01 · Overview

The finding

The largest block of partner time we could measure is the one thing the firm does not track, does not bill, and cannot cost.

Six of your eight partners walked us through their own work, separately, with no shared brief and no sight of each other's answers. All six described the same sequence. Peter then described 22 face to face meetings inside ten weeks, at two and a half to three and a half hours each, and told us nothing tracks any of it. Asked what a missed follow-up costs, he said “I don't know”, and then “I don't think anybody does”. That is not evasion. It is the sound a process makes when nothing has ever measured it.

6of 8

partners interviewed, each one separately

2h 13m

total partner time, link opened to finished

142

answers given, every one on the record

7 786

words in your partners' own voices

days of consultant time on site

Draft for review Trevor Murgatroyd and Gerhard Albertyn still to complete Nothing here is inferred

The signal board

Eight things came up. Ranked by how many of the six said each one.

The cohort

Six of eight partners are in.

Every finding above is counted out of the six who have completed, never out of eight.

DLDave LakeFoundervoice 19m

PSPeter van den SteenFounder and Directorvoice 17m

MLMartin LiebenbergPartner and Directorvoice 14m

DCDenis ChifunyisePartnervoice 11m

ACAshleigh CohenPartnertyped

JCJosh CunliffePartnertyped

TMTrevor MurgatroydFounder and Directorinvited

GAGerhard AlbertynPartnerinvited

Josh's voice session never picked up his microphone. That was a fault on our side, not his. He completed the whole interview in writing and it is used here on exactly the same footing.

Where the work lands

Six things the work produces. Five land on a single point, and the sixth lands nowhere.

Nobody was asked about concentration, cover or key-person risk. Every line below came out of a partner describing their own week. The map is the finding: the one thing that would tell the firm about the other five is the one thing nothing holds.

Each line, in the words it came from.

“The pen was held by Dave Lake, who is the master of putting together reports, and his linguistic skills are supreme.”

Peter van den Steen, asked who actually puts the review together

“I do”

Ashleigh Cohen, asked who signs off the reconciled model

“I will, the business rescue practitioner.”

Martin Liebenberg, asked who decides to abandon, liquidate or deregister

“Ultimately, we (the BRPs) will have the final say on what goes into a response or on the strategy, but we highly regard WW advice.”

Josh Cunliffe, on the arbitration strategy

“I am responsible for carrying out follow-ups on the meetings that I conduct unless I delegate it to someone else, which is rarely the case because it's a relationship I need to nurture.”

Peter van den Steen, on the relationships that produce the work

“I do not have a tool to track it. We don't use a tool to track it, and we've recently had a meeting as partners in the business to discuss the fact that we need to coordinate our efforts and track it. We have not found a solution yet.”

Peter van den Steen, on the row with nothing at the end of it

The one question nobody was asked, and the one answer we have

Only one partner was asked what happens to her work when she is unavailable. Ashleigh's answer was that she prepares it beforehand, or works while she is away, or makes sure she is reachable. That is not a cover arrangement. It is a person absorbing the risk personally, and it is the only data point the firm has on any of these five single points.

The six, one at a time

What each partner does, what each one costs, and the thing only they told us.

The areas are each partner's own labels for their week, not ours. The last line on every card is what would be missing from this report if that person had not taken the call, which is the clearest answer to why the two outstanding partners matter.

Dave Lake

voice, 19 min

Founder

Keeping people aligned, the deep investigation, strategically assessing the business, and preparing the information. He puts the strategic assessment at the top himself.

“the answer, ironically, is in our ignorance”

2 weeks to a rough view80 to 90% reliableone investigation per engagement

Only from himThe anatomy of the plan document, the house rule that volume is the enemy, and the reason the firm reads a business better than the people who have run it for twenty years.

Peter van den Steen

voice, 17 min

Founder and Director

Formal business rescue under Chapter 6, consensual restructuring short of it, and an ever-expanding network of bankers and attorneys that he works personally.

