The Readiness Audit

Once a leader believes the argument of this book, the instinct is to commission the big version of what comes next. You can watch it happen in a room. Someone at the offsite writes "we need to fix this" on the whiteboard and underlines it twice, and within twenty minutes the conversation has become a project. Hire a firm. Run a proper assessment. Six weeks, forty interviews, a maturity model with five levels and a colour-coded heat map, a workstream for each colour, a deck presented back to the board with a roadmap and a number. It feels responsible. It looks like leadership. It is, most of the time, an elaborate way of not starting.

There is a managing partner I can picture, a composite of several, who nearly did exactly this. She had felt the gap between the deal and the done for years, and she had finally read it described, and she wanted to know whether the diagnosis fit her firm or only firms in a book. The consultancy she called quoted her a sensible six-figure engagement to find out. She was about to sign when she asked herself a smaller question. Before I pay someone to spend six weeks telling me where I stand, could I not spend one honest afternoon finding out myself?

She could. So can you. The uncomfortable truth about a readiness audit is that the most valuable version of it is not the expensive one. It is a short, sharp, self-administered interrogation that any leader can run on their own firm, with their own real examples, in an afternoon, behind a closed door. It does not produce a report. It produces a decision. And the questions it asks are not a quiz to be scored. They are the discipline of this whole book, compressed into a form you can hold.

The turn: not a grade, a starting point

The wrong question to ask about your firm is how mature it is. Maturity models are graded reassurance. They sort you into a level, place you comfortably above the laggards and safely below the leaders, and hand you a thirty-item improvement plan that no one will ever finish, because a plan with thirty priorities has none. You file the report. You feel briefly diagnosed. Nothing changes.

The right question is narrower and far more useful. Of all the places where your firm quietly loses its decisions, which one is costing you the most right now, and can you find it before you go home? You are not trying to measure everything. You are trying to locate the single worst seam, the place where the mechanisms this book has described are biting hardest, so that when you begin, you begin where it matters and not where it is easiest.

Not a grade. A starting point.

That is what this chapter is. Five diagnostic areas, each drawn from an idea you have already met in these pages, each posed as questions you ask about specific, real events in your own firm rather than in the abstract. For each one I will tell you what a good answer sounds like and what a bad answer sounds like, and, more importantly, what a bad answer is telling you about the machinery underneath. Do not score them. Read them. At the end you will not have a number. You will have a shortlist of one.

A warning before you start, because it determines whether the exercise works at all. Use real examples, named and specific, not representative ones. "We generally document our handovers" is not an answer. "The Meridian account, handed over in March, from Priya to the delivery team" is an answer. The whole value of this audit is that it refuses to let you answer in the general, where every firm looks competent, and forces you into the particular, where the truth lives. Every organisation sounds well run in the abstract. Competitive advantage only reveals itself in specific decisions.

Area one: can you see your decisions?

This is the decisionless-stack test, and it is the one to run first, because it calibrates everything after it.

Pick one consequential decision from the last quarter. A real one, with money or risk attached. You held a price when you could have walked. You took on a client despite the reservations someone voiced in the review. Choose the decision, name it, and then try, from wherever you sit, to reconstruct four things about it. Who actually decided. On what evidence. Having considered what alternatives. And, if the outcome is already in, whether anyone has compared the two.

Set yourself a limit. If you cannot reconstruct it in the time it takes to make and drink a coffee, that is not a scheduling problem. That is the finding.

A good answer sounds like this. You can name the owner without hedging, and everyone who was in the room names the same person. You can point to the evidence the decision actually rested on, not evidence assembled afterward to justify it. You can list the alternatives that were genuinely on the table and say why they were set aside. And when you ask two other people who were there, their accounts agree with yours and with each other. The decision, in other words, exists as an object outside any single person's memory. It can be examined. It can be learnt from.

A bad answer takes several forms, and you should recognise them because each points at a different failure. The first is disagreement: you ask three people who was in the room and you get three different stories, which means there was never a decision, only a drift that everyone now narrates to their own advantage. The second is evaporation: the person who knew has left, and what they weighed left with them, so you are reconstructing a choice from its wreckage exactly as the delivery director did in Chapter 4, walking nine systems and finding everything except the one thing she came for. The third, and the most quietly damning, is the notes field: you find the decision recorded as "agreed to hold price," a conclusion with no reasoning, no alternatives, no owner, no fear, the stack's confession that the decision mattered and its shrug at capturing it.

Whatever form your bad answer takes, do not rush past it. This is your baseline. It is the honest measure of whether your firm can see its own decisions at all, and it sets a ceiling on everything else in this audit, because a firm that cannot reconstruct one deliberately chosen decision from last quarter certainly cannot reconstruct the thousand small ones it did not think to mark.

What this usually means: Decision capture does not exist.

Area two: does your firm remember?

