I had a laptop, a few client deadlines, and the loose plan of being back in San Diego by Thursday. That was a lie I had been telling myself for years, in different shapes. The “working trip” framing was just permission to leave without making it mean anything.
It meant something.
Most of what came with me on the move fit in two suitcases and a backpack. Everything else got sold, donated, or stuffed into the storage unit I keep telling myself I’ll empty one of these days. The seven things I actually used in the first month:
That list is the rest of the series in miniature. What I thought I’d need versus what I actually used. Who I thought I was versus who I am when nobody’s watching.
I’ve spent the last few years building businesses, learning AI, going to networking events, and giving advice about how to grow a service business. Most of that advice still works. But the version of me giving it has changed in ways the LinkedIn post version of my life never quite captures.
This series — “Falling In, Falling Out” — is about the part that doesn’t fit on a LinkedIn post. Moving to Mexico. Falling for the country. Falling for someone here. And then what happened when the version of the story I was telling myself stopped matching the version that was actually happening.
It’s not a confessional series. It’s a what-I-learned series. The lessons just happen to come from places no business book is going to send you.
A dozen or so posts, roughly monthly. Some about Mexico — the food, the rhythm, the weird gringo loneliness that hits when you realize your Spanish is good enough for the carnicería but not for the conversation you actually want to have. Some about relationships — not the gory details, but the patterns I keep noticing about how I show up, when I leave, and what I think love is supposed to look like when I’m being honest.
If you stick around, I’ll try to make it worth the click.
— Michael
]]>I changed both within forty-eight hours.
La Paz is a real city. About a quarter million people, a working downtown, decent espresso if you know where to look, traffic that’s irritating in the way only mid-sized-city traffic is. It is not Cabo, which is the comparison everyone makes — Cabo is a built-for-tourists strip; La Paz is a place where actual Mexicans live and have lived for several centuries. The difference shows up everywhere. Mercados that exist for the people who shop there, not for the cruise crowd. Restaurants that don’t have menus in English unless you ask for one. A pace that’s slower than San Diego but not performatively slow.
The malecón is the thing the brochures get right. Two miles of palm-lined waterfront with the kind of sunset that makes you understand why the conquistadors named the sea what they did. Mar de Cortés. The water turns four different shades of blue between noon and sundown and I still don’t have a word for the one it goes when the light starts to slant. I walked it every night for a week and never got tired of it.
What got me to extend the trip wasn’t the views. It was the food. I had a carnitas taco at a stand on Cinco de Mayo street that I’m still thinking about months later — fatty, crisp at the edges, dressed with nothing but lime and a salsa that was just chiles and salt. It cost me about thirty pesos. I ate three.
The thing that made me stay an extra week, though, was a different kind of feeling. After three days, the man at the corner tienda recognized me and started pulling out the brand of coffee I’d bought the previous morning before I asked. After four, the woman at the aguas frescas stand by the marina started making the jamaica without sugar because that’s how I’d had it the first time. Nobody at home in San Diego knows how I take my coffee. I’m not exaggerating.
I extended the flight. Then I extended it again. Six weeks later I still hadn’t been to Todos Santos.
The city keeps surprising me. Last week I found a record store run by a guy who’d been a session musician in Mexico City in the eighties and had opinions about every album in the place. The week before, I stumbled into a posada with a brass band that played until three in the morning. Last night, a fisherman at the marina told me an entire history of the Sea of Cortez fishery — the bonanza years, the collapse, what’s coming back — while we waited for a boat to refuel.
This place is loud and slow and quiet and busy and I’m not done figuring it out.
]]>This list is the inverse of that. These are the five tools I see survive the 90-day test in actual small businesses I work with — agencies, service shops, consulting practices in the $500K to $10M range. The ones that don’t get cancelled at renewal because somebody can’t justify the line item.
Claude. I’ll get the obvious one out of the way. Anthropic’s Claude has become the default for writing, editing, drafting proposals, summarizing client calls, and thinking out loud. The reason it sticks where ChatGPT often doesn’t — and this is opinion — is the output reads less like a press release. For service businesses where every email goes out under your name, that matters. The honest limitation: it does not replace a domain expert. It accelerates one.
GoHighLevel. Not new, not sexy, but the most-installed CRM and marketing automation platform I see in the agencies I talk to. It bundles email, SMS, funnels, calendars, and a pipeline into one tool, and the AI features added in 2025 (auto-replies, conversation summaries, voice transcripts) shifted it from “a thing my agency uses for me” to “a thing I can run myself.” Real ROI shows up at month three when the SMS follow-up sequences start filling the calendar without anybody touching them.
Descript. If you make video or podcasts, this is the one. You edit by editing the transcript, the AI removes ums and pauses, and the studio-sound feature is genuinely indistinguishable from a treated room for most use cases. I’ve watched two-person agencies go from “we’d love to do video but we don’t have the time” to publishing weekly because Descript collapsed a four-hour edit into forty minutes. The limitation: it’s a tool, not a strategy. It won’t tell you what to record.