“That's where I think we are falling short. I do not have a tool to track it.”

22 meetings in 10 weeks6 to 8 weeks per review1 to 2 h ExCo pre-read

Only from himThat the partners have already met about the tracking gap and found no solution, and that AI has been inside the drafting for six to eight months.

Martin Liebenberg

voice, 14 min

Partner and Director

Business rescue updates on live mandates, then the winding up, transferring and deregistration of what is left. Currently at the tail end across his clients.

“I use my email, so unless an item is cleared out, I will keep the email in my inbox.”

year six on one groupup to a month per sale agreementchased every 2nd or 3rd day

Only from himThe entire wind-down stage of the mandate, and why the group structure decides the shape of the plan before anybody writes a word of it.

Denis Chifunyise

voice, 11 min

Partner

Operational and financial reviews, engaging clients on the deep issues that led to distress, and being part of the solution, formal or informal.

“Usually email is the first thing that I begin with in order to set up my day, then usually compile a to-do list”

2 days to a week per model1 day to a week for the question seta few months in the worst case

Only from himThe consulting question set that carries the review, and the only correction any partner made when we read our understanding back to them.

Ashleigh Cohen

typed

Partner

Client financial modelling, the meetings that feed it and the meetings that review its output, and a large amount of Metis work that sits outside any mandate.

“I run quite a few internal processes that I dont bill for.”

3 to 5 h a week per client6 to 8 h per prospectabout 90% conversion

Only from herThe unbilled internal portfolio, seven categories deep, with no hours against any of it. It is the only bucket in this report with no number, and she is the reason we know it exists.

Josh Cunliffe

typed

Partner

Running mandates, two live creditor disputes heading to arbitration, business development with the lawyers and bankers who feed the firm, and scoping new opportunities.

“we (BRPs) were not around when historic events took place”

4 to 5 h per proposal3 to 4 a monthR300k to R700k smaller mandates

Only from himThe arbitration workstream and its archaeology problem, and that Metis is opening an investing arm to take equity in underperforming businesses.

Trevor Murgatroyd

invited

Founder and Director

Not yet interviewed. Anything he alone does is absent from every count, every stage and every figure in this report.

What is missingUnknown, by definition. That is the honest answer and it is the reason this report is marked a draft.

Gerhard Albertyn

invited

Partner

Not yet interviewed. The coverage matrix under Method carries two empty columns and this is one of them.

What is missingUnknown, by definition. Fifteen to twenty minutes closes it.

Five answers, four separate conversations

Five answers from four partners, none of whom was asked about the others. Read them in order.

“No, we have weekly meetings and we all maintain our own notes in different ways. Some people keep it in notes, some people write it down in a diary, others manage their emails. Each person is different.”

Martin Liebenberg, asked what tracks the close-out tasks on a live rescue

“I use my email, so unless an item is cleared out, I will keep the email in my inbox. Only once the task is complete will I move it out.”

Martin Liebenberg, on his own method. He checks it “all day”.

“Usually email is the first thing that I begin with in order to set up my day, then usually compile a to-do list or add to an existing to-do list.”

Denis Chifunyise, on how the working day starts

“That's where I think we are falling short. I do not have a tool to track it. We don't use a tool to track it, and we've recently had a meeting as partners in the business to discuss the fact that we need to coordinate our efforts and track it. We have not found a solution yet.”

Peter van den Steen, on tracking the network that produces your work

“I ensure that it is prepared beforehand, or I will work while Im away. I will ensure I am reachable when it is due or agree beforehand with client for a delayed review meeting.”

Ashleigh Cohen, asked what happens to the model sign-off when she is unreachable

Four different systems, and one plan for the single point of failure, which is that the single point of failure does not go offline.

Two interviews, one chain, neither partner made it

Josh is paying, on a live file, for a record nobody kept. Dave explains why the people who remember are gone.

