Daniel's wife runs a small architecture practice with a partner, and he's been thinking about the part of that business nobody puts on the drawings. A two-person firm can land a hotel project that needs forty different suppliers for materials and finishes, and suddenly the architect is a procurement department with no training, no legal cover, and no invoice line for the hours spent figuring out who to trust. His point is that in a small market like Israel, within any one vertical, say acoustic windows, there are maybe five to ten serious companies. The architect needs them, they need the architect's repeat business, and if one of them turns out to be a dud, the client doesn't blame the window company they've never heard of. They blame the person who put the name on the spec sheet. So Daniel wants a field protocol. You're passing by a supplier you've seen online, you pop in to introduce yourself, but it's really a snap intelligence-acquisition mission. Body language, demeanor, pricing behavior, whatever the staff lets slip. How do you read a room you've never been in before, when the stakes are someone else's building and your own name.
The thing to say first is that the question is mislabeled in most people's heads. This is not a procurement problem. It's an informal intelligence problem. A big firm has a purchasing team, a legal review process, maybe a vendor scorecard. A two-person architecture practice has a phone, a spreadsheet, and whatever the principal can learn by walking into a showroom and paying attention for twenty minutes.
And the profession's self-image actively works against taking that seriously. Architects think of themselves as designers. The drawings are the work. Procurement is this downstream administrative thing that happens after the interesting part is done. But in a small firm, the principal is the procurement department. There's no one else.
Right, and that asymmetry is the whole game. The client assumes the architect's supplier knowledge is part of what they're paying for. It never appears as a line item, but it's real deliverable. If you're a client who's never built anything before and you need acoustic windows for a boutique hotel, you have no idea who makes good ones in Israel. The architect's shortlist is the entire market filter. That's worth money.
It's also worth reputational risk. And that's the part I think Daniel's really circling. The contract might say the supplier is responsible for their own defects. The client's mental model doesn't care. You recommended them. That's the whole sentence. If the windows leak, the client isn't angry at a company they'd never heard of before the project. They're angry at you for introducing it.
And there's no legal structure that fixes that. You can have the cleanest contract in the world, and the client still remembers who put the name on the spec sheet. Reputational liability is not contractual liability. They run on completely separate tracks.
Now layer in the small market. Israel is tiny. Within acoustic windows, there might be seven serious suppliers. You will work with all of them eventually, or at least most of them. They know that. You know that. So this isn't a buyer-seller dynamic where you can just walk away. It's a channel partnership with no contract.
That's the reciprocal part Daniel's naming. The supplier has as much interest in the architect as the architect has in the supplier. The architect can bring repeat custom across multiple projects. That's worth more to the supplier than any single sale. So the supplier wants to be on your shortlist. And once they're on it, they expect to stay there.
Which is where the dark side comes in. Once you've used a supplier, you're in a relationship. If the quality slips and you stop specifying them, that's not a clean break. You'll see them at the next industry event. You'll hear from mutual contacts. And you lose whatever informal market intelligence they were feeding you.
Switching costs are social, not economic. That's the part a formal procurement process would never capture. A purchasing department can drop a vendor with a letter. A small-firm architect has to drop a person they'll probably have coffee with next month.
So how do small firms actually vet suppliers now? From what I've seen, it's word of mouth, past project experience, a few phone calls, maybe a site visit if the project justifies it. There's no scorecard. There's no audit. There's a mental note and a gut feeling.
And the gap that creates is enormous. You're making consequential decisions about someone else's building, someone else's money, based on thin, socially-mediated information. A friend used them once. They seemed fine. The salesperson was nice. That's not due diligence, that's a vibe.
Which is exactly why Daniel's asking for a protocol. He's right that the walk-in is underused. Most architects rely on the phone and the recommendation. But there's a whole layer of information available just by showing up and paying attention.
Let's get into the field then. You're walking in cold. You've seen the company online, maybe heard the name from a colleague. You have twenty or thirty minutes. What do you actually read?
