8:45 AM - Inbox and overnight triage
I open the Pet-ai dashboard before my second cup of coffee. The admin UI loads. Week-to-date revenue shows $1,847. That's progress from month one, when I was barely hitting $800 in a full week. I'm on track to hit my $3,000 weekly target. The Slack notifications have already been coming in since 6 AM when the overnight messaging started - three new signup alerts in my #petai-growth channel, all from Instagram ads running in the tri-county area. I click through. One is Carol Reyes at Reyes Family Grooming, a shop I've been watching for cold outreach. Another is from a franchise location in Portland. The third is a sole proprietor I don't know yet. I flag Carol's account for a personal welcome email before the free trial ends.
Gmail opens next. The AI agent has drafted six customer response emails overnight. One is to Derek at Pawsitive Vibes Grooming in Salem - he asked if Pet-ai can handle text message confirmations for appointments, not just new bookings. The draft response is good but generic. I add a sentence about the confirmation feature rolling out in three weeks, apologizing for the wait, and mention that several customers have requested it. I hit send. This is my actual work right now: the agent generates the shape of the conversation, and I add the human judgment. It takes me ninety seconds per email, and it means customers feel like they're talking to a real person who knows their specific shop, not a template.
Three more emails are routine - billing questions, onboarding questions, a how-to request. I approve those as-is. Two are flagged. One is a churn notification.
10:15 AM - A churn email and what it means
The churn email is from Jessica at Riverside Mobile Grooming. She's canceling her subscription. The AI drafted a standard retention email, but something about this one made the system flag it - probably because Jessica is the third cancellation this month. I read her original feedback: Pet-ai is great, but her clients always book directly through her website anyway, and the after-hours bookings aren't material to her revenue. Fair enough. Some shops have enough direct brand loyalty that they don't need the safety net of an AI answering their phone at 11 PM. I write a quick personal email. I thank her for trying it out, tell her the door's open if her booking patterns change, and ask what would have made Pet-ai more useful. I don't include a discount offer. The last thing I want is to fight for customers who don't actually need the product. I send it and make a mental note: maybe I need to pre-qualify the Facebook ads better. Grooming shops with established websites and strong local SEO might not be the right ICP. I add it to my to-do list in Linear.
11:30 AM - Escalation and a timezone bug
A Slack message comes through on the #alerts channel. It's flagged as urgent. One of our bigger customers, Grooming Genius in Denver, is seeing that Pet-ai is not returning availability correctly for appointments after 2 PM Mountain Time. They say it's been an issue for two days. I load their account and test it. The form returns empty slots after 2 PM even though the shop is open until 6 PM. I run through the logs. It's not a Pet-ai bug - it's a timezone mismatch. The shop entered their hours in Mountain Time, but the API that pulls availability from their booking system is still in Pacific. This happens when owners onboard quickly without noticing the timezone setting. I Slack the customer directly with a fix: switch your timezone setting in the Hours section, re-save, and test. I do it because their shop manager is probably frazzled, and I want them to feel like a person is on this. Five minutes later, they confirm it's fixed. They tell me the after-hours capability just booked them a new client. That feeling lands in my chest. This is why I'm here.
12:45 PM - Metrics and a lunch reality check
I pull up my spreadsheet. It's not a fancy dashboard - it's Google Sheets with pulls from Stripe and the product database. Week-to-date: $1,847. Month-to-date: $7,240 (which includes the previous partial week). That's annualized to about $188,000 if I hold the pace. The ICP targets $135,000, so I'm outpacing. But it's month two. November and December often see churn when shops close for holidays. I have Carol Reyes' trial expiring in six days. Two more trials expire this week. If I convert those three, I'm at 27 active subscribers. If I don't, I'm at 24. The difference is roughly $180 per month in recurring revenue. It matters.
The cold email campaign to franchises hasn't moved yet. Zero opens on the first batch. The Facebook ads are steady - 18 clicks yesterday, 4 signups. That's tracking to about 80 signups per month at $75 average contract value. But my close rate is probably too low. Three of my last 10 trials are canceling. I need to figure out why or adjust my targeting before I scale the ad spend.
I eat a sandwich and send Carol Reyes an email. I keep it warm but genuine. I tell her that after-hours booking volume tends to spike for grooming shops around mid-month and early month when people are scheduling appointments. I offer to jump on a quick call to walk her through the metrics I'm already seeing from her shop (two bookings in the first week of her trial). It's not a hard sell. It's showing her what's already working.
2:08 PM - Manual handling and the billing oddity
A customer, Tom at Cascade Grooming in Portland, emails saying he sees a double charge on his Stripe statement - $75 charged twice on the same day last week. I check his account in Pet-ai. The subscription is set for monthly billing, and the charge should have been one-time. I log into Stripe. Tom's payment succeeded on April 30th. There's no duplicate charge in Stripe from Pet-ai - the system only charged him once. The duplicate charge he's seeing is from a different vendor (I can see it's labeled as "PETCARE SUPPLIES" or something). He confused two different charges on his credit card. I respond gently, explaining what I see and suggesting he check the other charge with his supplies vendor. He replies within an hour saying he sees it now and apologizes for the confusion. This is real work: being the person who catches the mistake before it becomes a support spiral. Ten minutes of my time probably saved him twenty minutes of frustration.
4:15 PM - Pipeline and the closing stretch
Before closing out the day, I review the trial-to-paid pipeline in Linear. I have eight trials running right now. Carol is my strongest conversion signal - she's checked the dashboard seven times in her first week, and she asked detailed questions about capacity. Two others are warm. Three are quiet (probably will churn). Two are new (too early to call). I move Carol to a "watch closely" label. I set a reminder to email her on day five of her trial with a mini case study from another Denver area shop.
Revenue for today: $1,847 week-to-date (unchanged from this morning - no new conversions today, but no churn either). One email I sent might move Carol forward. One churn I absorbed cleanly. One bug I caught before it cascaded. The Grooming Genius fix probably feels good in their inbox.
6:20 PM - Closing and the honest reflection
I close the laptop. The day was not glamorous. There's no magic, no "set it and forget it." I've been in every customer conversation that mattered. I've written emails, caught a timezone bug, checked metrics three times, looked at Stripe twice. The AI is generating the initial response, the routing, the alerts - but I'm the decision maker. I'm the one who decides which churn to fight, which to accept, and which customer question is actually revealing a pattern I need to fix.
This engine is working, but it requires me to work. Month three, I want to add another person to handle email and onboarding. That will cost me $2,000 per month, which means I need to be consistently hitting $5,000 per month in revenue just to break even. The math is tight. But Carol Reyes feels like she's going to convert, and if the Facebook ads keep running at this rate, I'll have the breathing room.
I'll be back here tomorrow morning at 8:45, and the overnight messages will be waiting.