes recurring subscriptions from two existing customers plus that one new moving job. The system is handling about 60% of the follow-up work I would otherwise be doing manually. The other 40% is still me - escalations, edge cases, calls that need a human voice.
The churn number sits at the top of the dashboard in red. One customer dropped last week: Jake Morrison at Morrison Mobile Moving Services. He said our dispatch alerts weren't integrating cleanly with his crew communication system. He paid for two months and left. That one stings, and it's the reason Mike Torres' feature request is stuck in my head.
I open Stripe to look at the cash flow. Four active subscriptions at $329 per month each means $1,316 in monthly recurring revenue. That's solid for month two. If I can keep churn below 10% and add one new customer every 10 days, I'll hit the growth curve the pitch deck promised.
2:08 PM - A customer with a problem
The Slack channel for active customers pings me. Jennifer Hart at Hart Family Movers posted a question: one of her dispatched jobs is showing in the system but the crew says they never received the notification. It's a 3-person move scheduled for tomorrow morning at 9 AM in Oakland. They're checking their phones, emails, and the app - nothing arrived.
I log into the admin dashboard and pull up the job. It was assigned 4 hours ago. The system shows "notification sent" but the log is blank. There's no record of it actually reaching anyone. This is critical. If the crew doesn't show up tomorrow, Jennifer loses the job and I lose a customer.
I call Jennifer directly. I tell her the system had a delivery failure on that notification, I'm manually alerting the crew right now, and I'm personally staying on top of this until the team confirms they're locked in. I don't want to be that vendor who hides behind automation. I jump into the job and mark the crew as manually notified in the system. I send Jennifer a follow-up email with screenshots showing what I did and a commitment to investigate the notification gap by end of week.
The root cause is probably a webhook timeout in the integration between our notification service and the app. It's on my list to dig into, but not now. Now I needed Jennifer to feel safe.
4:30 PM - Pipeline review and a win
I spend 30 minutes looking at the full pipeline. 14 leads in inquiry status waiting for quotes. 5 in quoted waiting for responses. 2 in negotiating where customers have replied but haven't committed. 1 that closed this week: that's the $1,200 job from Carol Reyes. She'd sent a message thanking me for the quote and the follow-up email. The message was short: "You made this easy. Booked you guys in for Saturday."
Small thing. I read it twice. This is why I'm running this instead of staying in my day job.
I run the numbers one more time. If the 2 in negotiating both close at average deal size of $1,100, that's $2,200 more coming through. The 5 quoted jobs represent about $5,200 in potential revenue. It's early, the funnel is real, and the system is actually amplifying my work instead of replacing it.
6:15 PM - Wrap and tomorrow's list
I close the laptop around 6:30. The day felt like work - real work, not passive. I reviewed 12 emails, approved 2 follow-ups, fixed 1 customer crisis, wrote 2 manual emails, and dug into the pipeline for 30 minutes.
The system moved leads forward. The system drafted quotes and follow-ups. But I had to sit in the middle and make sure it wasn't missing context, wasn't sending the wrong message at the wrong time, wasn't failing in ways that would wreck a customer relationship.
By the end of the month, if we close half of what's in the pipeline, we'll hit $4,500 in new revenue. Churn worries me. Jake's departure showed me that the product has real gaps. But Jennifer Hart seemed solid after that call. Mike Torres is asking for features because he wants to use the system better.
Tomorrow I'm diving into that notification webhook issue and I'm building a simple dashboard for customers so they can see their own funnel. Small things. But they're the difference between this working and blowing up.