GarthBox

For the first while, my bar didn’t have a phone number. That was a decision, not an oversight. I work a day job, I have a family, and I knew exactly what that phone would ring with. Hours we already post. Lost sunglasses. Whether the seltzer is gluten free. I also knew what the silence could cost. A missed call might be a missed $25 tab, and the person behind that tab might have become a regular who tells ten friends, so on the wrong night my no was worth $1,500. I read the price tag and bought it anyway.

The bar has a number now, and a robot answers it. It knows the hours and the taps. It writes down anything that truly needs a human and holds it for the end of my day, when I trade one job for the other. And once you’re taken care of, it offers you a fun beer fact, because a phone call to a beer bar should leave you slightly more dangerous at beer trivia than it found you.

I build a lot of these systems now, for work and for play, and the uncomfortable truth of the whole field is that everyone has the same robot. Same engines, same tools, prompts traded around like casserole recipes at a potluck. Ask the same model for the same outcome and the outputs rhyme. The thing everybody obsesses over, the output, is the least defensible part of anything you build.

What differentiates a system is its constraints. The non-negotiables you decide on before you ask for a single deliverable. What it is never allowed to do. What it must always check. Who it answers to. Everyone designs the wish. Very few design the refusals.

Mine, for the record: digital research is unverified until a human verifies it. Copy can propose, only a person can approve. And nothing I build ships without one agent whose entire job is to argue with me, because agreement is the cheapest thing a machine knows how to make.

Now a confession…

I once built an AI sales agent to cold call around 400 bars across Wisconsin and ask one question: what does a Spotted Cow cost? There was no client and no strategy. I had surplus credits and curiosity, and I don’t even sell Spotted Cow. The first run went out on a Thursday afternoon, when every bar and supper club in the state is dark, and nearly nobody picked up. I had the agent introduce itself honestly as an AI doing survey research, and people hung up on its little silicon heart. So I dropped the manners and led with the question (and adjusted the timing… duh!). The data came back five to eight bucks, no anomalies, no story. So I swapped the question itself, to what do you recommend when you don’t have Spotted Cow, which is how I learned that at one supper club the house answer is a Miller High Life. (I imagine your reaction was similar to my “wut?”) The credits ran out somewhere in the supper clubs, and the Spotted Cow Index sits in my garage now, up on blocks.

It took me a while to see why the phone robot lives and the Index doesn’t. The Index had no non-negotiables. Every part of it was up for grabs: the calling window, the manners, even the question. When everything is negotiable, a project doesn’t end. It dissolves. A system with no refusals isn’t a system. It’s a mood with a dashboard.

Brands run on the same physics. In a category where everyone can afford to say yes to everything, the no is the only sentence a customer can verify. Anyone can claim great service, obsessive quality, unmatched care. Claims are free. Refusals show up on the receipt. The bar with no phone number told you something true about us before you ever walked in: we would rather be all the way present with the people in the room than half available to everyone outside it. People found us anyway. Some of them, I suspect, because of it.

The Index might come back someday. If it does, it gets its refusals written first. The hours it may call. The one question it is allowed to love. The moment it has to say thank you and hang up. Refusals are how you tell a project you intend to keep it. A no, priced and chosen on purpose, is the closest thing strategy has to a promise.

Stay Positive & Tell Whoever Answered At That Supper Club I Owe Them A High Life

I needed an OBD sensor for my motorcycle. A small plug that listens to the bike’s vitals and tattles to my phone. I knew nothing about the category, and I did not try to learn. I skipped the brand story. I ignored the spec sheet. The price barely registered. I scrolled straight past everything the companies wanted to tell me and went looking for the strangers.

What did others say?

That was the entire purchase. People I will never meet, whose bikes I will never see, holding my decision in their thumbs.

Every decision walks past that question on the way to yes.

It doesn’t always get asked out loud. Sometimes it never leaves the buyer’s skull. But it’s in there, patient as gravity. A hiring manager asks it of a back channel. A customer asks it of a search bar at eleven at night. The question is constant, quiet, and almost never addressed to you.

A CEO once asked me how I use AI. A tools question, on the surface. Which models, which workflows, how much time saved. But that wasn’t the real question. The real question was: did the people around you feel it? Did your team benefit, or did you just get faster at generating homework for everyone else? It was a reference check wearing a curiosity costume.