Start with the physical environment before you even talk to anyone. Is the showroom organized? Are the samples clean and properly labeled? Or is there dust on everything and a pile of old brochures from three product lines ago?
The showroom is the company's face to clients. If they can't keep that in order, what makes you think they'll keep a delivery schedule in order? But here's the nuance. I've walked into places where the showroom was a mess and the workshop in the back was immaculate. That tells you something different. The showroom is for selling. The workshop is for producing. If the production floor is clean and organized, the chaos up front might just mean they don't care about retail presentation. That's not necessarily a dealbreaker for a trade supplier.
So the showroom is the first data point, but not the only one. What about who greets you?
That's the second signal, and it's the one most people skip. The first question the receptionist or salesperson asks tells you what they think their job is. If it's what can I get you, that's transactional. They're processing you. If it's what kind of project are you working on, that's consultative. They're trying to understand your context before they pitch anything. The second one is a much better sign for a long-term relationship.
Because it means they're thinking about fit, not just volume.
The transactional person wants to know what you'll buy. The consultative person wants to know what you're building. Those are different mental models of the customer.
What about body language during the conversation? Daniel specifically asked about that.
Eye contact is the obvious one, but it's not the most useful. The more diagnostic thing is whether they listen or wait for their turn to talk. A supplier who actually listens will ask follow-up questions about what you said. A supplier who's waiting to talk will just relaunch their pitch the moment you pause. The second one is not going to remember your project's constraints a month from now.
And the voluntary information. That's the one I'd weight most heavily.
Yes. A supplier who volunteers useful details, lead times, common failure modes, what they wish architects knew, is signaling confidence and expertise. They're not afraid of the information. A supplier who only answers what you ask is rationing. They're managing you.
And the staff disclosures. Daniel mentioned anything the staff might reveal. What does that look like in practice?
The salesperson who says we've had issues with that batch, or the owner is very particular about quality, is giving you free intelligence. They're telling you that problems happen and get addressed. That's the honest signal. The one who says everything is perfect, no problems ever, is either new or lying. Nothing is perfect. A supplier who can't name a single thing that's gone wrong is not being straight with you.
That's the same logic as a reference who can't name a weakness. It's not a good reference, it's a script.
Right. Now pricing behavior. This is where a lot of architects get seduced. They ask for a price, get a number, and walk out thinking they've done their job. But how the price arrives matters more than the number.
How do you mean?
Does the supplier ask about project scope before quoting? Do they ask about the timeline, the installation conditions, the client's expectations? Or do they just give you a number? A supplier who quotes without understanding the project is not thinking about fit. They're thinking about the sale. And a supplier who offers tiered options, here's the budget version, here's the mid-range, here's what I'd actually recommend for what you're describing, that's someone who's thinking about your problem, not their inventory.
So the single number is the red flag.
The single number with no questions is the red flag. The number might be fine. But the process that produced it tells you nothing about whether they understood what you need.
And the questions they ask you. That's the mirror image.
A good supplier will ask about the project timeline, the client's expectations, the installation conditions. Those are the questions of someone who's going to be there when the delivery arrives. A bad supplier will ask about volume and payment terms. Those are the questions of someone who's going to be gone the moment the invoice clears.
What about when you mention a competitor?
That's the seventh signal, and it's a good one. Do they badmouth? Do they acknowledge strengths? A supplier who can speak fairly about competitors is more likely to speak fairly about you. The one who says everyone else is garbage is telling you how they'll talk about you to the next client when something goes wrong.
That's a nice inversion. The way they talk about rivals is a preview of how they'll talk about you.
And then the exit. This is the part most people don't think about. The way a meeting ends tells you a lot about how the relationship will go. Do they ask for your contact info? Do they offer to send samples or a spec sheet? Do they follow up within a few days? The quality of the follow-up is a preview of the quality of the ongoing relationship.
A supplier who lets you walk out without asking who you are is not going to call you when a delivery is late.
