The Engagement: Proposal to Trusted Advisor

Consulting Galaxy · Branch: Foundations (cn-foundations) · ⬜ White Belt: Analyst · ~8 focused hours · Mastery gate ≥ 85%

You have, somewhere, produced a piece of analysis you were genuinely proud of, and watched it change nothing. Nobody argued with it. Nobody found a flaw in it. It simply sat in an inbox, or got a polite nod in a meeting, and the decision it was built to inform got made some other way, for some other reason, by people who were never quite persuaded to act on what you gave them. That experience is not a fluke of one unlucky memo. It is the single most common failure mode in advisory work, and it has almost nothing to do with the quality of the thinking. A consultant is not hired to be right in private. A consultant is hired to make a client confident enough to act (on a recommendation, on a number, on a plan that costs money and careers if it is wrong), and that confidence is not a mood you create in the last meeting with a good slide. It is built, or quietly destroyed, at every single stage of an engagement: in the proposal that first argued you deserved the work, in the contract that defined exactly what you owed, in the small moment three weeks in when you either absorbed a client's off-scope request for free or handled it like a professional, in the number a junior analyst got wrong and a senior one caught before the client ever saw it, in the feedback that made the junior analyst better instead of merely told them they were wrong, and in the one conversation, late in the engagement, where telling the client the truth cost your own firm money.

This node is the map of that whole arc (proposal to trusted advisor), and it exists because the node before this one taught you how a consultant thinks (structure, hypothesis, answer-first) while this node teaches how a consultant is actually retained, run, and trusted to put that thinking to work on a real, paying, anxious client. You will learn what a good proposal argues and how it differs from a sales pitch; how a Statement of Work turns a persuasive pitch into a document you can be held to; how a fee is actually built from people and time rather than pulled from the air; why scope creep is the single most common way a good engagement quietly goes bad, and the discipline that prevents it; how an engagement team is actually organized, delegated to, and quality-controlled; how feedback is given so it changes behavior instead of just recording an opinion; and, at the heart of it, the trust equation, a deceptively simple piece of arithmetic that explains why the most technically brilliant person in the room is sometimes the least trusted, and why the single highest-leverage thing an advisor can do is act, visibly, against their own short-term interest.

Here is all of it, compressed into one sentence you will spend eight hours earning the right to believe: trust is engineered in the paperwork and the small decisions, not manufactured in the big pitch. The proposal, the SOW, the change-control conversation, the QC pass, the feedback you give a junior teammate, the fee you quote and the fee you don't ask for when the honest answer is "you need less of us," and every one of these is a place where trust compounds or leaks, and none of them is the dramatic moment a beginner imagines "being a great consultant" is made of. By the end of this node you will be able to read a proposal and say precisely what it is and is not arguing; run a scope-creep request through a real test instead of a gut feeling; compute, by hand, why one advisor is trusted more than a more "impressive" rival; and reason all the way through an ethics dilemma to the action that actually protects the relationship, because a relationship, like a filing, is only as strong as the discipline you apply to it when nobody but you would notice if you cut a corner.

Learning objectives

By the end, you can:

1. Map an engagement end-to-end: problem statement → proposal → signed SOW → workplan and workstreams (each carrying a day-one hypothesis) → analysis and QC → the "so what" → the deck, and state, at each stage, what gets produced and who on the team owns it.
2. Read and draft a consulting proposal: name the six things a good proposal argues (situation, complication, objective, approach, team, timeline and fee) and explain why a proposal is an argument for being hired, not a sales brochure or a contract.
3. Read and draft a Statement of Work (SOW): state the components that make it the contract-grade reference the whole engagement is run against (deliverables, in-scope/out-of-scope, assumptions and dependencies, acceptance criteria, change control, governance) and distinguish it cleanly from the proposal that preceded it.
4. Price an engagement at concept level: build a fee from person-weeks × blended day rate by role, and choose defensibly among fixed-fee, time-and-materials, and value-based structures given who should bear the estimation risk.
5. Run the scope-creep discipline: test any mid-engagement request against the SOW's deliverables list, and route an out-of-scope ask to a quantified change order instead of silent absorption or a flat refusal.
6. Describe how an engagement team is actually run: the typical Partner/Principal → Engagement Manager → Associate/Consultant → Analyst ladder; delegate a task by matching it to a level with explicit success criteria and a check-in cadence; and enforce quality control (the two-person rule, the pyramid check, the fact-check pass) as a non-negotiable discipline rather than a courtesy.
7. Give and receive feedback that changes behavior: apply the Situation–Behavior–Impact (SBI) model in real time, in both directions on the team, and distinguish it from a saved-up annual review.
8. State, compute, and apply the trust equation (Maister, Green & Galford): Trust = (Credibility + Reliability + Intimacy) / Self-Orientation, and define each term precisely, compute a worked comparison between two advisors, and explain why lowering self-orientation is the single highest-leverage lever in the formula.
9. Reason through the ethics of advice: independence, conflict-of-interest disclosure, the "advisor's dilemma" of recommending less work when that is the honest answer, and when the professional response is to walk away from an engagement.
10. (Productivity objective: the duality below.) Use AI as a research and drafting amplifier for proposals, SOWs, delegation plans, and fee benchmarks, while keeping every commercial commitment, every scope boundary, and every judgment about who is ready for what task fully in your own hands.

The duality, stated once. Objectives 1–9 are the understanding objective the mastery gate rewards: you earn the belt by explaining the engagement's mechanics and reasoning correctly through a scope, delegation, or ethics scenario by hand. Objective 10 is the productivity payoff you keep afterward: knowing where an AI tool can draft a useful first pass on the paperwork of an engagement, and exactly where your own judgment about scope, people, and the client relationship must never be delegated to it. A tool can draft a proposal skeleton. It cannot decide what you owe a client, or whether a teammate is ready for a task, or whether the honest recommendation costs your firm money: those judgments are what this belt certifies.

Prerequisites & connections

Builds on. CN0.01 (The Consulting Mind), which teaches the problem-solving cycle this node's engagement map assumes (define → structure → prioritize → analyze → synthesize → recommend), and which states, once and honestly, the scope this whole galaxy works within: partner-level judgment is compounded from hundreds of live engagements and cannot be manufactured by any curriculum, so what is taught here is the teachable floor beneath that judgment: the structures, the paperwork, the team discipline, and the relationship mechanics a great advisor stands on. Nothing else is assumed: this node opens the White Belt and needs no finance-galaxy background to follow, though its arithmetic (fee build-ups, weighted averages, simple ratios) is the same comfort with numbers Phase 0–2 of the Finance galaxy assumes.

Feeds forward. CN1.01–CN1.03 take the "hypotheses per workstream" and "the so what" steps named here and teach their actual mechanics in full (MECE issue trees, hypothesis-driven analysis, and the Minto-style answer-first synthesis), so this node's engagement map is the skeleton those nodes put flesh on. CN7.02 returns to team and client leadership at the delivery stage (running client working sessions and steering committees, stakeholder management up and down), and is this node's direct sequel once you are inside a live engagement rather than opening one. CN7.03, the Partner Capstone, requires you to run an entire engagement solo under time, including a steering-committee-style interim update and client pushback, and everything this node teaches about the engagement's shape, the fee, the team, and the relationship is load-bearing there.

What this node explicitly hands off, and does not re-teach. The consulting galaxy is additive to the Finance galaxy, never a duplicate of it, and this node is disciplined about its own boundaries too. The mechanics of MECE issue trees, hypothesis-driven problem solving, and case math belong to CN1.01–CN1.02/CN2.01: this node names "the day-one hypothesis" and "the workstream" as steps in the map without teaching how to build one. The full mechanics of the answer-first pyramid (governing thought, MECE support, SCQA framing) belong to CN1.03 (the "So What" crown) and to M10.03 (The Communication Toolkit), which already teaches Barbara Minto's Pyramid Principle in depth; this node uses "the so what" only as the named final step of the engagement arc and points there for the craft. Deck and storyline design belongs to the future Presentation galaxy; this node teaches only that the recommendation must be answer-first, never how to lay out a slide. And this node's taster-and-deepening relationship with the Finance galaxy's own elective runs both ways: E11.03 (The Consulting Case-Interview Lab) and its case CL-11.3 are the Cycle-2 on-ramp into this galaxy: if that is where you arrived from, this node is "the whole kitchen" the taster promised; if you continue past CN0, you will meet issue trees, case math, and full cases again, taught to a deeper bar.

The engagement as the unit of work

An engagement is the unit everything in this galaxy assumes: a bounded piece of advisory work, sold, scoped, staffed, delivered, and closed, for one client, over a defined stretch of time. Its shape is remarkably constant across firms, industries, and case types, and it is worth holding the whole arc in view before zooming into any one stage, the same way you would look at a full income statement before studying any one line:

Problem statement → proposal → signed SOW → workplan and workstreams (each with a day-one hypothesis) → analysis, exhibits, and QC → synthesis (the "so what") → the recommendation and the deck → delivery, feedback, and, if it went well, the next engagement.

