Reading a Client Before You Walk In

Most first meetings are spent catching up. Finding out what you could have known before you arrived. Not the confidential things. The shape of things. How confident the delivery team really is. Whether governance is a rhythm or a ritual. Where the Chair and the CEO would score the same question differently. The instruments already exist, they are free, and they were built to be handed around. This guide shows you two ways to point them at a client’s world before you walk in.

First, who this is for. Mode two assumes a client big enough to have a board and a public record. If your next client is a four-person business, mode one still works exactly as written and mode two has nothing to feed on.

Mode one runs on the core templates from unfsckd, which are open: no signup, no email, no login. Mode two uses the role-based intakes, which sit behind a free account. Neither mode needs a consultant or the client’s permission to run. You are not reporting on the client. You are preparing yourself.

Try the instrument at home first. Run the Delivery Confidence Snapshot across your own delivery team, each person scoring independently, and the numbers rarely line up. The PM scores delivery an 8. The tech lead scores it a 5. That gap is not a disagreement to smooth over in a stand-up. It is the read: the two people closest to the work are seeing different work. Pointing the same instrument at a client is that exact move, run from the outside in.

Mode one: point it at your own read

Pick one template. Start with the Delivery Confidence Snapshot. Fill it from what you already know: their site, their public record, what they said in the enquiry, what the industry says about businesses like theirs. Score the dimensions you can.

Then the inversion that makes the exercise pay. Every dimension you cannot score is a discovery question you did not know you had.

Your blanks are the agenda. Walk in with eight dimensions, five of them scored and three of them empty, and the three empties are your first meeting. That is a sharper opening than any generic discovery script, because it is built from the specific places your read runs out. You are not asking “tell me about your challenges.” You are asking the three questions you already know you cannot answer.

Do this for one template and you have a first meeting. Do it for the two or three that matter most to the engagement and you arrive with a map of exactly where your own understanding is thin.

Mode two: model the client’s seats

Mode one runs on what you know. Mode two stress-tests it.

Gather what is public or given: the website, annual reports, the org structure, the brief they sent you. Give it to an AI and ask it to hold the client’s seats. The role-based intakes give you the exact question set per seat: Chair, CEO, COO, CFO, Secretary, CISO. Have the model answer each intake as that person, drawn only from the evidence, and flag every answer as inferred.

Then put mode two beside mode one. Where the modelled answers and your direct read diverge is where your assumptions live.

That divergence list is the most valuable thing you carry into the room. It is not a set of conclusions. It is a set of places where two honest reads of the same evidence disagree, which means at least one of them is wrong, and you do not yet know which. Every line on it is a question worth asking in the first conversation, and a risk worth pricing into the engagement.

An invented example, and I am labelling it invented because this guide’s whole argument is about labelling things. You are reading a mid-market firm before a delivery engagement. Your own read scores their governance a confident 7: the board meets monthly, the papers look thorough. The model, holding the CFO’s seat from the same public evidence, scores it a 4 and flags that not one decision in the last three board packs was ever reversed. Those two answers cannot both be right. Maybe the governance is real and the CFO seat is underinformed. Maybe the papers are theatre and the monthly cadence is a ritual. You do not know which, and that is the point. That single divergence just became the first real question of your first meeting, and you found it from public documents, before anyone shook your hand.

In practice most divergences end duller than that. Take the same two scores with a boring ending: it turns out you were scoring the papers and the model was scoring the decisions, and neither of you was wrong about the thing you were actually looking at. That is still the first real question of the meeting; it just does not make as good a story.

The honesty rail

Modelled answers are hypotheses. They are labelled as inferred, verified in the room, and never presented back to a client as their actual position. The model tells you what to ask. It does not tell you what is true. Anyone who presents a modelled read as client intelligence has stopped diagnosing and started fabricating, and the client will know before you do.

This is not a style note. It is the line that keeps the method legitimate. The moment a modelled answer gets repeated as fact, you have manufactured false confidence, and the first thing the client says in the room will expose it. Held to the rail, the same modelled answer becomes the sharpest question you bring. Break the rail and it becomes the reason they stop trusting you. Same input, opposite outcome, decided entirely by whether you labelled it.

What this changes

Both modes end in the same place: a first conversation that starts at the real questions instead of taking an hour to reach them.

You walk in with a scored read, a blanks list, and a divergence list. The client experiences someone who did the work before asking for theirs. Nothing you present is dressed up as certainty, and everything you ask is precise.

The diagnostics are free. Mode one is about an hour per template; mode two takes as long as their public record takes to read, which for a listed client is most of a day and for a private one might be twenty minutes because there is nothing to read.

That is the whole move. The difference in the room is the difference between discovery and knowing where to look.

The core templates are free and ungated; the role-based intakes need a free account: get them at unfsckd.com. Want the read done with you before a real engagement? Book a Clarity Call.

Questions