This is the amnesia test, and it asks not what your systems hold but what walks out of the building when a person does.

Name a key person who left in the last year. Not a junior. Someone who carried real context: an account lead, a senior deliverer, a pursuit lead, the person three clients asked for by name. Now ask, plainly, what left with them. Not their tasks, which were reassigned in a week. Their judgement. The reasons behind the calls they made, the map of which stakeholder went quiet under pressure, the knowledge of why the Meridian renewal had to be handled gently, the hundred small understandings that made them good. Where is that now? Can someone who never met them use it?

Then run the mirror image of the same test, because memory has two faces, losing it and inheriting it. Take a person who joined a client-facing role in the last few months and ask them, honestly, how they came up to speed. Did they inherit context or reconstruct it? Did they read a living record of what had been decided and why, or did they spend their first six weeks interviewing whoever would spare an hour, rebuilding from fragments a picture that already existed in a mind that had left?

A good answer is one where the departure was survivable in substance and not just in staffing. The leaver's reasoning is still usable by someone who never shook their hand, because it was captured as it was made, not stored in their head. And the joiner inherited rather than excavated. When this is true, attrition is a human loss and a logistical task, but it is not an amnesia event.

A bad answer is the one most firms give, and it is worth hearing how ordinary it sounds. "We are honestly still finding out what she knew." "Two of her accounts wobbled after she left, and we are not entirely sure why." "The new joiner is bright, he is getting there, it just takes time to pick up the context." That last sentence is the tell. The context did not need to be picked up. It existed. It simply existed nowhere a person could inherit it, so every departure forces a reconstruction and every arrival pays for it again. This is the mechanism from Chapter 5 rendered as a line on your own payroll.

A firm whose memory lives in its people does not have a memory. It has a hostage arrangement with its staff, renewed monthly, and every resignation is a ransom.

What this usually means: Your memory is person-dependent.

Area three: is your loop closed?

This is the seam test, and of the five it is the one most likely to be your answer, because the seam between selling and delivering is where people businesses lose the most and see the least. Run it deliberately.

Take three engagements your firm delivered in the last year. Pick them for range, not for comfort: one that went well, one that went badly, one that was merely ordinary. For each, ask two questions that run in opposite directions across the seam.

The first question runs forward, from the sale into the delivery. Did the delivery team inherit the reasoning behind the sale, or only the contract? Go and ask the person who led the delivery, directly. Did you know why it was priced the way it was? Did you know which parts of the scope were firm and which were a polite fiction agreed in the room to get to signature? Did you know what promise, made at four o'clock on a Thursday, actually sealed it? If the delivery lead knew these things, the seam carried the decision. If the delivery lead started by reconstructing what the client "really wanted" from a document that recorded what was agreed and almost nothing of why, the seam carried only the record, and you have found the exact gap the introduction to this book opens inside.

The second question runs backward, from the delivery into the next sale. Did the outcome of the work ever reach the people selling the next one? When your firm next pitched something similar, did the person shaping the proposal know how this engagement had actually turned out: the true margin, not the sold margin; the scope that crept; the escalations that ate the plan? Or did they build the new proposal on the same optimistic assumptions that this one had just quietly disproved, because the lesson of the delivery never travelled back to the mouth of the loop?

A good answer is a loop that closes in both directions. Delivery leads can state the commercial logic of the sale they inherited, in their own words, without being coached. And sellers can cite the real outcome of the last comparable engagement, including the parts that hurt, and can show you where those outcomes changed how the next thing was priced or scoped or staffed.

A bad answer is a loop broken at the seam, which is to say the ordinary condition of the industry. Delivery inherited a contract and rebuilt the context from meetings. The next proposal repeated the pricing that lost money the last time, because the person who wrote it never learnt that it had. If this is what you find across your three engagements, mark it heavily. A broken sell-to-deliver seam is not one problem among five. It is the seam where the workforce starts, where the promise is made that all the later work must honour, and a firm that loses its decisions here loses them at the source.

What this usually means: Your sell-to-deliver seam is broken.

Area four: can you answer the three confidences, today, with evidence?

This is the confidence test, and the word that matters in it is today. Almost every firm has the artefacts. The question is whether the artefacts can be trusted, and whether they are connected to each other, or whether they are three separate fictions maintained by three different people who never reconcile.

Do not ask whether you have a forecast, a delivery view, and a P&L. You do. Ask the harder version of each.

For Win Confidence, ask: could you say right now, without a week of preparation, which deals in your pipeline are real and which are hope, and defend the distinction with evidence rather than the optimism of the person whose bonus depends on the answer? A good answer rests on what the client has actually done and committed, not on the salesperson's mood. A bad answer is the forecast everyone privately discounts, the one where the leader applies a silent haircut because experience has taught them the number is brave. If you are mentally reducing your own forecast as you read this, you have your answer.