Two things about this arc are easy to miss on a first look. First, it does not end at the deck. An engagement that produces a brilliant recommendation and is never asked back has, by the standard this node teaches, only half-succeeded: the point of doing the work well is to become the advisor the client calls before the next problem becomes a crisis, not merely to have been right this once. Second, the arc is not a straight line: a workstream's early analysis routinely kills the day-one hypothesis that spawned it, sending the team back to re-hypothesize, and a scope conversation can reopen mid-engagement when new facts surface. What stays constant is the discipline applied at each stage, not the guarantee that each stage is visited exactly once. This node walks the arc stage by stage: what each stage is for, what a professional produces there, and where the relationship is quietly won or lost.

The proposal: what a consultant is actually arguing

A proposal is not a brochure and not yet a contract. It is a specific, falsifiable argument for why this firm, with this team, using this approach, in this time, for this fee, is the right choice to solve this client's problem, and a proposal that could be sent to any client with the names swapped has already failed, regardless of how polished it looks. A good proposal argues six things, in this order, because the order itself is the argument:

1. Situation: the client's business context, stated in the client's own language and numbers, proving you understood the business before you pitched a solution to it.
2. Complication: the specific problem, quantified wherever you can (falling margin, a stalled market entry, a merger integration going over budget), not a generic category ("we help companies grow").
3. Objective: what a good outcome looks like and by when, stated as a decision the client will be able to make, not a document you will hand over.
4. Approach: the workstreams you propose and the day-one hypotheses each will test (enough to prove you have a plan, deliberately not the full analysis, since a proposal that gives away the whole answer for free has priced itself at zero).
5. Team: who is staffed, and the specific experience that makes them credible for this problem, not a generic list of past clients.
6. Timeline and fee: the phases, the milestones, and the commercial terms (§ below), stated plainly enough that nobody is surprised later.

Notice what is conspicuously absent from that list: a guarantee of the answer. A proposal that already knows the conclusion before the work starts is either lying or has not actually scoped a problem worth paying to solve: the honest proposal commits to a process (this team, this approach, this timeline) and a day-one hypothesis to test, not a pre-baked conclusion. This is the same discipline you will drill in full later: hypothesize, then let the analysis confirm or kill it. Here, it shows up first as a commercial promise: you are selling a disciplined way of finding the answer, not the answer itself.

The Statement of Work: turning a pitch into a contract

Once a proposal is accepted, it is rewritten into a Statement of Work (SOW): the document the entire engagement is subsequently managed against, and the single most important artifact in this whole module, because almost every dispute that damages a client relationship traces back to something the SOW did or did not say. A complete SOW carries:

Hold the distinction between proposal and SOW sharply, because conflating them is a common beginner error: the proposal persuades; the SOW commits. A proposal can be aspirational and a little bit of a sales document without lying; a SOW cannot, because the SOW is what a court, a client's procurement team, or, far more often, an aggrieved client three months from now will actually reread. Everything from here forward in the engagement (the fee build, the scope-creep test, the workplan) is downstream of what the SOW actually says, which is exactly why the discipline of writing it precisely, before the relationship gets busy and informal, pays for itself the first time a client asks for "just one more thing."

``ai-augment-json { "skill": "Drafting a first-pass proposal outline and a matching SOW skeleton from scoping-call notes", "use": "Have a chat assistant turn rough notes from a client scoping conversation into a structured proposal outline (situation, complication, objective, approach, team, timeline, fee) and a matching SOW skeleton (deliverables, in/out-of-scope, assumptions, change-control clause) as a fast first draft you then rewrite in your own judgment — never a document you send unedited.", "tools": ["Chat assistants — Claude, ChatGPT, Gemini", "A word processor's built-in AI drafting add-in"], "prompt": "Here are my rough notes from a scoping call, industry and situation only, no real client specifics: <NOTES>. Draft a one-page proposal outline with these sections: Situation, Complication, Objective, Approach (3-4 workstreams, each with a one-line day-one hypothesis), Team, Timeline, Fee & commercial terms. Then draft a matching SOW skeleton with explicit in-scope and out-of-scope bullets and a change-control clause. Flag anything you had to guess.", "verify": "Every scope boundary, deliverable, and fee figure the draft invents must be checked against what you actually agreed on the call — a fluent draft can quietly widen or narrow scope in a way that reads fine but does not match reality. Trace every commercial term back to your own notes before anything goes to the client (Guardrail #1 Trace and #2 Match, applied to a proposal instead of a filing).", "diy": "You must already know what a good proposal argues and what a SOW must contain (this section) well enough to catch a draft that omits the change-control clause or invents a scope boundary nobody discussed. AI drafts the skeleton; you own the commercial commitment.", "market": "IN" } ``

Pricing an engagement at concept level

A consulting fee is not a number pulled from confidence; at concept level it is built the same way a workplan is built: bottom-up, from people and time. The core unit is the person-week: one person, staffed at a given FTE (full-time-equivalent fraction: 1.0 is fully staffed, 0.3 is roughly a day and a half a week), for one week of the engagement. A blended day rate by role (Partner/Principal, Engagement Manager, Consultant, Analyst, each priced differently, reflecting seniority and cost) turns person-weeks into money:

Fee (by role) = FTE × weeks on the engagement × day rate × working days per week.

Sum every role's fee and you have the engagement's cost build-up: the number a proposal's "fee" line is ultimately defended against, even when the client only ever sees the final figure. Worked Example 1 recomputes this in full for a real-shaped scenario; hold the formula, because it is what turns "we think this should cost about ₹1 crore" from a guess into a number you can defend line by line if a client asks.

Three commercial structures then decide who bears the risk that the estimate is wrong:

• Fixed-fee. One number for the whole SOW, regardless of how many actual hours it takes. The firm bears the estimation risk: if the analysis runs long, the firm's margin absorbs it, not the client's budget. Clients favor this when the deliverable is well-defined and they want budget certainty.
• Time-and-materials (T&M). The client pays for actual hours or days worked, at agreed rates, typically with a not-to-exceed ceiling. The client bears more of the estimation risk: if the work runs long, the invoice grows, in exchange for not overpaying if it runs short. Favored when the scope is genuinely hard to estimate up front (early-stage diagnostics, exploratory work).
• Value-based / success fee. Some or all of the fee is tied to a measured outcome (a percentage of realized cost savings, a bonus on a successful deal close). This aligns incentives most tightly with the client's actual result, but it requires a measurable, attributable outcome and a level of trust in the measurement that most engagements have not yet earned, which is why it shows up more often with a repeat client than a first-time one.

None of these is "the right one" in the abstract; the right choice is the one that puts the estimation risk on the party better positioned to manage it, given how well-defined the scope actually is (Practice P13 gives you a concrete case to reason through).

Scope creep: the small print where engagements actually die

Scope creep is not a single dramatic event; it is a slow accumulation of small, individually reasonable-sounding requests that fall outside the signed SOW and get absorbed without a change order, until, weeks later, the team has quietly done 20% more work for the original fee, the schedule has slipped, and nobody can point to the single moment it went wrong because there wasn't one. It is the most common way a well-run engagement turns into a resented one, on both sides: the client feels nickel-and-dimed if they're charged for what felt like a small ask, and the firm feels exploited if it wasn't. The fix is not to refuse every request, and it is not to say yes to everything to "keep the relationship warm": both are failures of the same discipline. The fix is a test, run every time, against the SOW's own deliverables list:

1. Is this explicitly in-scope? If yes, it was already priced in: do it, no conversation needed.
2. Is this explicitly out-of-scope, or simply not mentioned? Silence is not an inclusion. If it is out-of-scope or unaddressed, it is a candidate for a change order: quantify the additional person-weeks and fee (using the same build-up as the SOW itself), and put the choice honestly to the client: add the scope and the fee together, trade it for something already in scope, or decline it and move on.
3. Who has the conversation? The Engagement Manager, not the analyst fielding the request on the ground: the discipline needs someone with the standing to say "yes, and here is what changes" without it reading as either a favor or a confrontation.

The reason this belongs beside the trust equation later in this node is not incidental: absorbing scope silently feels like generosity, but it teaches a client that the SOW is not a real document, which corrodes the very reliability (doing what you said, on the terms you said it) that trust is built from. A transparent change order, even one the client sometimes declines, is the more trust-building move, because it proves the paper trail means something.

The workplan and the weekly rhythm

The SOW's timeline becomes, in practice, a workplan: the engagement broken into workstreams, each carrying its own day-one hypothesis (its full mechanics taught later, not re-taught here) and its own set of exhibits and deliverables, sequenced against milestones. Most engagements settle into a weekly rhythm that repeats until delivery: a Monday planning touchpoint (what does each workstream need to answer this week, what data or access is still outstanding), a mid-week check (is any workstream stuck, does a hypothesis need to be killed and replaced before more time is sunk into it), and a Friday client touchpoint (a short, informal update, not a full presentation, so the client is never surprised by what the final deck says). This last habit deserves its own emphasis: an engagement that saves every finding for the final meeting is running a real risk of a bad surprise in the room, and a later module will teach the full discipline of "pre-wiring" a steering committee so that nothing in the final meeting is heard there for the first time. Here, hold the simpler version: frequent, informal, honest updates are cheaper than one large risky reveal.