Perplexity. This one snuck up on people. Perplexity does AI-summarized web search with sources, and it’s quietly replaced Google for a specific job: the “I need to understand a competitor / a market / a regulation in fifteen minutes” job. For agency owners pitching new verticals, it’s the difference between two days of research and an hour. Limitation: it’s only as good as the public web on a given topic. For B2B niches with thin online footprints, you’ll still need to call people.
n8n. The unsexy fifth pick. n8n is open-source workflow automation — Zapier, but you self-host and the AI nodes don’t bill per call. The reason it lands here is the cost curve. At ten workflows running ten thousand operations a month, Zapier is a meaningful expense. n8n is twenty bucks of server time and a weekend learning the interface. For agencies operating on margin, it pays for itself in the first month. Limitation: somebody has to learn it, and that somebody usually isn’t the owner.
That’s the list. The pattern, if you want one: every tool here either replaces a recurring time cost (Claude, Descript, Perplexity, n8n) or replaces a recurring software cost (n8n, GHL). Anything that doesn’t do one of those two things — and an alarming amount of “AI for business” software doesn’t — gets cancelled by month three. The five that do tend to survive.
]]>What was my experience: I joined a chapter, showed up every week, gave my 60-second commercial, sat through other people’s, did one or two 1-2-1s, didn’t pass meaningful referrals, and didn’t receive any. Six months in, my honest assessment was that I’d spent about a hundred hours and had a few new business cards and not much else. I came close to dropping out twice.
I didn’t. And in month seven something changed, and then in month nine it changed again, and by the end of year one I was passing and receiving referrals every other week and my agency was getting business that the rest of my marketing wasn’t producing. So the question I’ve been asked a few dozen times since — what flipped? — is the actual point of this post.
Three things flipped, and they’re connected.
The 1-2-1 is the whole game, not the meeting. This was the biggest reframe. In the first six months I treated the weekly meeting as the thing and the 1-2-1s as homework. That is exactly backwards. The weekly meeting is where you maintain visibility — important, but on its own it produces almost nothing. The 1-2-1 is where you actually learn what someone does, who their ideal client is, what a referral feels like to them, and what you’d need to see to send one. Until I started treating 1-2-1s as the work and the weekly meeting as the warm-up, nothing happened. Once I did, the curve bent quickly. I now do at least two 1-2-1s a week with people in or adjacent to my chapter, and they are by far the highest-leverage hour of my week.
Referrals compound when you know someone’s ideal client better than they do. This sounds backwards. It is not. Most business owners describe their ideal client in vague aspirational terms (“any small business owner who wants to grow!”). When you push them in a 1-2-1 — what’s the actual size, vertical, situation, trigger event — you almost always learn more than they’ve articulated, and you can hold that knowledge for them. Now when someone in your network mentions a person who fits, you spot it before they do. That ability is the single most valuable thing a BNI chapter teaches you, and most members never develop it because they don’t push past the polite version of the question.
Showing up consistently matters more than any single introduction. The math of referrals is boring and unfair. If you show up to forty meetings in a year, give a sharp 60-second every week, do two 1-2-1s a week, follow up on every referral within 24 hours, and don’t drop the ball when someone refers to you — you will end the year with momentum no individual brilliant introduction can match. I have watched a member give a knock-it-out-of-the-park pitch about a perfect product, generate a flurry of interest, and produce nothing because he didn’t follow up cleanly. I have watched another member give the same fine pitch every week for two years, never miss a meeting, return every email same-day, and quietly close more business than anybody else in the chapter. The reliable one wins.
That’s the stuff I’d go back and tell month-one me. None of it requires being a great salesperson, charismatic, or extroverted. It requires showing up, doing the unglamorous work, and treating other people’s businesses like you’d want them to treat yours.
The corporate version of BNI is fine but the real story is just consistency. If you’re new and frustrated at month four, that’s normal. Keep going. Make the 1-2-1s the thing.
]]>The lazy version of all this is to make people pick. Have a website for the agency, a Substack for the writing, a separate brand for the speaking, a recipe Instagram nobody finds, a LinkedIn that sort of holds it together. That’s what I had for a while, and what it does is teach the algorithm — and people — that I’m a marketer, full stop. Everything else gets quietly sanded off.
Which is fine if you’re trying to build the most efficient possible personal brand. I’m not. I’m trying to build a place where the parts of my life that actually inform each other can sit on the same page. The agency work is sharper because I cook on weekends. The speaking is better because I’m in BNI every week. The Mexico writing is honest because I’m not trying to monetize it. They feed each other.
So that’s what this is. Speaking topics over here. Recipes here. Whatever I’m working on at the moment here. The blog has whatever I’m thinking about — Mexico, AI, networking, sometimes the stuff in between.
If you came here for one of those things, the others might surprise you. Poke around. There’s no funnel. There’s not even a popup. There’s just stuff I’ve made and things I’m doing, and if any of it lands, the newsletter is how you keep getting it.
Welcome.
]]>