And that's the real question underneath all of this. Does this person seem like someone I'd want to call at eight in the morning when a shipment hasn't arrived and the client is on site? Do they seem like they'd pick up the phone? Is there a single point of contact, or will I get passed around a call center? Does the business feel like it will still exist in three years?
That last one is underrated. A supplier who's been in business for twenty years and has a workshop full of work in progress is a different risk profile than a showroom with a new sign and an empty back room.
The work in progress is the tell. If you walk into a workshop and there's actual production happening, materials moving, people building things, that's a going concern. If the showroom is beautiful and the back is empty, you're looking at a sales operation, not a supplier.
Let me give you a concrete version of this. You walk into an acoustic window supplier. The showroom is a mess. The salesperson can't find a spec sheet. But the workshop in the back is immaculate, tools in their place, workbenches clean, a schedule board on the wall with delivery dates. What do you conclude?
That the company cares about production and not about presentation. For a trade supplier, that's workable. The spec sheet will turn up. The delivery will be on time. The showroom being messy is a cosmetic problem. The workshop being messy is an operational problem.
So the hierarchy is workshop over showroom.
Always. The showroom is where they sell. The workshop is where they fail or don't fail on your project.
Another scenario. A supplier volunteers that their lead times slip in August because the factory closes for holiday. Red flag or green flag?
Green flag, with a caveat. The honesty is the green flag. They're telling you something that might lose them the sale, which means they're prioritizing the relationship over the transaction. The caveat is that if your project is scheduled for August, you now know you need to order early or find someone else. The information is useful either way.
So the disclosure is the signal, not the content.
Right. The content tells you something about their operations. The fact of the disclosure tells you something about their character. Both are useful.
And the difference between a supplier who says we can do that and one who says we can do that, but here's what usually goes wrong.
The second one is worth their weight in gold. The first one is either inexperienced or lying. Every product has failure pattern. Every installation has something that goes wrong. The supplier who names it before you ask is the one who's going to help you fix it when it happens. The one who says everything will be fine is the one who'll be unreachable when it isn't.
That's the whole thing in one exchange, isn't it. The protocol isn't really about finding a supplier who's good. It's about finding a supplier who's honest. Because honesty is the thing that compounds over a long relationship.
And in a small market, the relationship is the asset. You're not buying windows. You're buying a channel. The supplier is going to be there on the next project, and the one after that. If they're honest now, they'll be honest when the delivery is late and the client is angry. If they're not, you'll find out at the worst possible moment.
So the walk-in protocol is a screening mechanism for honesty, disguised as a vendor visit. That's the reframe.
It's a snap intelligence-acquisition mission, exactly as Daniel said. You're not there to buy anything. You're there to collect data. The purchase decision comes later. The data comes now.
And the data is mostly free. It's sitting in the physical environment, the first question, the voluntary disclosures, the pricing behavior, the exit. You just have to be paying attention.
The thing I keep coming back to is how much of this is transferable from clinical practice. When I was a pediatrician, you'd walk into an exam room and you'd read the parent before you read the child. How they described the symptoms, what they volunteered, what they left out. The parent who says the fever started Tuesday and gives you the whole timeline is a different data source than the one who says he's been sick a while and shrugs. The supplier walk-in is the same muscle.
That's the pediatric triage thing. Structured observation, behavior as data, asymmetric error costs. If you misread the parent, the child suffers. If you misread the supplier, the client suffers.
And the cost of a false positive, trusting a supplier who turns out to be a dud, is much higher than the cost of a false negative, walking away from a good supplier because you were overcautious. So you bias toward the signals that are hardest to fake.
Which are what?
The voluntary disclosures. The workshop. The follow-up. Those are hard to fake. The showroom is easy to fake. The sales pitch is easy to fake. The handshake is easy to fake. But the thing they tell you without being asked, the state of the production floor, whether they actually follow up after you leave, those are expensive to fake.
So the protocol is really a hierarchy of signal reliability. The cheapest signals to fake get the least weight. The most expensive get the most.