For Delivery Confidence, ask: could you name, today, which of your live engagements are quietly slipping toward a loss, while there is still time to act, rather than discovering it at month-end when the loss is booked? A good answer is a current, evidence-backed reading of the live state of the work, sensitive enough to catch a programme going sideways in the week it starts to go, not the quarter it finishes. A bad answer is a wall of green status decks that turn red only after the damage is done, because status is self-reported by the people least able to say they are in trouble, and the true state of delivery lives in a delivery lead's stomach and nowhere you can see.

Sit with this one longest, because it is the only one of the three you have no existing instrument for. You already know what a forecast is, and Win Confidence only asks you to make it honest. You already know what a P&L is, and Commercial Confidence only asks you to make it live. There is no equivalent standing discipline for delivery. Most firms have never named the thing at all. They run engagement dashboards that count hours booked and milestones closed, which is motion and not health, and a programme can show green on every box on Friday and be structurally lost by the following Wednesday. Delivery Confidence is a reading of a different kind: whether the work will actually land where it was sold to land, formed early enough that the answer can still be changed. It is the one instrument nobody hands you, and it is the one your margin most quietly depends on. If you build only one thing out of this whole audit, there is a fair case that you build this.

For Commercial Confidence, ask: does your forecast of profit connect to the live state of delivery, or is it last month's finance plus hope? A good answer is a view of the money that moves when the work moves, so a failing engagement shows up in the commercial picture before the accounts catch it. A bad answer is a P&L that is accurate, trusted, and entirely retrospective, a beautifully kept record of a decision it is now too late to change.

Then ask the question that matters more than any of the three on its own, because it is the one that turns three numbers into Business Confidence. Are they connected? When an engagement starts slipping, does that show up anywhere in your commercial view before finance closes the month? When delivery is straining to honour what sales promised, does anyone selling the next deal feel it? If the three confidences are current but disconnected, each true in its own silo and never reconciled with the others, you do not have Business Confidence. You have a forecast, a status deck, a P&L, and, as this book has said before and will say again, a bad feeling.

What this usually means: Your Business Confidence is disconnected.

Area five: where does your margin actually leak?

This is the leak test, and it is the most concrete of the five, which makes it the best at ending arguments. Run it exactly as written.

Name the last three engagements that lost margin. Not lost money outright, necessarily; the ones that came in materially below the margin you sold them at. Name them now, from memory, before you open a system. Then, for each, answer one question: which decisions drained it?

Notice that this is not the question "what happened." What happened is easy and useless: it overran, it went long, the client was difficult. The audit question is which decisions drained the margin, and it wants specifics of the kind Chapter 14 catalogued. The scope that was absorbed instead of raised as a change request, because it was easier to say yes in the moment. The wrong grade of person staffed on the work, because they were the ones free. The escalation handled three weeks late, because no one judged it urgent until it was a crisis. The delivery heroics that held a broken plan together and hid, until the person doing the heroics burnt out, that the plan was broken at all.

A good answer sounds almost anticlimactic. You can name the three engagements without looking. For each, you can point to two or three specific decisions that drained the margin, with an owner and a moment attached to each. And, tellingly, the three engagements did not leak in the same three ways, because your firm noticed the first leak and closed it, so the second and third failed differently. That last detail is the real signal. It means the loop from Chapter 4 is intact: decision, outcome, revised judgement, a firm that actually learns.

A bad answer comes in two grades, and both matter. The milder grade is that you cannot name the three engagements without pulling a report, and when you do pull it, the explanation attached to each loss is "delivery overran" or "scope grew," with no decisions named, no owners, no moments, just the residue of choices nobody kept. The margin leaked and the firm recorded the size of the puddle but not the crack it came through. The worse grade is when you can name the decisions and they are the same decisions every time: the same underpricing, the same over-servicing, the same late escalation, engagement after engagement, year after year. That is not a leak. That is a designed outflow the firm has never chosen to close, because the loss shows up in finance where it looks like bad luck, and never next to the decision where it would look like a pattern.

You cannot fix a leak you can only measure. You can only fix a leak you can name, and to name it you need the decision that opened it, not the invoice that recorded the cost.

What this usually means: Margin is being measured, not managed.

Reading the five: finding the one place to start

You now have five honest answers about your own firm, which is more than most leaders get from a six-week assessment. Resist, hard, the thing you will now want to do, which is to write a plan that addresses all five. That plan is the thirty-item report that gathers dust. The whole design of this audit is to prevent it.

Read across the five instead, and one will be worse than the others. It usually is not close. There will be an area where the answers were most uncomfortable, where the money or the risk visibly concentrates, where the mechanism this book describes is not a theory but a wound you can point to. That is your seam. Not all five. One. The worst-performing seam is where you begin, and beginning in one place rather than everywhere is not a compromise forced on you by limited resources. It is the correct strategy, and the next chapter is about why.