A related habit worth naming now, because you will meet its mechanics properly later, is the ghost deck: a skeleton of the final presentation's structure and exhibit titles, built early and refined weekly, even before the numbers inside it are final. It forces the team to know what argument they are building toward at every stage, rather than accumulating analysis and hoping a narrative appears at the end. Building the ghost deck is a structuring skill (CN1); running the rhythm that keeps it honest, week over week, is an engagement-management skill: this node's job.

Running the team: roles, delegation, and the discipline of QC

Most engagement teams, across firms, settle into a version of the same ladder:

Delegation is done well when a task is matched to the level that can execute it with the appropriate amount of oversight: not zero (abandonment, which produces silent errors and missed deadlines) and not constant (hovering, which wastes the senior person's time and stunts the junior one's growth). The formula worth internalizing: task matched to level + an explicit, one-line success criterion ("done" means the weighted-average margin reconciles to the underlying revenue mix, not just "a number appears") + a defined check-in cadence (a daily five-minute stand-up, a mid-week deeper review). A task handed over with none of the three is not delegation; it is hoping.

Quality control (QC) is the discipline that catches errors before a client ever sees them, and it holds regardless of how senior or confident the person who made the error is: in fact, the most expensive mistakes are disproportionately made by the most confident person in the room, precisely because confidence suppresses the instinct to double-check. Three specific QC habits are worth naming as non-negotiable:

• The two-person rule. Any number or claim that will reach the client gets a second reviewer before it ships: no exception for seniority, and no exception for "I already checked it myself twice." A second, different pair of eyes catches a different class of error than the same person re-reading their own work.
• The pyramid check. Confirms the top-line message (the "so what") is genuinely supported by the exhibits assembled beneath it, not merely that each individual fact is correct in isolation. A slide can contain twelve true numbers and still support a false conclusion if they were assembled carelessly; the pyramid check is the review of the assembly, not just the arithmetic.
• The fact-check pass. Traces every cited figure back to a named source before it ships: the same Trace / Match discipline the AI-Augmented Analyst branch teaches for verifying an AI's output applies with equal force to a human teammate's slide; a wrong number from a colleague is exactly as dangerous to the client relationship as a hallucinated one from a model, and the review discipline that catches it is identical.

Worked Example 3 shows this discipline catching a real, reconcilable error before it reached a client.

``ai-augment-json { "skill": "Drafting a delegation plan and a QC checklist for a new workstream", "use": "Use AI to turn a workstream's objective into a first-pass task breakdown — who owns what, at what level, with what success criterion and check-in cadence — and a QC checklist naming which numbers or claims need a second reviewer, as a scaffold the Engagement Manager then edits and owns.", "tools": ["Chat assistants — Claude, ChatGPT", "A workplan or project-tracking tool's built-in AI drafting feature"], "prompt": "This workstream's objective is: <OBJECTIVE>. My team, by role only (no real names): <ROLES>. Draft a task breakdown assigning pieces of this workstream to the right level, with a one-line success criterion and a check-in cadence for each. Then draft a QC checklist naming every number or claim in this workstream that should get a second reviewer before it reaches the client.", "verify": "The draft cannot know your actual team's real skill level, current workload, or this specific client's sensitivities — re-match every task to the real people on your team (not the generic levels the model assumed), and add any QC item specific to this engagement that a generic checklist would miss.", "diy": "You must already know the delegation formula (task matched to level, explicit success criteria, a real check-in cadence) and the two-person QC rule (this section) to judge whether a draft under- or over-delegates. AI proposes a scaffold; you dispose of whatever does not fit your actual team.", "market": "US" } ``

Feedback that actually changes behavior

QC catches an error once; feedback is what stops it from recurring, and most beginners give it badly in one of two directions: either so vague it cannot be acted on ("good job, just tighten it up") or saved up for a single end-of-engagement review, by which point the behavior has repeated for weeks and the specific moment is long forgotten. A far more effective structure, taught widely under the name SBI (Situation–Behavior–Impact), forces specificity in exactly the way vague feedback lacks it:

• Situation: name the specific moment and context ("in Wednesday's QC review of the market-sizing slide…").
• Behavior: describe the observable action, not a character judgment ("…you used a simple average of the three product lines' margins…" rather than "you were careless").
• Impact: state the effect it had, concretely ("…which would have overstated the blended margin by nearly two points on a slide the client would have used to set a price point").

The habit that makes SBI actually change behavior is timing: given in real time, close to the moment, it is specific and low-stakes enough to land as coaching rather than judgment. Saved up for a formal review, the same content reads as an accusation, arrives too late to be actionable on the work it describes, and, because six weeks of small moments have blurred together, is rarely specific enough to be SBI at all anymore. And feedback is not a one-directional privilege of seniority: a junior teammate who tells an Engagement Manager, specifically and in the moment, that a client touchpoint ran long without reaching a decision is offering exactly the same discipline upward that QC offers downward: the team that only gives feedback down its own hierarchy is missing half the signal available to it.

The trust equation: engineering the relationship on purpose

Everything so far in this node (the proposal, the SOW, the fee, scope discipline, QC, feedback) is machinery. The trust equation, formalized by David Maister, Charles Green, and Robert Galford in The Trusted Advisor, is the model that explains why running that machinery well is what actually earns the relationship, rather than treating "being trusted" as a separate soft skill you either have or don't. The formula:

Trust = (Credibility + Reliability + Intimacy) ÷ Self-Orientation

Each term names something specific and checkable, not a vague vibe:

• Credibility: do your words ring true? Are you technically competent, and do you say things that turn out to be accurate? ("I trust what she says about pricing.") This is the term the proposal and the analysis itself are mostly building.
• Reliability: do your actions match your words, consistently, over time? ("I trust her to actually do what she said, by when she said it.") The SOW, the workplan rhythm, and the scope-creep discipline are what this term is built from: every kept commitment adds to it, and every silently absorbed scope change or missed check-in quietly subtracts.
• Intimacy: is it emotionally safe for the client to share something sensitive or risky with you, without fear of exposure or judgment? ("I trust her with this problem, not just with this fact.") Intimacy is not friendliness, and it is not "we get along": a client can golf with an advisor every weekend and never once feel safe telling them the business is actually in trouble; intimacy is about psychological safety around the client's real vulnerabilities, which is exactly why it is the term most beginners misdiagnose (Common mistake #5, below).
• Self-orientation: whose interest is actually being optimized: the client's, or the advisor's own (billings, being seen as the smartest person in the room, winning the follow-on sale)? This is the denominator, not one more thing to add, and that placement is the whole insight of the formula, worked through in full arithmetic in Worked Example 2.

Because self-orientation divides the sum of the other three rather than adding to it, a small reduction in self-orientation moves total trust by more than an equivalent-sized increase in credibility or reliability: the denominator is a multiplier on the whole relationship, while the numerator terms are merely additive. Concretely: raising credibility from a rating of 9 to 12 (on the illustrative 1–10-ish scale the book itself uses to make the formula tangible) adds 3 to a numerator that then still gets divided; lowering self-orientation from 6 to 3 doubles the entire fraction. This single mechanical fact (not a platitude about "always put the client first," but an arithmetic consequence of where a variable sits in a formula) is why the single highest-leverage thing an advisor can do is visibly act against their own short-term interest: turn down billable follow-on work that is not truly justified, admit a mistake before it is discovered, or tell a client the honest but fee-reducing truth. Worked Example 2 computes this comparison in full, and Worked Example 4 shows the highest-stakes version of it: the ethics of recommending less work.

The ethics of advice: independence, conflicts, and the courage to say less

Trust's self-orientation term is more than a relationship-building tactic: it is the mechanical seed of the whole professional ethic this node closes on. Three specific disciplines follow directly from taking self-orientation seriously rather than treating it as a nice sentiment:

Independence. A recommendation must be argued from the analysis, not bent toward whatever outcome benefits the firm: most commonly, the temptation to recommend a larger or longer engagement than the facts actually justify, because a follow-on phase is easier revenue than winning a new client from scratch. The discipline is structural, not just willpower: separate the person who scopes and prices the next phase from the incentive to inflate it, and hold every follow-on recommendation to the same evidentiary bar as the first one.

Conflict-of-interest disclosure. When a firm's relationship with one client creates a real or perceived conflict with another (most commonly, being asked to advise a client on a question where the firm already advises that client's direct competitor), the discipline is to disclose, not to self-adjudicate silently. Disclosure is typically resolved one of two ways: an ethical wall (separate, non-overlapping teams with no shared access to either client's confidential information), if both clients knowingly consent to proceed on that basis; or declining one of the two engagements, if a wall cannot credibly protect confidentiality or either client is unwilling to proceed. Neither resolution is optional once a real conflict exists: the failure mode is not picking the wrong resolution, it is skipping the disclosure that would have surfaced the choice at all.

The advisor's dilemma. The hardest and highest-leverage version of independence is the moment the honest analysis says the client needs less of you, not more: that a proposed follow-on phase would cost more than it would save, that a program should stop, or that a smaller and cheaper engagement would serve the client better than the one already being pitched. Saying this costs the firm money in the moment it is said. It is also, per the trust equation's own arithmetic, one of the highest-leverage trust-building acts available, because self-orientation is a denominator and this is the clearest possible demonstration that it is low. Worked Example 4 walks through exactly this decision with real, reconciled numbers.