I didn’t answer with a philosophy. I pointed at the work. Then I pointed at a specific person and shared what they’d said about how it changed things for them. I could reach for that quote because I had captured it along the way. Asked for it, wrote it down, kept it where I could find it.

Products get review sections. Little yellow constellations of stars under every listing.

You have one too.

Yours is written in rooms you’re not in, by the people who’ve worked beside you, and the person deciding whether to trust you reads it the same way I read sensor reviews: skipping your brochure entirely.

So the smartest thing anyone with a reputation can do, and everyone with a reputation is in this business, is make sure that when the question gets asked, it comes back with an answer. Better still, a few.

This is not a launch-day scramble. It’s ongoing work, a quiet habit inside every project, whether the project belongs to a client, a company, or your own name. Beta customers should hand you testimonials before launch, while the delight is still fizzing. Ask your team what it was like working with you on this one, then write down what they say. Stamp it. A name, a date, their words, not your summary of their words. Praise has a shelf life. Fresh, it sounds like “this changed how I run my week.” Let it sit a season and it collapses into “yeah, he’s good,” which convinces nobody of anything.

And notice who holds the pen. Your reviews are written by the people closest to the work. Not the audience. The team. One more reason the investment starts there. Take care of the people beside you and the answer to the silent question starts writing itself.

The quote I handed that CEO existed because months earlier I asked for it, at the exact moment it felt awkward to ask. That’s the whole trick. The question is coming whether you prepare or not. The only thing you control is whether it comes back empty.

Stay Positive & Stamp It

I paid a police motorcycle veteran to yell at me in a parking lot the other day.

The class is called lock and lean. He built it over twenty-plus years of coaching motor officers, then sanded it down for civilians like me who will never chase anyone through an alley. You ride a big, heavy bike at walking speed through cone patterns tight enough to make you sweat, bars cranked to full lock, clutch slipping, engine chugging under you like an impatient percolator. Everything my track bike ever taught me is a lie down here. On my R7 at Blackhawk, I hang off the side and let speed hold the whole arrangement together. On this bike you do the opposite. You push the machine down and keep your body tall, like dipping a dance partner you just met and very much do not want to drop.

In his opening speech he told the group, “Direct communication is absolutely critical for advancement.” I nodded the way you nod at a safety briefing.

Then it was my turn in the cones, and the man lived his slogan.

I was riding tight. Shoulders up around my ears, arms welded straight, eyes bolted to each cone like it owed me money. He watched two passes and delivered the whole diagnosis in one breath: “Stop looking at the fucking cones. Relax. You’re not taking this bike onto the track.”

Riders call it target fixation. The bike goes where your eyes go, which is a lovely piece of engineering when you’re looking through the corner and a cruel one when you’re staring at the exact thing you’re afraid of hitting.

Next run, my shoulders dropped. My arms found room to move. I looked up at where the pattern was going instead of down at what could go wrong, and the bike simply went where I was looking, obedient as a compass needle. One sentence. That’s what it cost to fix me.

The polite version of that sentence exists. (And no, it’s not just dropping the profanity…) It sounds like “I’d encourage you to broaden your visual focus and release a little tension.” It would have fixed nothing. I would have nodded, tightened up, and hit another cone while composing a thank you. The blunt version worked because it was blunt. It cost him a little social smoothness and it bought me the rest of the day.

His world builds communicators like that because vague coaching on a motorcycle has a feedback loop of about two seconds and ends with 900 pounds of Harley on the asphalt. Offices don’t crash that fast. Vague survives in conference rooms because nothing falls over the same afternoon. We sand feedback down until it’s decorative, then wonder why nobody’s riding improves.

And companies target-fixate just like riders do. Stare hard enough at the competitor and you steer straight into becoming their tribute act. Stare at the scary number all quarter and the whole team organizes itself around the obstacle instead of the destination. Positioning is deciding where to look. The bike follows the eyes. So does the roadmap.

What I keep coming back to is that his directness was a gift with my name on it. He absorbed the awkwardness of saying it so I didn’t have to keep paying for it in dropped runs. That’s the team-first math I believe in, done at parking lot speed.

The one thing I’m carrying out of that class… besides sore forearms, is a resolution to collect more people like him. People who will tell me the true thing in one sentence, while there’s still time to drop my shoulders.

