The Sweep: how we decide whether to trench a business or walk.

A guy from a home services roll-up called me last Tuesday. Thirty minutes, no charge. Standard Sweep. Three minutes in, he asked if we could build him an "AI agent that closes deals for us." I asked what problem he was actually trying to solve. He said his team was not closing enough deals. I asked how his reps took notes today. He said he was not sure. I asked where his SOPs lived. He laughed.

I told him this was not a fit and passed. He was annoyed. That is fine.

I want to talk about the Sweep, because it is the single most important thirty minutes of any potential engagement, and it is not what most prospects expect it to be.

What the Sweep is

The Sweep is a thirty-minute call. It is free. It is not a discovery deck. It is not a sales pitch. There is no slideware. There is no follow-up quote generator running in the background. It is one call, on the calendar, where I ask a small number of specific questions and listen to the answers.

I take notes on paper. I do not share my screen. I do not demo anything. If I sound like I am on a sales call, I have already failed at the Sweep.

At the end of the thirty minutes, one of three things happens. I say we should keep talking. I say we should not. Or I refer the caller to someone else who is a better fit than I am. Roughly two out of five Sweep calls end in the second or third bucket. That number is on purpose.

What I am listening for

There are four signals that make me lean in during a Sweep. They are not fancy. Most consultants would not care about them. I care about them because they predict whether a Trenching engagement will actually work.

The first is that the operator can point at a specific problem. Not "we want to grow." Not "we want AI." A specific problem with a specific victim. "Our ops manager loses an hour every Monday night rebuilding a dashboard from three tools." That is a problem I can hear. "We need to level up as a company" is not.

The second is that the operator can point at where their SOPs live. This does not mean the SOPs are good. It means they exist somewhere, in a drawer or a Google Doc or a Notion or a stack of laminated pages in the shop. If SOPs exist in some form, we can build on them. If they do not exist at all, we are not building software, we are doing therapy, and there are better people for that job.

The third is a clear ownership structure. I want to know who signs the check, who owns the operational calendar, and who I will actually be sitting next to during the engagement. If the answer to any of those three is "well, it is complicated," I write that down and I ask again. Complicated ownership means complicated engagements. Complicated engagements fail.

The fourth is a real budget for a real fix. Not a specific dollar number. A posture. I need to hear the caller talk about the cost of the problem before I talk about the cost of the work. If the problem is costing the business real money and the caller can articulate that number roughly, we can have a conversation about spending real money to fix it. If the caller is looking for the cheapest possible outcome, we are already at the wrong table.

What makes me walk

Here is the list of things I hear on Sweep calls that make me pass. Not one of them is disqualifying by itself. All four together are.

The magic pill posture. The caller wants a tool, not a change. They believe there is a piece of software, somewhere, that will solve the thing without them having to do anything different. This person is going to be disappointed by every engagement they ever sign, including one with me. I will not be their next disappointment.

The "we want AI" opener. AI is a hammer, not a project. When the pitch starts with the technology and not the problem, I know we are going to spend the engagement trying to justify a decision that was made before I got there. That is not Trenching. That is a vendor demo with extra steps.

The refusal to let us inside. Trenching means I sit inside your business. I read the Slack. I ride along. I watch the CS rep's screen. If the caller flinches at any of that, if they want to hand me a scoped statement of work and receive a deliverable at the end without me being on-site, we are not doing Trenching. We are doing something else. There are firms that do that. I know a few of them. I will refer you.

Confusion about who is buying. If the person on the Sweep call cannot tell me who owns the outcome, they are not the buyer. That is fine. It just means the business is not ready. Come back with the buyer. I do not work through proxies on discovery.

Why we turn work away

Trench Logic is small on purpose. I take a limited number of engagements. Every engagement I sign, I am not signing something else. If I take a project that is going to fail, I have burned that slot, and I have burned the caller's money, and I have burned my team's time, and I have made the caller's next consultant harder to trust.

Turning away a bad fit is a form of respect. It respects the prospect's money and their calendar. It respects my team's time. It respects the client I will take instead, the one who actually has a shot at a successful engagement.

I would rather turn down a check today than take a check and fail on the deliverable in month three. I have done that math both ways and it is not close.

What a good Sweep sounds like

The Sweep calls that turn into engagements have a very specific texture. The caller does more of the talking than I do. They walk me through where a specific process is breaking. They can name the person on their team who feels the pain first. They can point at the drawer, the Google Doc, the Slack channel where the SOPs actually live. They know what the problem is costing them, roughly, and they say the number out loud without me pulling it out of them.

By minute twenty of that kind of call I am already sketching a rough shape in my notebook. Not a scope. A shape. What week one would look like. Who I would need access to. What the failure mode of the current setup is likely to be. That is not because I am a genius. It is because they gave me enough real information for the shape to draw itself.

How to know the Sweep is right for you

Not every business needs a Trenching engagement. Most do not. If your operation is running well, your data reconciles, and your team is not burning cycles on manual work that a tool should be doing, you do not need me. Save the thirty minutes. Buy your ops manager lunch instead.

The Sweep is right if you know something is off, you can point at where it hurts, and you are willing to let someone sit inside your business long enough to actually see it. If that is you, we should talk. If it is not, I will tell you that on the call, and I will save both of us the follow-up email.

Thirty minutes. Free. One call. That is the Sweep.