When to walk away. All of the above assumes the conflict or pressure can be managed within the engagement. Occasionally it cannot: a client insists on shading a finding, misrepresenting data to a regulator or board, or pressures the team past what disclosure and walls can resolve. The professional response, rare and always a last resort, is to say so plainly and, if the situation is not corrected, to end the engagement. A firm that cannot, even in principle, reach that decision does not have a real independence discipline: it has a policy that only holds until it is inconvenient.

From analysis to the deck: the hand-off you do not re-teach here

The engagement's arc ends, formally, at the recommendation and the deck, and this node names that step deliberately without teaching its mechanics, because doing so properly is a full node's worth of craft on its own. The discipline you owe the client at this final stage is simple to state and hard to execute: answer first. State the recommendation before the supporting analysis, structure the support so it is genuinely exhaustive and non-overlapping (MECE), and never let a client sit through twenty minutes of build-up before learning what you actually think they should do. That is the whole of what belongs here. The how (governing thought, MECE grouping, the SCQA opening, the pyramid itself) belongs to this branch's crown module and to this program's own Communication Toolkit; the design of the deck that carries it belongs to the future Presentation galaxy. This node's job was only ever to place "the so what" correctly on the map and insist that it come first: the craft of building it is a deliberate, well-marked hand-off, not an oversight.

Where this leaves you

Hold the whole arc together now. A proposal argues for the work in six specific moves. A signed SOW turns that argument into the reference the whole engagement is managed against, including, especially, against a client's later, reasonable-sounding request that falls outside it. A fee is built from people and time, not confidence, and priced through a structure that puts the estimation risk where it belongs. A team is delegated to and quality-controlled with the same rigor whether the person making the claim is a first-week analyst or the partner who sold the work. Feedback, given in the moment and in both directions, is what makes the team better rather than merely graded. And underneath every one of those mechanics sits the same single variable, engineered on purpose rather than hoped for: self-orientation, the one term in the trust equation that divides rather than adds, and the one an advisor controls completely, in every single interaction, independent of how talented or well-resourced the rest of the team happens to be. That is the payoff this node exists to install: trust is engineered in the paperwork and the small decisions, not manufactured in the big pitch, and every mechanic in this node, from the SOW's change-control clause to the courage to recommend less work, is that one sentence made operational.

Common mistakes & how experts think differently

1. "The proposal is a sales document, so oversell it." A proposal that promises more than the SOW and the team can actually deliver is not persuasive, it is a future credibility problem with a delay built in: the moment reality falls short of the pitch, the credibility term of the trust equation takes the damage all at once. Experts write a proposal precisely as good as what will actually be delivered, and let the six-part argument (situation → complication → objective → approach → team → timeline/fee) do the persuading instead of exaggeration.
2. "Scope creep is just being helpful: say yes to the client's small ask." Every "yes" outside the SOW either becomes a transparent change order or a silent, invisible tax on the team's time and the engagement's quality: there is no third option where it is simply free. Experts run the in/out-of-scope test every time and treat "yes, and here is what changes" as the generous move, not "sure, no charge," which quietly teaches the client the SOW was never a real document.
3. "Delegate by handing over the task and disappearing" (or its mirror image: hovering over every step). Both are failures of the same formula. Experts match the task to a level that can execute it, state an explicit one-line success criterion up front, and set a real check-in cadence: delegation is a design decision made before the work starts, not an amount of attention decided moment to moment.
4. "Feedback is an annual-review event." By the time a formal review happens, the specific moment is gone, the behavior may have repeated for weeks, and the feedback arrives too diffuse to be actionable. Experts give SBI feedback close to the moment it happens, specific and behavior-focused rather than character-focused, and expect it to flow both up and down the team, not only from senior to junior.
5. "Trust-equation intimacy means being friendly with the client." Confusing intimacy with sociability is the single most common misreading of the formula. Intimacy is the emotional safety to share something genuinely risky (a business in real trouble, a mistake, an unresolved fear) without fear of exposure or judgment; a client can attend every dinner you host and never once feel safe enough with you to say the number is worse than the board has been told, and that gap is exactly what the term measures.
6. "The QC pass is optional once the team is senior enough to trust." Seniority does not retire the two-person rule; if anything it heightens the risk, because the most confident person in the room is often the least likely to re-check their own work. Experts apply the identical QC discipline (two-person rule, pyramid check, fact-check pass) regardless of who authored the number, and treat "I'm sure, I checked it myself" as a reason to check again, not a reason to skip the second reviewer.
7. "A good advisor always has the answer in the room and never says less-of-us." The instinct to always propose more work, and never admit a gap, is exactly the self-orientation the trust equation punishes, and the confident wrong answer given rather than a candid "I don't know yet, and here's when I will" carries the same failure signature the AI-Augmented Analyst branch teaches about a model's plausible-but-wrong output: fluent confidence is not evidence of correctness, whether the source is a person or a machine. Experts treat "you need less of us" as the single highest-trust sentence available, precisely because it is the one that costs the speaker money to say.
8. "Ethics is a form you sign once, at onboarding." Independence and conflict management are not a one-time compliance checkbox; they are re-examined every time the scope, the client contact, the fee structure, or the firm's other client relationships change. Experts re-run the conflict-of-interest question at every material change in the engagement, not only at its start, and treat "we already cleared this" as the beginning of the question, not the end of it.

The thread through all eight: the expert never treats "being trusted" as a personality trait or a mood to project in the room. They treat it as the visible output of a set of unglamorous disciplines (a precise SOW, an honest fee build, a real QC pass, timely feedback, and a self-orientation kept deliberately low even when lowering it costs money), applied consistently enough, for long enough, that the client stops needing to double-check you. That is the whole of what "trusted advisor" means, operationally, and it is exactly what this node's paperwork was teaching all along.

Worked examples

Worked example 1: Pricing a proposal and pricing a scope-creep request (India, ₹)

A boutique strategy firm is scoping a 10-week national profitability diagnostic for Suvidha Retail, a synthetic Indian retail chain (every figure here is illustrative and internally consistent, not a real firm's real rates). The proposed team: an Engagement Manager at 0.3 FTE, two Consultants at 1.0 FTE each, and one Analyst at 1.0 FTE, over the full 10 weeks, 5 working days per week. Illustrative blended day rates: EM ₹1,25,000; Consultant ₹60,000; Analyst ₹35,000. Three weeks in, the client asks for a 3-city market-expansion sizing that was never in the original deliverables list: should the firm quietly do it, refuse it, or run it through the change-control clause?

Solution. Step 1: build the fee role by role. Fee (by role) = FTE × weeks × day rate × days/week.

• EM: 0.3 × 10 = 3.0 person-weeks; weekly rate = ₹1,25,000 × 5 = ₹6,25,000; fee = 3.0 × ₹6,25,000 = ₹18,75,000.
• Consultants (2, each 1.0 FTE): 1.0 × 10 × 2 = 20.0 person-weeks total; weekly rate = ₹60,000 × 5 = ₹3,00,000; fee = 20.0 × ₹3,00,000 = ₹60,00,000.
• Analyst: 1.0 × 10 = 10.0 person-weeks; weekly rate = ₹35,000 × 5 = ₹1,75,000; fee = 10.0 × ₹1,75,000 = ₹17,50,000.

Step 2: sum the build-up. Total fee = ₹18,75,000 + ₹60,00,000 + ₹17,50,000 = ₹96,25,000 (₹96.25 lakh, just under ₹1 crore). This is the number the proposal's fee line is defended against, role by role, if the client ever asks how it was built.

Step 3: run the scope-creep test on the mid-engagement ask. Is the 3-city expansion sizing in the original deliverables list? No: the SOW's deliverables were a national profitability diagnostic across two workstreams, and market-entry sizing for new cities was never named, so it fails the in-scope test regardless of how reasonable it sounds on the call. It is a change-order candidate, not a free add and not a flat refusal.

Step 4: price the change order the same way. Suppose it needs one additional Consultant at 0.5 FTE for the 4 remaining weeks: person-weeks = 0.5 × 4 = 2.0; weekly rate = ₹3,00,000 (same Consultant rate); change-order fee = 2.0 × ₹3,00,000 = ₹6,00,000 (₹6 lakh), about 6.2% of the original fee (₹6,00,000 ÷ ₹96,25,000 ≈ 6.23%).

Read it. Notice what quietly absorbing this would have cost: ₹6 lakh of unpriced work is a small-sounding ask in the room and a real, measurable margin leak on paper: precisely the gap between "feels like a favor" and "is actually 6% of the whole engagement's fee." Raising it as a transparent change order, and letting the client decide whether to add the scope, trade something out, or decline, is the reliability-building move; silently doing it for free teaches the client that the SOW's deliverables list was decorative.

Worked example 2: The trust equation, computed by hand (US, $)

Meridian Foods Co., a synthetic US mid-market client, is choosing between two advisors pitching the same engagement. Rate each, illustratively, on a 1–10 scale. Advisor A: Credibility 9, Reliability 7, Intimacy 4, Self-orientation 6. Advisor B: Credibility 7, Reliability 8, Intimacy 7, Self-orientation 2. Advisor A is, on paper, more "impressive" (higher credibility). Which one does the trust equation say the client should actually trust more, and by how much?