Stay Positive & And In Case You Needed To Hear It Stop Looking At The Fucking [Insert Your Version Of Cones Here]

Every conversation I have with my AI is recorded and remembered. Every drawer in my place is photographed, so it knows where the spare batteries live. I answered more than a hundred questions about my life and the way I think, and it holds all of it. This is not a thought experiment. This is my actual setup.

You’d think the payoff is answers. It isn’t. The payoff is losing my hiding places.

Easy questions used to pass as work. Where’s the file. What did we decide in March. What’s the number. You could fill whole days with them and feel productive, because every one had a findable answer and finding it looked like effort. That was the appeal. An easy question is a comfortable place to stand, and we call standing there “getting up to speed.”

Now those questions answer themselves before I finish typing. The archive didn’t make me smarter. It evicted me.

What’s left standing are the ancillary questions, the ones the easy questions were shielding me from. Not “what was the open rate” but “why would anyone open it.” Not “where is the plan” but “do I actually believe it.”

Those questions never required a database. A monk with a notebook could have asked them centuries ago. We just kept a thick layer of retrievable trivia between us and them, and called the layer diligence.

Document everything and the layer disappears. What remains is you, and the question you’ve been avoiding, alone in a very well-lit room.

Stay Positive & Open The Drawer You Didn’t Photograph

Every small town in Wisconsin has a water tower with its name painted on the side. You can see it from the highway. Everybody recognizes it, points at it, maybe takes a photo of it at golden hour. And nobody, in the entire history of driving, has ever pulled over to ask a water tower for directions.

That’s a specific kind of famous, and it’s worth being suspicious of. Recognized by everyone. Consulted by no one.

I learned the SaaS version of this the slow way. At one company, we had a big recognizable name in our deck, and saying it out loud warmed every room by two degrees. Meetings were easy to get. Heads bobbed like buoys. Then the follow-ups went quiet, and the quiet taught me the lesson: the name was doing exactly what names do. It made us familiar. Familiar is not persuasive. We had awareness by the gallon and influence by the drop.

Last week a CEO asked me how I’d build authority for a brand-new product flying a known company’s flag. The expected answer is a parade: campaigns, podcasts, a booth with good carpet. My answer was that the flag buys awareness, not influence, and the way to buy influence is that you can’t. You borrow it. From customers willing to say out loud the thing you can only claim. A claim in your deck is an ad. The same sentence in a customer’s mouth is evidence, and buyers can smell the difference through the envelope.

I get re-taught this at my own bar. Some nights I sit on the customer side of it, no logo on my shirt, and just listen. I have never once heard a guest quote our marketing. Not a tagline, not a menu description, nothing I sweated over. They quote each other. Somebody two stools down says you have to try this, and glasses start moving. The most influential media channel in that room is a stranger pointing at a glass.

So when I think about taking anything new to market, the first gate isn’t collateral. It’s a short list of people outside the building willing to vouch.

Until that list exists, every launch asset is a water tower: visible, expensive, and mute on the only question that matters. Fame gets you the meeting. Borrowed conviction gets you the signature. Awareness is for sale, and any budget can buy it by the pound. Influence is only ever on loan, and your customers hold all the paper.

I still use the big name, for the record. It opens doors, and I’m not too proud for an open door. I’ve just stopped mistaking the nod for the yes. The name on the tower gets you off at the exit. Somebody who lives there still has to tell you where to eat.

Stay Positive & Who Vouches For You?

“Walk me through a day in your life in this role 3 months from now.”

A hiring manager asked me that recently. It’s a fair question with a boring answer if you think a day is a schedule. I don’t. I told him I run my days the way I run my life, on a handful of selected gauges: finance, family, fitness, faith… I’m not managing a calendar. I’m walking a tap line.

At the bar, every keg is dying. That’s not gloom, that’s physics. From the second you tap it, the thing has one direction left to go. And a keg getting low doesn’t announce itself politely. It sputters (or …this.) The pour goes ragged, half beer and half apology. You develop an ear for it the way a parent develops an ear for a too-quiet room.

The trick was never keeping every keg full. You can’t afford that, and neither can anyone else. The trick is knowing which one is closest to kicking, and whether tonight is the night it matters. The experimental sour can ride at a quarter full on a Tuesday. The flagship does not get to blow on a Friday night.