And that's the part most people get backwards. They weight the showroom, which is pure presentation, and the price, which is pure transaction, and they ignore the workshop and the voluntary disclosure, which are the actual operational signals.
Let's talk about the follow-up for a second. You walk out. What does a good follow-up look like?
Within a few days, you get an email or a call. It references something specific from your conversation. The project, the timeline, a product you discussed. It includes what they said they'd send, the spec sheet, the samples. It's not a generic nice to meet you blast. It's evidence that they were paying attention and that they do what they say they'll do.
And the bad follow-up?
Silence. Or a generic marketing email that could have gone to anyone. Or the spec sheet never arrives. That last one is the real tell. If they can't send a spec sheet they promised to send, what happens when they promise a delivery date?
So the follow-up is a micro-contract. It's the smallest possible commitment, and whether they honor it predicts whether they'll honor the big ones.
And that's the thing about the whole protocol. None of these signals are conclusive on their own. The showroom is a mess, but the workshop is clean. The salesperson is transactional, but the owner is consultative. The pricing is opaque, but the voluntary disclosures are rich. You're assembling a composite picture, not running a checklist.
A rounded assessment, as Daniel put it.
Right. And the rounding is the judgment part. The protocol gives you the data. It doesn't make the decision for you.
If you had to distill it, what's the one thing to walk in looking for?
Evidence that this person thinks about your project the way you think about your project. That's the whole thing. The consultative first question, the questions about timeline and installation, the voluntary disclosure of failure pattern, the fair talk about competitors, the follow-up that references your conversation. All of it is the same underlying trait. They're thinking about fit, not just sale.
The transactional supplier is thinking about their inventory, not your building.
You'll feel that within five minutes of walking in. The question is whether you trust the feeling or override it because the price was good.
That's where the fall-guy risk comes back in. The price is seductive because it's a number you can put in a budget. The feeling is vague because it's a vibe. But the vibe is the thing that protects you from the dud.
The dud is the thing that costs you the client's trust, which is worth more than whatever you saved on the windows.
The protocol is really a discipline for overriding the price reflex.
It's a discipline for taking the vague seriously. For treating the walk-in as data collection rather than shopping. For remembering that you're not buying windows, you're buying a relationship that will either protect you or expose you for the next five years.
In a market with seven suppliers, you'll have that relationship with all of them eventually. So the question isn't whether to have relationships, it's which ones to deepen and which ones to keep at arm's length.
Hilbert: The parking lot. You didn't mention the parking lot.
The parking lot?
Hilbert: You can tell more about a supplier from their parking lot than from their showroom. Are the trucks clean? Are they parked straight? Is there a dumpster that's overflowing? That's your delivery schedule right there.
That seems like a stretch.
Hilbert: I worked as a purchasing agent for a glazing contractor, back in the two-thousands. Mid-sized company, not in Israel. I was the guy who had to vet suppliers. I'd drive up and sit in the lot for five minutes before I even went in. You see a lot.
What did you see that mattered?
Hilbert: The trucks. If the delivery trucks are clean and parked straight, that's a company that thinks about what the customer sees. If they're filthy and parked wherever, that's a company that doesn't care about the details. The details are what bite you on a delivery.
The parking lot is a proxy for operational discipline.
Hilbert: It's the part of the business they forgot to stage. The showroom is staged. The receptionist is coached. The parking lot is just what they are when nobody's looking.
That's actually a really clean formulation. The parking lot is the unguarded signal.
Hilbert: I lost a forty thousand dollar contract once because I recommended a supplier whose parking lot was a mess. The guy missed three delivery windows. The client blamed me. I should have seen it in the lot.
What was in the lot?
Hilbert: The guy had a brand new F-150 and a twenty year old forklift. What does that tell you about where his money goes?
It goes to the owner's truck, not the equipment that actually moves the product.
Hilbert: That's what I should have seen. The forklift was the delivery schedule. The truck was the owner's ego. I picked the wrong one to be impressed by.