Solution. Step 1: compute Advisor A's trust score. Trust = (C + R + I) ÷ S = (9 + 7 + 4) ÷ 6 = 20 ÷ 6 = 3.33.

Step 2: compute Advisor B's trust score. Trust = (7 + 8 + 7) ÷ 2 = 22 ÷ 2 = 11.0.

Step 3: read the gap. Despite a lower raw credibility score, Advisor B's trust score is more than three times Advisor A's, because Advisor B's self-orientation (the denominator) is a third the size. A client choosing on "who sounds smartest in the room" would pick Advisor A and, on this model, get the less trustworthy advisor.

Step 4: show why lowering the denominator beats raising the numerator. Suppose Advisor A worked on the numerator instead, raising credibility from 9 to 12 (a large jump, and one credibility alone rarely earns quickly): new trust = (12 + 7 + 4) ÷ 6 = 23 ÷ 6 = 3.83, barely moved. Now suppose Advisor A instead lowered self-orientation from 6 to 3 (visibly turning down unnecessary billable work, admitting an early mistake, giving the client bad news early): new trust = (9 + 7 + 4) ÷ 3 = 20 ÷ 3 = 6.67, nearly double, from a change to one variable that did not require becoming more technically impressive at all.

Read it. The arithmetic is the whole lesson: because self-orientation divides the sum rather than adding to it, it is a multiplier on the entire relationship, not one more ingredient alongside the other three. This is why "act against your own short-term interest, visibly" is not a platitude in this node: it is the single highest-leverage move available in the formula, and it costs nothing technical to execute, only nerve.

Worked example 3: A QC catch: weighted average versus simple average (India, ₹)

An Analyst on the Ananta Consumer Products engagement (synthetic FMCG conglomerate) is building a market-sizing slide and needs the company's blended gross margin across three product lines: Line 1: revenue ₹40 cr, gross margin 45%; Line 2: revenue ₹25 cr, gross margin 30%; Line 3: revenue ₹15 cr, gross margin 60%. The Analyst's draft slide states a blended gross margin of 45%, computed as the simple average of the three margins. The Engagement Manager, running the two-person QC rule before the slide ships, stops and recomputes.

Solution. Step 1: name the correct method. A blended margin across lines of different sizes must be revenue-weighted, not a simple average of the three percentages: a line that is half the business should count roughly twice as much toward the blend as a line that is a quarter of it.

Step 2: compute the correct, weighted figure. Total revenue = ₹40 cr + ₹25 cr + ₹15 cr = ₹80 cr. Weighted gross margin = (40 × 0.45 + 25 × 0.30 + 15 × 0.60) ÷ 80 = (18 + 7.5 + 9) ÷ 80 = 34.5 ÷ 80 = 43.125% (43.1%).

Step 3: quantify the analyst's error. The simple average of the three margins is (45% + 30% + 60%) ÷ 3 = 45.0%, an overstatement of 1.875 percentage points (45.0 − 43.125) versus the correctly weighted figure.

Step 4: see why the direction is not fixed. Here the simple average happened to overstate the truth, because the highest-margin line (60%) is also the smallest by revenue, so weighting pulls the blend down. That is a property of this data, not a rule: a differently-weighted mix can just as easily make the simple average understate the truth (Form A item A11 shows exactly that reversal). The lesson is not "simple averages are always too high"; it is that an un-weighted average of ratios across unequal bases is the wrong tool full stop, and must always be re-derived from the underlying totals.

Read it. On its own, 1.9 points sounds small. On a slide feeding a client pricing recommendation, it is exactly the kind of quiet, plausible-looking error the two-person rule exists to catch before it reaches the room, and it cost the EM under a minute to recompute from the totals, versus the cost of a client later catching it themselves.

Worked example 4: The advisor's dilemma (US, $)

Northgate Retail Group, a synthetic US client, hired a firm for a Phase 1 cost-transformation engagement: fee $650,000, and the team's own tracked results show $2,100,000 in annualized savings actually delivered. The client is delighted and the firm's account team has already drafted a Phase 2 proposal at a fee of $850,000. But the team's honest workplan analysis for Phase 2 (the same rigor applied to Phase 1) projects only $300,000 of net new annual savings achievable, because the easiest cost levers were exhausted in Phase 1. Should the team present the Phase 2 proposal as drafted?

Solution. Step 1: hold Phase 1 to its own standard. Phase 1's return was $2,100,000 ÷ $650,000 = 3.23×, a clearly value-creating engagement by any reasonable test, and the reason the client trusts the firm enough to consider a Phase 2 at all.

Step 2: apply the identical standard to Phase 2. Phase 2's projected return is $300,000 ÷ $850,000 = 0.35×: the client would be paying $850,000 for $300,000 of value. Judged by the same yardstick that justified Phase 1, this is not a marginal case; it is value-destructive.

Step 3: name the two options and their cost. Presenting the Phase 2 proposal as drafted earns the firm $850,000 in the short term and, the moment the client's own finance team runs the same arithmetic (which a client this analytically engaged eventually will), destroys the credibility and reliability the firm built in Phase 1 all at once. Presenting the honest finding (that a large Phase 2 is not justified, and offering, if genuinely warranted, only a small, targeted sustainment engagement instead) costs the firm the $850,000 fee today.

Step 4: apply the trust equation to the choice. This is Worked Example 2's arithmetic in its highest-stakes form: telling the client the truth is a direct, visible reduction in self-orientation, the trust equation's denominator. It is also the single most expensive sentence in this entire node to say out loud, and the single most valuable one for a long-run relationship: the client who is told "you don't need a big Phase 2" is precisely the client who calls this firm first, unprompted, the next time a real problem appears.

Read it. Notice this was not a close call once both phases were held to the same standard: the dilemma is not in the arithmetic, which is unambiguous, but in whether the team has the discipline to run the same test on its own follow-on proposal that it would run on any other investment the client was considering, and the courage to act on the answer when it costs the firm money.

Worked example 5: Reading a proposal's opening for what it actually argues (India, ₹)

Two drafts of the opening paragraph of a proposal for DeccanJet, a synthetic regional Indian airline, land on an Engagement Manager's desk. Draft 1: "We are a leading advisory firm with deep aviation experience and a track record of transformative results for our clients across sectors." Draft 2: "DeccanJet is burning approximately ₹4 crore a month against a cash reserve of ₹28 crore (a runway of 7 months at the current rate), driven primarily by underutilized regional routes flown at off-peak frequency. The question this proposal answers is which of your 14 regional routes to cut, shrink, or reprice within 60 days, and by what approach we would reach that answer." Which draft is doing the job of a proposal's opening, and why?

Solution. Step 1: check the runway arithmetic in Draft 2, because a proposal's numbers must reconcile before anything else is judged. Runway = cash reserve ÷ monthly burn = ₹28 cr ÷ ₹4 cr/month = 7.0 months: the figure checks out exactly as stated.

Step 2: name what each draft argues. Draft 1 states a category ("aviation experience," "transformative results") that could be sent, unedited, to any airline client anywhere: it argues nothing specific to DeccanJet and could not be falsified or challenged, because it commits to nothing checkable. Draft 2 states a situation (the burn and the reserve, both numbers a client can immediately confirm or dispute), a complication stated with urgency a vague draft cannot manufacture (7 months, not "a challenging environment"), and an objective framed as a decision the client will make (which routes, within 60 days) rather than a document the firm will produce.

Step 3: connect it to the six-part structure. Draft 2 is doing, in miniature, exactly what the proposal section of this node named: situation and complication first, specific and quantified, setting up an objective the client will recognize as the actual question keeping them up at night, before a single word is spent on the firm's own credentials.

Read it. The tell is not tone or polish (Draft 1 might well be more elegantly written); it is specificity that could be checked and could be wrong. A proposal opening that cannot be disputed because it says nothing falsifiable has not yet proven the firm understood the client's actual problem, and a client reading Draft 1 has no way to tell this proposal apart from the one the losing firm sent.

Worked example 6: A capacity gap surfaces before it becomes a crisis (US, $)

An 8-week operating-model engagement for a synthetic US industrial client, Bridgeforge Industrial, has three workstreams whose combined analysis needs, per the workplan, 50 + 45 + 35 = 130 person-days. The team staffed against the SOW is two Consultants, full-time, for the full 8 weeks. Is the team big enough, and if not, what are the Engagement Manager's actual options?

Solution. Step 1: compute the person-days available. 2 Consultants × 8 weeks × 5 days/week = 80 person-days.

Step 2: compute the gap. 130 person-days needed − 80 person-days available = 50 person-days short: the team is staffed for barely 60% of the analysis the workplan itself calls for, a gap large enough that it will not be absorbed by working a little later most nights.

Step 3: name the real options, because "the team will just work harder" is not actually one of them at this size of gap. (a) Add capacity: staff an additional Analyst or Consultant for some weeks, which is itself a scope-and-fee conversation if it was not already priced into the SOW's assumptions. (b) Extend the timeline: hold the team size and push the 8 weeks out, which is a client conversation about the delivery date, not a silent slip. (c) De-scope: cut one workstream's ambition (narrower analysis, fewer exhibits) to fit the 80 person-days actually available, which must be raised with the client as explicitly as any other scope change, not quietly decided by the team alone.