Josh Cunliffe, on the Concor arbitration

“It's often not that easy because we (BRPs) were not around when historic events took place, so it involves discussing historic events with previous M&R employees and reviewing historic comms, documents and circustances to try and understand the circumstances and get to a firm undertstanding of the claims and our arguments.”

Dave Lake, on how a rescue begins

“more often than not, one has to make significant changes to the management team so that you don't have the same management team which got the company into trouble working with us to try and get it out.”

The part worth sitting with

Josh is reconstructing Murray and Roberts's institutional memory out of people who have already left. Dave, four days later and without having heard him, describes the policy that removes them. Neither partner drew the line between the two, because neither was in the other's conversation. Metis runs the same exposure on its own file: the record of what is open lives in individual inboxes and individual heads, and it is one retirement away from being somebody else's archaeology.

Every duration your partners named, on one scale

From five minutes to six years.

Twenty three figures, taken from the six interviews and plotted on a logarithmic scale. The top lane is effort, meaning hours a partner actually spends. The bottom lane is elapsed, meaning calendar time a thing takes. That distinction is the one our own interviewer got wrong once, which is written up in Method. The dashed row is the most valuable number in the cohort, and we do not have it.

Nearly six orders of magnitude of duration, and every one of them is tracked in the same place: somebody's inbox.

02 · Findings

What was said, and by how many

Eight things came up. Here is exactly how many of the six raised each one.

The marks show the count. Nothing is presented as consensus unless six people said it. A single mark means one partner said it and the other five did not, which is often the more interesting number, and twice in this report it is the most useful thing in the document.

6 of 6

Every engagement opens by unpacking someone else's business by hand.

All six partners described the same opening act in six different vocabularies: the deep investigation, the deep dive review, the pre-assessment, the independent business review, the organogram, the model build, the scoping meeting. It is the firm's product, it is what the referral is actually buying, and it is done entirely by people.

“the restructuring stuff is really about, first of all, understanding, getting a deep understanding of the business, what makes it tick, why it's in distress, and then assessing what are the alternatives”

Dave Lake

“It was a deep dive review. My recollection, because it's a number of years ago now, is that that review probably took about, it must have been six or eight weeks, something like that.”

Peter van den Steen, on Stefanutti Stocks

“It's also trying to understand the business, sit down with management, just allow them to unpack the business to you so you can see areas where you need to focus.”

Martin Liebenberg

Dave · Peter · Martin · Denis · Ashleigh · Josh

5 of 6

Waiting on somebody else's information is the dominant delay, and Dave puts the first number on it.

Five partners described the chase in different forms: a cagey client, an absent tax authority, a query loop that will not close, employees who have already left. Dave is the only one who has quantified it, and he adds a reason the digging cannot be skipped even when the information is good. Peter is the exception and says so plainly, which is why this is five and not six.

“on average, I would say most companies are 80 to 90% reliable. And there's that 10 to 20% where people are being optimistic in order to, for themselves, psychologically see a way through the crisis.”

Dave Lake

“if the banks and other third parties outside of the company are going to trust the work that we do, we have to go and investigate, we have to unpack the numbers and make sure that they are reliable”

Dave Lake, on why you dig even when the numbers are sound

“Some clients are usually misinformed with the process that's run, so a bit cagey in terms of what they supply. In the worst case, it's taken even a few months.”

Denis Chifunyise, on waiting for the financial reports

“It wasn't really necessary. The team was actually quite on point, and the people within the organization were very, very cooperative”

Peter van den Steen, the dissent, asked who he had to chase

Dave · Martin · Denis · Ashleigh · Josh | Peter dissents

4 of 6

Nothing outside an individual's inbox knows where the work stands.

Four partners described how they personally keep track. No two described the same method, none of the four is readable by anybody else, and the firm-level answer is that there is no answer yet. Peter volunteered that the partners have already met about exactly this and have not found a solution. He is not describing a gap he has just noticed. He is describing one the firm has already looked at.

