Capital Refinery
Scenario · 7 minute read

Before Monday's IC — what the diagnostic does for a Friday-afternoon CIM.

The CIM lands at 4pm on Friday. The committee meets at 8am Monday. Here is how one partner's weekend changes when the diagnostic runs first.

A partner's weekend has a shape. The CIM lands Friday afternoon. The associate is told to “build a model and a memo by Sunday night.” The partner reads the memo Sunday evening for the first time. Monday at 8am, the committee asks the questions. Whatever the associate missed becomes whatever the partner has to defend on the fly. This page is about a different shape.

Friday 4:12pm — the CIM lands.

Eighty-page CIM. Financial model in xlsx. A QoE excerpt — partial, not yet final. Three years of audited financials. A summary credit agreement, redacted in two places. The banker wants a meeting Tuesday. Your IC is Monday at eight.

The old workflow: forward the bundle to the associate. “See what you can do by Sunday night.” The associate spends Saturday extracting numbers, building the deal model from scratch, drafting a memo against the firm template. Sunday they do another pass. Quality varies by who picked it up.

The new workflow: the partner drops the same files into the diagnostic intake at 4:30pm. Tags the lane (corporate PE). Closes the laptop. Drives home.

Friday 11pm — the pack comes back.

A lane-aware IC memo. A risk-signals scoreboard graded against firm policy — observed value, threshold, verdict. A KPI table where every figure cites a page or cell in the source documents. A “what must be true” list, a “what would change the call” list, a “likely pushback” list. A blockers section that names what the data does not yet support.

The partner does not look at it Friday night. They never do. The diagnostic does not change the cadence of human attention; it changes what is waiting when the attention shows up.

Saturday 8am — the first pass.

The partner reads the pack over coffee. Not the way you read a memo — the way you read a list of claims you have to defend Monday morning. Three things surface in the first half-hour:

An add-back the QoE did not flag. The diagnostic flagged a $1.4M legal-settlement add-back as recurring across three of the last four quarters. The QoE summary called it one-time. The diagnostic cited the line in the audited financials where the same vendor name appears in consecutive years. The partner doesn't know yet which is right. She knows it is now a question.

A customer concentration figure that looks high. Top customer at entry projection: 18%. Diagnostic shows top customer trailing-twelve at 24%. Both are plausible. The diagnostic cites the page in the CIM and the tab in the financial model. The partner notes the gap and moves on.

A covenant tighter than firm precedent. The proposed leverage covenant has a 0.25x cushion above projected. The diagnostic surfaced two recent firm precedents at 0.50x and 0.75x. Not a deal-killer; the kind of thing the partner would rather know before negotiating than after.

She forwards the pack to the associate at 8:47am with a three-line note: “Look at these three. Don't redo the rest.”

Sunday afternoon — the focused pass.

The associate spent Saturday on the surface analysis they would have spent the whole weekend on under the old workflow. Sunday is now available for depth. They come back to the partner at 5pm with three answers:

The add-back is partially defensible. One quarter is a genuine one-time settlement; two quarters look like the same vendor on a recurring services contract. The associate has a number for the cushion impact: roughly 18 basis points narrower than the QoE represented. Material, but not disqualifying. A negotiation item.

The concentration is real but stable. Top-three customer share has been 41–43% for the last four quarters. Top customer alone drifted from 21% to 24% on a single contract expansion that is now contracted through 2028. The risk is real; the trajectory is not deteriorating.

The covenant is tight but the precedent set the diagnostic surfaced is internally consistent — both prior deals had similar revenue concentration profiles. The associate drafted three alternative covenant structures, ranked by likelihood the lender accepts.

Monday 8am — the room.

The partner walks in with the diagnostic pack, the associate's annotations, and three answers ready. The committee asks about customer concentration. She has the cited number, the four-quarter trend, the contract expiry. Asks about the add-back. She has the QoE language, the diagnostic's challenge, and the basis-point impact. Asks about the covenant. She has the precedent set and three alternatives.

The committee debates the call. They do not debate the numbers under the call. The deal proceeds, or it doesn't — but the question of whether the team did the work is settled before the question of whether to invest is asked.

Or: the deal you should kill on Friday.

Most CIMs that land Friday afternoon should not make it to Monday. The diagnostic returns the blockers fast enough — by 5:30pm Friday, in most cases — that the partner can decide before the weekend whether the deal is worth the team's time at all.

A blocker pack reads cleanly: missing audited financials, covenant package not yet drafted, sponsor concentration above firm policy, lane mismatch with the CIM's own framing. The partner doesn't need to feel guilty about prioritization. She can tell the banker no on Monday morning, with a specific, named reason. That conversation is shorter and ends the relationship in better shape than “we passed.”

What this changes structurally.

The partner is back in the analytical loop. Not as the reader of someone else's memo, but as the person who reads the source-backed claims and decides where to push. The associate is not compressed into a 48-hour extraction sprint; they are deployed onto the questions that actually matter. The committee is not debating numbers no one in the room can defend.

The diagnostic is not the answer. It is the substrate that lets the people in the room talk about the deal instead of the data. The judgment is still the partner's. The defense, six months later, is still the partner's. The work is now rigorous enough to defend.

What to send.

Three things. The CIM. The financial model in xlsx. Either the QoE / audited financials (corporate PE) or the credit agreement and lease abstracts (private credit / real estate). The diagnostic returns the pack in the same shape every time, on the same day in most cases. No data prep, no schema mapping, no intake call.


If your committee meets Monday morning and you have a CIM in hand by Friday afternoon, the diagnostic is the cheapest way to take the weekend back. Run it on one deal. The next CIM that lands becomes the test.

Pressure-test one deal before the next committee.

Send the source materials Friday. Get the diagnostic pack back the same day. Walk into Monday with the questions already framed.