Step 4: connect it to the discipline already taught. Whichever option is chosen, it is resolved the same way a scope-creep request is resolved: quantified, and put to the client as an explicit choice, rather than absorbed as unpaid overtime that erodes both the work's quality and the team's goodwill.

Read it. The value of doing this arithmetic in week one (not week six, when the team is already exhausted and behind) is that a 50-person-day gap is a straightforward, calm planning conversation before the engagement starts and a credibility-damaging crisis if it is only discovered once deadlines start slipping.

Practice set

Work each problem fully before reading its solution. All company names and figures are illustrative and internally consistent unless stated otherwise.

P1 (guided). The engagement map. List, in order, the stages of an engagement from problem statement to the deck, and name who typically owns the SOW and who typically owns a single workstream's day-one hypothesis.

Solution. Problem statement → proposal → signed SOW → workplan and workstreams (each with a day-one hypothesis) → analysis, exhibits, and QC → synthesis (the "so what") → recommendation and the deck. The SOW is owned by the Engagement Manager (drafted, negotiated, and subsequently managed against), typically signed off by the Partner/Principal on the firm's side. A single workstream's day-one hypothesis is owned by the Associate/Consultant leading that workstream, who forms and tests it with their Analyst.

P2 (guided). Fee build-up. A 6-week US engagement staffs an Engagement Manager at 0.25 FTE ($2,400/day), a Consultant at 1.0 FTE ($1,400/day), and an Analyst at 1.0 FTE ($900/day), 5 days a week. Compute the total fee.

Solution. EM: 0.25 × 6 = 1.5 person-weeks; weekly rate $2,400 × 5 = $12,000; fee = 1.5 × $12,000 = $18,000. Consultant: 1.0 × 6 = 6 person-weeks; weekly rate $1,400 × 5 = $7,000; fee = 6 × $7,000 = $42,000. Analyst: 1.0 × 6 = 6 person-weeks; weekly rate $900 × 5 = $4,500; fee = 6 × $4,500 = $27,000. Total = $18,000 + $42,000 + $27,000 = $87,000.

P3 (guided). Trust equation. Compute the trust score for an advisor rated Credibility 8, Reliability 6, Intimacy 5, Self-orientation 4.

Solution. Trust = (8 + 6 + 5) ÷ 4 = 19 ÷ 4 = 4.75.

P4. Scope-creep classification. A signed SOW's deliverables list reads: "(1) a diagnostic of current-state margin by product line; (2) a prioritized list of cost-reduction levers with sizing." Mid-engagement, the client asks the team to also benchmark the client's margins against three named competitors. Classify the request and state the disciplined next step.

Solution. Competitor benchmarking is not named in either deliverable and is not a natural sub-component of "current-state margin diagnostic" (which is about the client's own numbers): it fails the in-scope test and is a change-order candidate. The disciplined next step is for the Engagement Manager to quantify the additional person-weeks this would take, present the client with the added fee (or a trade against existing scope), and let the client decide, not to add it for free and not to refuse it outright without a conversation.

P5. Weighted average versus simple average. A retailer's three products sell: Product A: 10,000 units at ₹250; Product B: 4,000 units at ₹600; Product C: 6,000 units at ₹150. An analyst reports the "average selling price" as the simple average of the three prices. Compute the correct, revenue-weighted average selling price and the size of the analyst's error.

Solution. Total revenue = (10,000 × 250) + (4,000 × 600) + (6,000 × 150) = 25,00,000 + 24,00,000 + 9,00,000 = ₹58,00,000. Total units = 10,000 + 4,000 + 6,000 = 20,000. Weighted ASP = ₹58,00,000 ÷ 20,000 = ₹290. Simple average of the three prices = (250 + 600 + 150) ÷ 3 = ₹333.33, an overstatement of about ₹43.33, or roughly 14.9% above the correct figure. The error runs high here because Product B, the highest-priced line, is also the smallest by volume, so an unweighted average leans on it too heavily.

P6. SBI feedback. An Analyst submitted a market-sizing exhibit on Thursday morning, one day after the agreed Wednesday-evening deadline, with no advance warning it would be late. Write the feedback as Situation, Behavior, Impact.

Solution. Situation: "For Wednesday's workstream deliverable, the market-sizing exhibit." Behavior: "It arrived Thursday morning, a day after the agreed deadline, with no message beforehand that it would be late." Impact: "That meant the Wednesday mid-week QC review had nothing to check, and the whole workstream's timeline slipped by a day it may not get back before the client touchpoint." (A forward-looking ask, "next time, flag a likely delay as soon as you see it coming, even before the deadline," completes the conversation but is not itself part of the SBI structure.)

P7. The advisor's dilemma, by the numbers. A Phase 1 Indian engagement delivered ₹1,80,00,000 in annualized savings against a ₹40,00,000 fee. A proposed Phase 2 is priced at ₹65,00,000 and is projected, by the team's own honest analysis, to deliver ₹20,00,000 in net new annual savings. Compute both returns and state the ethical action.

Solution. Phase 1 return = ₹1,80,00,000 ÷ ₹40,00,000 = 4.5×. Phase 2 projected return = ₹20,00,000 ÷ ₹65,00,000 ≈ 0.31×: the client would pay more than three times what the engagement is projected to return. The ethical action is to present this finding honestly rather than the drafted Phase 2 proposal, and offer only a smaller, genuinely justified scope of work if one exists, protecting the relationship's self-orientation term even though it costs the firm the larger fee.

P8. Delegation. A workstream needs a first-cut market-sizing model built. The team has a first-week Analyst, a two-year Consultant, and the Engagement Manager. Who should build the first cut, and what does correct delegation require beyond just picking the person?

Solution. The Consultant is the best match: experienced enough to build a defensible first cut with light oversight, unlike a first-week Analyst who would need heavy hand-holding on a task this open-ended, and a poor use of the Engagement Manager's time on something that does not require their level. Correct delegation additionally requires an explicit success criterion (e.g., "a top-down and a bottom-up estimate that land within 2× of each other, with assumptions listed") and a check-in cadence (e.g., a look at the structure after day one, before too much time is sunk into a flawed approach); handing over the task alone, without either, is not yet delegation.

P9. Capacity gap. A workplan calls for four workstreams needing 40, 35, 30, and 20 person-days respectively. The staffed team is three Consultants for 7 weeks, 5 days a week. Is the team big enough, and by how much?

Solution. Total needed = 40 + 35 + 30 + 20 = 125 person-days. Available = 3 × 7 × 5 = 105 person-days. The team is short by 20 person-days: a gap to raise with the client as an explicit choice (add capacity, extend timeline, or trim a workstream) before it becomes a missed deadline.

P10. Proposal quality. Classify this proposal excerpt and explain why: "Our firm brings unmatched insight and world-class expertise to every engagement, delivering results our clients can count on." What is missing that Worked Example 5's Draft 2 had?

Solution. This is a generic credentials claim, not an argument specific to any client's situation: it names no complication, no number, no objective, and could be sent unedited to any prospective client in any industry. What is missing is exactly what made Draft 2 work: a quantified, checkable situation and complication (a real number the client would recognize as their own problem) and an objective framed as the decision the engagement will help the client make, not an adjective about the firm.

P11. Trust equation: what changed. An advisor is rated Credibility 6, Reliability 9, Intimacy 8, Self-orientation 4. Compute the trust score. Then compute the new score if self-orientation improves to 2, and state which lever moved the number more per unit of effort typically required to change it.

Solution. Current trust = (6 + 9 + 8) ÷ 4 = 23 ÷ 4 = 5.75. With self-orientation improved to 2: (6 + 9 + 8) ÷ 2 = 23 ÷ 2 = 11.5: the score exactly doubles. Lowering self-orientation is the higher-leverage lever per unit of effort: raising credibility, reliability, or intimacy by an equivalent few points typically requires months of consistently proven competence or care, while lowering self-orientation can be demonstrated in a single visible act (turning down unjustified work, admitting an error early).

P12. Spot the missing SOW component. A draft SOW lists deliverables, a timeline, and a fee schedule, but nothing else. Name two components it is missing and the specific risk each absence creates.

Solution. Missing in-scope/out-of-scope language: without it, any client request that is not explicitly excluded can be argued as implicitly included, defeating the entire scope-creep discipline. Missing a change-control clause: without it, there is no agreed process for pricing and approving a scope change, so every mid-engagement ask becomes an improvised, case-by-case negotiation instead of a repeatable, low-drama process.

P13. Fixed-fee versus time-and-materials. An engagement's scope is genuinely hard to estimate up front: the team's honest range is 250 to 550 hours at an agreed rate of $150/hour ($37,500 to $82,500), with a most-likely estimate of 400 hours ($60,000). A firm proposes a flat fixed fee of $70,000. Whose risk does each structure put where, and which would you recommend given how wide this range is?