Here is the part that makes the audit more than an audit. Choosing that one place is itself a decision, and you should treat it as your first decision object. Write it down properly, the way this book has asked all along. What you are choosing (to start at, say, the sell-to-deliver seam). Who owns it (you, or a named person, not "the leadership team"). On what evidence (this audit, these three engagements, that departure, those unreconstructable choices). Against what alternatives (the other four areas, and why they can wait). And leave room for the outcome, so that in six months you can come back and mark your own working. If you do that, you have not merely diagnosed your firm. You have practised the discipline in miniature, on the most important decision in the whole programme, which is where to begin. The audit's real output was never a report. It was one well-made decision, captured. You have already started.

The audit on one page

Everything above this line is the reasoning. Here is the audit itself, stripped to something you can print, photograph, or put in front of a leadership team and work through behind a closed door in an afternoon. It is deliberately not a scorecard, and the boxes are not there to be totalled. They are there to force a specific answer where a general one would hide. Fill it in with real names and real engagements. Let the last block, the worst seam, be the one decision you carry into the next chapter.

                   DECISION READINESS AUDIT
             One firm. One afternoon. One decision.

 AREA 1   CAN YOU SEE YOUR DECISIONS?
   Reconstruct one real decision from last quarter: its
   owner, its evidence, its alternatives, its outcome.
   [ ] Fully     [ ] Partially     [ ] No
   Decision tested: ____________________________________
   What would not reconstruct: _________________________

 AREA 2   DOES YOUR FIRM REMEMBER?
   When your last senior person left, what left with them?
   [ ] Captured     [ ] Partly     [ ] Walked out the door
   Who left: ______________  What we lost: _____________

 AREA 3   IS YOUR LOOP CLOSED?
   Forward, sale into delivery:   [ ] Reasoning inherited
                                  [ ] Contract only
   Backward, delivery into sale:  [ ] Outcome learned
                                  [ ] Repeated the mistake
   Engagement tested: __________________________________

 AREA 4   THE THREE CONFIDENCES, TODAY, WITH EVIDENCE?
   Win:         [ ] Evidence      [ ] Optimism
   Delivery:    [ ] Live reading  [ ] Green decks
   Commercial:  [ ] Moves live    [ ] Retrospective
   Connected?   [ ] Yes           [ ] Three separate fictions

 AREA 5   WHERE DOES YOUR MARGIN LEAK?
   Three engagements that came in under the margin you
   sold, and the decision that drained each.
   1. ______________  decision: _______________________
   2. ______________  decision: _______________________
   3. ______________  decision: _______________________

 ------------------------------------------------------------
 MY WORST SEAM IS:  ________________________________________

 Why this one, not the other four: _________________________

 Owner: ___________________   Decision date: ______________

 Review on: _______________   Outcome: ____________________
 ------------------------------------------------------------
       Not a score. Your first decision object.

The counterpoint: audits produce reports that nobody acts on

The strongest objection to a chapter like this is one every experienced leader has earned the right to make, and it should be answered without flinching. Audits do not work. You have seen them. The consultancy comes in, the interviews are held, the maturity model is applied, and out comes a handsome document with a heat map and a list of recommendations, ranked, costed, and doomed. It is circulated. It is admired. It is filed. Six months later nothing has changed except the invoice, and the honest lesson a scarred executive takes from the experience is that diagnosis is a genre of procrastination.

The objection is correct about the audits it describes, and the reason those audits fail is precisely why this one is built differently. A conventional assessment fails because its output is a comprehensive programme. It surveys everything, grades everything, and recommends improving everything, and a recommendation to improve everything is a recommendation to do nothing, because it hands a busy firm thirty priorities and no first move.

This audit is designed to produce the opposite. Its entire output is a single decision: where to begin. It does not grade you across a matrix; it points you at one seam. It does not hand you thirty items; it hands you one, chosen on evidence, small enough to act on this quarter. The dust gathers on comprehensiveness. It does not gather on a single decision with an owner and a date.

The bridge

So you know where you stand, and you know where to start. You have found the worst seam, and you have made your first real decision about it, and captured it. What you have not done is build anything, and the temptation now is to swing from too small to too large, to take the one seam you have identified and wrap a two-year transformation around it, the kind of programme that bets the firm on getting the decision layer right all at once.

You do not have to, and you should not. The next chapter is about how a real leader, with a real firm and no appetite for staking everything on a single grand rebuild, actually builds a decision layer: at one seam at a time, starting with the very one you have just found, with a first move sized to that specific gap rather than a generic programme, making capture a by-product of the work rather than a second job, earning trust in the record before automating anything, and coordinating before you automate. You have located the wound. Now we turn to how you close it, deliberately, incrementally, and without betting the firm to do it.