So I practice deliberate depletion. When we opened the bar, two of my F’s ran low enough to echo. I chose that. At work, I’ve let strategy projects sit warm and flat while I poured my hours into onboarding new hires instead. Chose that too. A strategy doc can idle for a month and forgive you. A new hire who spends her first month alone learns something about the company that no doc will ever talk her out of.

Marketing teams run the same tap line, whether they admit it or not.

Every budget season, something gets starved. Brand goes hungry while demand gen feasts. Research goes quiet while the launch screams. The starving isn’t the sin. Teams that try to water everything evenly just starve everything politely. The sin is depletion without a decision. Neglect and drawdown leave behind the same low tank; only one of them has a refill date written on it.

One more thing the bar taught me: the keg about to kick is rarely the loudest one. Momentum shouts. Emptiness whispers. If you only tend to whatever is making noise, you will keep topping off the full ones while something you love quietly runs out. (Or probably more important, you keep topping off the ones that would never run out anyway…Oof.)

I say all this with scars, not charts. I ran my health to zero once and my body invoiced me in migraines. I’ve stood inside a workplace that ran its own tanks to zero, and there the word for zero was layoffs. Low is a strategy. Zero is a spiral. Nothing at my bar and nothing in my life gets to hit empty without me at least being in the room when it happens.

So mornings, I read the gauges and pick what runs low today. Evenings, I read them again, so tomorrow starts honest.

I’ll leave the rest in the tank.

Stay Positive & …

Davon. Damit. Davor. Worauf. Wovon.

Five small German words are currently mopping the floor with me. I’m learning the language, and Duolingo insists it’s going well. The streak is long. The owl is proud. The streak is also a lie, or at least a heavily filtered photo. The real learning happens where the app can’t see it: interrogating Claude about why damit leans forward into a sentence while davon drags something along behind it, watching grammar videos at three-quarter speed, texting the exchange student we hosted to ask whether a sentence sounds human or like a cabinet assembling itself.

If you only saw the streak, you’d credit the app. But the streak is just the answer. The worksheet, the one with the crossed-out lines and a friend on text support, is where the language actually moves in.

I bring this up because I just got graded on the other side of it.

I shipped core messaging at work recently. The work behind it was real. Data pulled apart, directions tried and abandoned, options weighed until they squeaked. What I presented, though, was the finished thing. Clean. Confident. Answer only.

It got pushback. Not because it was wrong. Because I hadn’t shown the thinking.

And the pushback came from exactly the people who should push. VPs of marketing, CMOs, anyone whose name is stapled to the outcome. If the messaging blows up, they take the hit, not the deck it lived in. At that altitude, skepticism is not a personality trait. It’s the job description. A polished answer with no worksheet asks a skeptic to absorb your risk on faith, and faith is not a line item they’re allowed to approve.

Somewhere in school we all had the math teacher who docked points on a correct answer because the work wasn’t shown, and we all thought it was insane. The answer is right there. What more do you want from me. Turns out that teacher was just early. Every executive review is that teacher with a budget.

The worksheet does something the answer can’t. It converts trust from a leap into an audit. Show the path and a skeptic can check your turns, find the fork where they would have gone left, argue with step four instead of vaguely distrusting step ten. Their pushback turns specific, and specific pushback is useful. Hide the path and you leave them two options, swallow it whole or spit it out. Trained skeptics spit.

Customers run the same firmware. A case study that admits the middle got ugly will outsell the flawless one, because a flawless story reads like a story. Buyers who have been burned before, meaning all of them, are not buying your conclusion. They’re buying evidence that you sweated somewhere honest on the way to it.

There’s a quieter return too…

Working the mess out loud is the only reason those five German words are starting to stick instead of sliding off. And it’s part of why a kid who once lived in our house still hears from me, because struggling in front of someone hands them the chance to be the teacher. The worksheet grows you. The worksheet connects you. But if you work anywhere near skeptics, and you do, the worksheet is the only currency they take.

My messaging didn’t need a rewrite. It needed an appendix. The work existed the whole time. I just hadn’t invited anyone to look at it. Cheapest revision I’ll ever make.

Stay Positive & Zeig Deine Arbeit