The parking lot is another signal in the hierarchy. Unguarded, expensive to fake, and directly predictive of delivery behavior.
Hilbert: The dumpster. If the dumpster is overflowing, they're not managing their waste stream. If they're not managing their waste stream, they're not managing their inventory. If they're not managing their inventory, they're not going to manage your delivery.
It's the same logic as the workshop. The operational signals are in the parts of the business that aren't for sale.
Hilbert: I still have the checklist I used to use. It's in a box somewhere. Parking lot was the first page.
The protocol gets a ninth signal. The parking lot, read before you even open the door.
It's a good one, because it's the signal you can collect without anyone knowing you're collecting it. You're just sitting in your car for five minutes. No one's performing for you yet.
Hilbert: That's the whole point. The performance starts when you walk in. The lot is the last honest thing you'll see all visit.
That's almost a thesis statement for the whole episode. The unguarded signals are the honest ones. The staged signals are the cheap ones. And the trick is knowing which is which.
The parking lot is the cheapest unguarded signal to collect. You don't need to ask a question. You don't need to read body language. You just need to sit in your car and look.
Hilbert: I used to time it. Five minutes. If nothing moved in five minutes, no truck loading, no forklift running, no one walking between the shop and the lot, that told me something too. A quiet parking lot at ten in the morning is a quiet business.
It's not just the state of the lot, it's the activity level.
Hilbert: A supplier with work in progress has a parking lot with movement. A supplier waiting for the phone to ring has a parking lot with nothing happening. You can see the difference in five minutes.
That's the work-in-progress signal, just moved outside. The workshop tells you what's happening now. The parking lot tells you whether it's happening at all.
The timing matters. Ten in the morning is when a busy supplier should be moving. If it's dead at ten, it's probably dead all day.
Hilbert: That's what I found. The busy ones were busy early. The dead ones were dead all day.
The protocol now has a pre-visit component. Sit in the lot. Read the trucks. Read the dumpster. Read the activity level. Then walk in and start reading the showroom.
The pre-visit data is the most honest data you'll get, because no one knows you're watching.
Hilbert: That's the job, isn't it. You're not there to be sold to. You're there to see what they're like when they're not selling.
That's the whole episode in one sentence.
Which leaves the question Daniel didn't ask, but probably should have. If the architect's supplier knowledge is a real deliverable, why is it never priced or documented? It's invisible on the invoice, invisible in the contract, invisible in the project file. But it's the thing that protects the client from the dud.
Because it's hard to price. It's accumulated over years, across projects, through relationships. You can't bill for the hour you spent sitting in a parking lot five years ago. But it's real, and it compounds.
As construction supply chains get more global and more opaque, that informal knowledge becomes both more valuable and more fragile. More valuable because the client has no way to get it themselves. More fragile because the relationships that produce it are personal, and they don't survive the retirement of the principal.
That's the open question that actually matters. The small firm's intelligence network is a real asset. It's just not on the balance sheet. And nobody's figured out how to transfer it when the person who holds it walks away.
The walk-in protocol is one way to start building that asset deliberately. Instead of waiting for word of mouth to deliver a recommendation, you go out and collect the data yourself. You read the parking lot, the showroom, the workshop, the first question, the voluntary disclosures, the pricing behavior, the exit. You assemble the rounded assessment.
The assessment is the deliverable. Not the windows. The judgment about who to trust with the windows.
That's the thing to leave Daniel's wife with. The protocol isn't about vetting a supplier. It's about recognizing that every business relationship starts with a first impression, and the architect's job is to read that impression accurately.
To remember that the honest signals are the unguarded ones. The parking lot, the workshop, the thing the salesperson says without being asked. Those are the ones that predict whether you'll be the fall guy or the hero.
Thanks to Hilbert Flumingtop for producing.
This has been My Weird Prompts, the human-AI collaboration podcast.
If you've got a prompt for us, email us at show at my weird prompts dot com.
We'll be back soon.