Solution. A fixed fee of $70,000 puts the estimation risk on the firm: if the work runs to 550 hours, the firm has effectively priced its own time at a discount; if it runs to 250, the client has overpaid relative to actual effort. Time-and-materials would put the risk on the client, who pays only for hours actually used (bounded, ideally, by an agreed not-to-exceed ceiling near the top of the range, e.g. $82,500). Given how wide this range is (250–550 hours, nearly 2.2× top to bottom), T&M with a not-to-exceed ceiling is the more defensible structure: a fixed fee this early, on scope this uncertain, forces one side to guess badly, and it is usually the firm that ends up eating the difference if the fee was set anywhere near the low end to win the work.

P14. The two-person rule under pressure. A senior Consultant, under a tight deadline the night before a client meeting, says there is no time for a second reviewer on a number they are confident is correct. What is the disciplined response, and why does seniority not change the answer?

Solution. The disciplined response is that the two-person rule still applies: find a second reviewer even if it means a shorter final review or a slightly later night, because the rule exists precisely for the moments it is most tempting to skip it. Seniority does not change the answer because confidence is not evidence of correctness (the same lesson the AI-Augmented Analyst branch teaches about a model's fluent output): the most expensive errors are disproportionately made by the most confident person in the room, exactly because their confidence suppressed their own re-check.

P15. Conflict of interest. A firm is approached by a new prospective client to advise on entering a market segment where the firm already advises that prospect's direct competitor on the identical question. Name the required first step and the two ways this is typically resolved.

Solution. The required first step is disclosure to both clients: the firm may not decide silently on its own that the conflict is manageable. It is typically resolved by (1) an ethical wall (separate teams with no shared access to either client's confidential information) if both clients knowingly consent to proceed on that basis, or (2) declining one of the two engagements if a wall cannot credibly protect confidentiality or either client will not consent.

P16. Trust-equation diagnosis from a client complaint. A client says, "I respect your team's expertise, but I don't feel like you really understand our business, and I'm not sure I'd tell you if something were actually going wrong." Which trust-equation variable is deficient, and, given current ratings of Credibility 8, Reliability 8, Intimacy 2, Self-orientation 3, compute the current trust score and the score if intimacy improved to 7.

Solution. The complaint is a direct description of low intimacy: the client does not feel safe sharing something risky, regardless of how competent or reliable the team is rated. Current trust = (8 + 8 + 2) ÷ 3 = 18 ÷ 3 = 6.0. With intimacy improved to 7: (8 + 8 + 7) ÷ 3 = 23 ÷ 3 ≈ 7.67, a meaningful jump achieved without any change to technical competence, underscoring that the fix for this specific complaint is time spent understanding the client's actual business and vulnerabilities, not another impressive exhibit.

Applied mini-project

Run one engagement's opening, on paper, end to end.

You are the Engagement Manager for a synthetic 8-week engagement of your own invention (choose an India or US client and industry; keep every figure synthetic and internally consistent). Produce, in order:

1. A one-page proposal using the six-part structure (situation, complication, objective, approach with 3 workstreams and a one-line day-one hypothesis each, team, timeline and fee): situation and complication must include at least one real, checkable number (a burn rate, a margin, a growth rate) that your later fee build-up will be consistent with in scale.
2. A SOW skeleton for the same engagement: deliverables, in-scope/out-of-scope bullets (at least 2 of each), assumptions/dependencies, a change-control clause, and governance (sign-off and meeting cadence).
3. A fee build-up by role (FTE × weeks × day rate × days/week for each of at least 3 roles), reconciled to a total, and a one-line justification for your chosen commercial structure (fixed-fee, T&M, or value-based) given how well-defined your scope actually is.
4. One scope-creep scenario, invented, run through the in/out-of-scope test against your own SOW, with a quantified change-order estimate if it fails the test.
5. A trust-equation self-assessment: rate a synthetic advisor (yourself, in character) on Credibility/Reliability/Intimacy/Self-orientation for this engagement at the 3-week mark, compute the trust score, and write two sentences on which variable you would move first and how, with the arithmetic showing the effect of that specific change.

Deliverable and self-check. One short document (roughly one to two pages) containing all five numbered items above. Score yourself 2 points each for: (a) the proposal's situation/complication contains a real, checkable number, not a generic claim; (b) the SOW contains all six required components, not just deliverables and fee; (c) the fee build-up arithmetic actually reconciles (recompute it, don't eyeball it); (d) the scope-creep scenario is correctly classified against your own SOW's actual deliverables list, not against what you now wish you had written; (e) the trust-equation arithmetic is computed correctly both before and after the proposed change. Pass at 8/10 with no zero on any item.

This node maps no Case-Engine case or drill set of its own in the galaxy plan: its gate is the mastery quiz alone (below). The reps for this material live in this mini-project now, and again, at far greater depth and under time, in the Partner Capstone once you have completed the branch.

Reading & resources

Everything below is an anchor to learn from and verify against, not a substitute for the judgment this node asks you to build. Free unless marked; level tags run beginner → advanced.

The trusted-advisor relationship (the spine of this node):

• David Maister, Charles Green & Robert Galford, The Trusted Advisor: the trust equation in full, with the book's own worked illustrations of credibility, reliability, intimacy, and self-orientation. [Paid] [Beginner–Advanced]
• David Maister, Managing the Professional Service Firm: the economics of a PSF (leverage, team structure), and why professional norms (not just individual talent) are what make an advisory firm trustworthy at scale. [Paid] [Advanced]
• Peter Block, Flawless Consulting: the consultant–client "contract" (in the relationship sense, not the legal one), and the craft of a genuinely flawless meeting. [Paid] [Beginner–Advanced]

The engagement, the team, and the craft of the work:

• Ethan Rasiel, The McKinsey Way and The McKinsey Mind: accessible, anecdote-rich treatment of how an engagement team actually runs day to day, including staffing, feedback, and client-management moments. [Paid] [Beginner]
• Barbara Minto, The Pyramid Principle: the full mechanics of the answer-first synthesis this node only names as a hand-off; read it when you reach CN1.03 or M10.03, not required here. [Paid] [Advanced]
• Conn & McLean, Bulletproof Problem Solving: the 7-step problem-solving cycle CN0.01 introduces and this node's engagement map assumes; a natural companion read. [Paid] [Beginner–Advanced]

Feedback and delegation (free, practitioner-oriented):

• The Center for Creative Leadership's public write-ups on the SBI (Situation–Behavior–Impact) feedback model: the short, practical original source for the model taught in this node. [Free] [Beginner]
• Harvard Business Review: across delegation, feedback, and professional-services management, HBR is the standing anchor for this node's team-management content; search its archive by topic as each comes up. [Mixed free/paid] [Beginner–Advanced]

Scope, fees, and change control (free, practice-oriented):

• Public project- and program-management bodies of knowledge (the kind of material behind a PMP-style curriculum) cover scope-statement and change-control discipline in a form directly portable to a consulting SOW, even though this node teaches the consulting-specific version of it. [Free overviews / paid full certification] [Beginner]
• Management Consultancies Association (MCA)-style professional codes of conduct: publicly available codes of ethics/independence from consulting industry bodies are a useful, free anchor for the independence and conflict-of-interest discipline taught here. [Free] [Beginner–Intermediate]

Do this, not just read: find a real, publicly available consulting-firm case study or a client testimonial write-up, and try to reverse-engineer its likely original proposal: what situation and complication would have opened it, what the fee structure was probably built on given the described scope, and where a change-order conversation most likely happened but was never mentioned. Fifteen minutes of doing that will teach you more about how the paperwork behind a polished case study actually worked than another chapter of theory.

Flashcards

This module's flashcards and mastery quiz are wired into the app: see the node's Quiz and Reviews.

Mastery check

Two parallel forms, 12 items each. Closed book, calculator allowed, ~25 minutes per form. Numeric answers within ±2% score as correct. Pass threshold: ≥ 85%, 11 of 12 on a 12-item form. This node maps no Case-Engine case or drill set of its own: its row in the galaxy plan gates on the mastery quiz alone, and passing either form both certifies the understanding objective and fires the White Belt.

Form A

A1 (MCQ). Which of the following best orders the consulting engagement's core arc, from a signed proposal onward? (a) team staffing → fee negotiation → the deck → SOW (b) problem statement & workplan → workstreams & hypotheses → analysis & QC → the "so what" & the deck (c) the deck → hypotheses → problem statement → QC (d) SOW → the deck → problem statement → workstreams

A2 (MCQ). A consulting proposal, as distinct from the SOW that follows it, primarily exists to: (a) legally bind the fee and deliverables in enforceable detail (b) argue why this firm, this team, and this approach are the right choice: situation, complication, objective, approach, team, timeline, fee (c) list every in-scope and out-of-scope item down to the sentence (d) log every change order raised during delivery

A3 (numeric). A Consultant works 1.0 FTE for 6 weeks at a day rate of ₹55,000, 5 days a week. What is the fee for this role?

A4 (MCQ). A client casually asks, mid-engagement, for one more analytical cut not in the SOW's deliverables list. The disciplined response is to: (a) do it for free to keep the relationship warm (b) refuse outright, citing the SOW (c) run it through the in/out-of-scope test, and if out, quantify it and raise a change order rather than silently absorbing or refusing it (d) escalate immediately to the partner without discussing it with the client

A5 (MCQ). In a typical engagement team, delegation is done well when a task is: (a) handed off with no further contact until the deadline (b) matched to a level that can do it with appropriate (not excessive) oversight, with explicit success criteria and a defined check-in cadence (c) always given to the most senior available person, regardless of workload (d) checked by the delegator at every single step

A6 (numeric). A workplan needs 70 person-days of analysis. The staffed team is 2 Analysts for 5 weeks, 5 days a week. How many person-days short (or over) is the team?

A7 (MCQ). The "two-person rule" in engagement QC means: (a) only two people are ever allowed to see client-facing exhibits (b) any number or claim that will reach the client passes a second reviewer's check before it ships, regardless of how senior or confident the author is (c) the team must always be exactly two people (d) the client and the partner must both sign every exhibit

A8 (MCQ). In the SBI feedback model, the three components are: (a) Situation, Behavior, Impact (b) Strengths, Blind spots, Improvements (c) Scope, Budget, Implementation (d) Structure, Business judgment, Intuition

A9 (numeric). An advisor is rated Credibility 5, Reliability 5, Intimacy 5, Self-orientation 5. Compute the trust score.

A10 (MCQ). In the trust equation, "intimacy" refers to: (a) personal friendship with the client (b) the emotional safety that lets a client share sensitive or risky information without fear of exposure or judgment (c) the number of years you have known the client (d) how often you socialize outside of work

A11 (numeric). Three US regions contribute: Region X: revenue $30M, margin 20%; Region Y: revenue $50M, margin 35%; Region Z: revenue $20M, margin 10%. Compute the correct revenue-weighted blended margin.

A12 (short). A team's own honest analysis shows a proposed follow-on phase would cost the client more than it would save. State the ethical action and name which trust-equation variable it protects.

Form A key. A1: b. The engagement's arc runs problem statement/workplan → workstreams/hypotheses → analysis/QC → the "so what" → the deck. A2: b. The proposal argues for the work; it does not yet commit like a SOW. A3: 1.0 × 6 = 6 person-weeks; weekly rate = ₹55,000 × 5 = ₹2,75,000; fee = 6 × ₹2,75,000 = ₹16,50,000. A4: c. The in/out-of-scope test and a transparent change order, never silent absorption or a flat refusal. A5: b. Task matched to level, explicit success criteria, a real check-in cadence. A6: available = 2 × 5 × 5 = 50 person-days; needed = 70; the team is 20 person-days short. A7: b. The rule applies regardless of seniority or confidence. A8: a. Situation, Behavior, Impact. A9: (5+5+5) ÷ 5 = 15 ÷ 5 = 3.0. A10: b. Emotional safety, not friendliness or social closeness. A11: total revenue = 30+50+20 = 100; weighted margin = (30×0.20 + 50×0.35 + 20×0.10) ÷ 100 = (6 + 17.5 + 2) ÷ 100 = 25.5 ÷ 100 = 25.5% (note this is above the simple average of 21.67%, the opposite direction from Worked Example 3, because here the largest region also carries the highest margin). A12: The ethical action is to tell the client honestly that the follow-on phase is not justified, even though it costs the firm the fee, and only propose a smaller, right-sized scope if genuinely warranted; this protects (lowers) self-orientation, the trust equation's denominator, and is one of the highest-leverage trust-building acts available precisely because it works against the firm's own short-term interest.

Form B

B1 (MCQ). A signed SOW is followed immediately by: (a) the final client presentation (b) standing up the workplan and workstreams, each carrying its own day-one hypothesis (c) the fee negotiation (d) the trust-equation self-assessment

B2 (MCQ). The Statement of Work (SOW), as distinct from the proposal that preceded it, primarily exists to: (a) persuade the client the firm deserves to win the work (b) serve as the contract-grade reference the whole engagement is managed against: deliverables, in/out-of-scope, assumptions, acceptance criteria, change control (c) replace the need for a workplan (d) score the team's trust equation

B3 (numeric). An Analyst works 0.5 FTE for 8 weeks at a day rate of $800, 5 days a week. What is the fee for this role?

B4 (MCQ). Scope creep is best defined as: (a) any request a client makes during an engagement (b) small, individually reasonable-seeming asks outside the SOW that are absorbed without a change order, silently eroding margin and the deliverable's focus (c) a large, formally negotiated change to the SOW (d) a delay caused by the client's data not being ready

B5 (MCQ). The typical engagement-team hierarchy runs, from the client relationship down to the core analysis, as: (a) Analyst → Consultant → Engagement Manager → Partner (b) Partner/Principal → Engagement/Project Manager → Associate/Consultant → Analyst (c) Engagement Manager → Partner → Analyst → Consultant (d) there is no standard hierarchy at all

B6 (numeric). A workplan needs 45 person-days of analysis. The staffed team is 3 Analysts for 4 weeks, 5 days a week. How many person-days short (or over) is the team?

B7 (MCQ). The "pyramid check" during QC specifically verifies that: (a) the exhibit's font and formatting match the firm's template (b) the top-line message (the "so what") is actually supported by the evidence assembled beneath it, not just individually correct facts, but correctly built into the argument (c) every slide has exactly three bullet points (d) the fee matches the SOW

B8 (MCQ). Effective engagement feedback is best characterized as: (a) saved up for a single end-of-engagement review so it lands with more weight (b) real-time, specific about an observed behavior and its impact, and flows both up and down the team, not only from senior to junior (c) always delivered in writing so there is a paper trail (d) avoided during the engagement itself to protect team morale

B9 (numeric). An advisor is rated Credibility 8, Reliability 8, Intimacy 8, Self-orientation 2. Compute the trust score.

B10 (MCQ). In the trust equation, "self-orientation" sits in the denominator because: (a) it is the least important of the four variables (b) as it rises, it divides down all the trust the other three variables built, however strong they are: a multiplier on the whole relationship, not just one more ingredient (c) it can never be measured (d) only junior consultants have self-orientation

B11 (numeric). Three business segments contribute: Segment 1: revenue $60M, margin 12%; Segment 2: revenue $25M, margin 40%; Segment 3: revenue $15M, margin 22%. Compute the correct revenue-weighted blended margin.

B12 (short). A firm is approached to advise a client on entering a market where the firm already advises that client's direct competitor on the identical question. Name the ethical discipline this triggers and the two ways it is typically resolved.

Form B key. B1: b. The workplan and workstreams, each with a day-one hypothesis, stand up immediately after the SOW is signed. B2: b. The SOW is the contract-grade reference; it does not persuade, it commits. B3: 0.5 × 8 = 4 person-weeks; weekly rate = $800 × 5 = $4,000; fee = 4 × $4,000 = $16,000. B4: b. Small, individually reasonable asks absorbed without a change order. B5: b. Partner/Principal → Engagement/Project Manager → Associate/Consultant → Analyst. B6: available = 3 × 4 × 5 = 60 person-days; needed = 45; the team has a 15 person-day surplus (not a gap), worth flagging to the EM as a staffing question, not just a risk to hedge against. B7: b. The pyramid check verifies the argument's assembly, not just each fact's individual correctness. B8: b. Real-time, specific, bidirectional. B9: (8+8+8) ÷ 2 = 24 ÷ 2 = 12.0. B10: b. It divides the whole fraction, making it a multiplier rather than an additive ingredient. B11: total revenue = 60+25+15 = 100; weighted margin = (60×0.12 + 25×0.40 + 15×0.22) ÷ 100 = (7.2 + 10 + 3.3) ÷ 100 = 20.5 ÷ 100 = 20.5% (below the simple average of 24.67%, a third distinct direction from A11 and Worked Example 3, underscoring that the direction of the error depends entirely on the data, never assume it). B12: This is a conflict of interest, which must be disclosed to both clients rather than self-adjudicated silently. It is typically resolved by (1) an ethical wall (separate, non-overlapping teams with no shared access to either client's confidential information) if both clients consent to proceed on that basis, or (2) declining one of the two engagements if a wall cannot credibly protect confidentiality or either client is unwilling to proceed.

Teach it back & journal

Feynman prompt. Explain to a friend who thinks "consulting is just smart people giving smart advice" why a technically brilliant recommendation can still fail to land, using the sentence "the client is not buying the analysis, the client is buying the confidence to act on it." Then walk them through the trust equation by hand with two advisors of your own invention, one more "impressive" and one with lower self-orientation, and show, with real arithmetic and not just an assertion, why the equation says to trust the second one more. If your friend can, unprompted, predict that the single highest-trust-building thing an advisor can do is tell a client they need less work, not more, you have installed the mechanism this node exists to teach.

Journal prompt. Think of a real service relationship in your own life (a doctor, a lawyer, a financial advisor, even a mechanic) where you either trusted them completely or never quite did, and be specific about which. Rate them, as honestly as you can, on Credibility, Reliability, Intimacy, and Self-orientation, compute their trust score, and identify which single variable was actually driving your feeling, not the one you would have guessed before doing the arithmetic. Then write a paragraph on this: which of this node's disciplines (a precise scope boundary, a transparent change order, real-time specific feedback, or a visible act of low self-orientation) would most change how a client experiences working with you, and what is the one concrete thing you will do differently the next time you are the one being trusted with someone else's problem?

This module's flashcards and mastery quiz are wired into the app: see the node's Quiz and Reviews.