“we all maintain our own notes in different ways. Some people keep it in notes, some people write it down in a diary, others manage their emails. Each person is different.”

Martin Liebenberg

“We have not found a solution yet.”

Peter van den Steen, on the partners' own meeting about tracking

The three answers that matter most in this report

Asked how the gap shows up in practice, Peter said “I don't know”. Asked who on the team would know, he said “I don't think anybody does”. Asked for a single instance of a lead slipping, he said “I can't think of any right now”. Three non-answers in a row look like an unhelpful stretch of interview. They are the opposite. You cannot produce an example from a system that keeps no record, so the absence of the anecdote is the finding, and it is the strongest evidence in the document that nothing is measuring this.

Peter · Martin · Denis · Ashleigh

4 of 6

Work arrives. The firm does not go and get it.

Four partners described the same intake, unprompted, and one of them put a conversion rate on it. This is a strength and it is stated as one. It is also the reason the network is load-bearing, which is what makes the absence of any record of that network the expensive part.

“We normally dont do outreach ourself, we usually get approached by banks and lawyers”

Ashleigh Cohen

“If work comes through in this way, almost 90% it will land a mandate”

Ashleigh Cohen, on a bank or lawyer introduction

“it's difficult for us to actively seek work - our clients are looking for advisory services to assist them with their distressed businesses - so they usually come to us and not the other way around”

Josh Cunliffe

Peter · Denis · Ashleigh · Josh

2 of 6

The same four groups have to be kept calm, and two partners named them identically.

Dave and Martin were interviewed on different days, neither heard the other, and neither was asked to enumerate anything. Both produced the same four counterparties in the same order of urgency: staff, then lenders and bankers, then suppliers, then customers. When two people independently produce the same list, the list is a real structure in the work and not a turn of phrase. It also runs underneath every one of the six stages rather than sitting in any of them.

“you need to calm people down, you need to reassure people that there's a way out of this because losing staff is a real danger during these times. You need to keep the bankers and lenders calm”

Dave Lake

“There is, I think, employee matters, calming the employees down, the suppliers calming them down.”

Martin Liebenberg, four days earlier

Dave · Martin

2 of 6

AI is already inside the firm, brought in by partners themselves.

Neither of them was asked about AI. Both volunteered it, in different parts of the business, and Peter's is already in production against the exact step another partner calls the thinking. This is not an appetite question for Metis. It has been answered already, by the partners, without a programme.

“over the last six to eight months, employed artificial intelligence to assist us, and it has significantly cut down the time that we need to put these plans together. We are getting better at briefing AI to make it happen”

Peter van den Steen

“business development like creating AI tools and improvement processes that we can utilise in the business”

Ashleigh Cohen, listing the internal work she does not bill for

Peter · Ashleigh

1 of 6

Concision is a house rule, and volume is the named anti-pattern.

One partner said this, and he is the one who holds the pen, which makes it the binding constraint on every document the firm produces and on anything the firm ever buys to help produce them. He also gave the full anatomy of the deliverable without being asked for it: problem, status quo, the strategic alternatives, what each one produces with its risks and opportunities, a recommendation for the board, and appendices behind it for anyone who wants to check.

“in our firm, we avoid voluminous statistical analysis, pages and pages of data about what happens in the company, what happens in other companies in the same industry, what happens in other companies in other geographic regions, etc. The kind of thing that the auditors do. We don't do that.”

Dave Lake

“it's a class document which tells a story from one end to the other. And it shouldn't be too long. You should keep it short as you can.”

Dave Lake

Dave

1 of 6

The firm's advantage over management is that it does not know the business.

Asked what else mattered about assessing a business strategically, Dave answered a question nobody had put to him: why a team that has been in the building for a fortnight can read a company better than the people who have run it for twenty years. It is the clearest articulation of the firm's method in the whole corpus, it was unprompted, and it is the argument for reading a business through a channel with no history in it.

“Often people ask why we have an advantage strategically to look at a business compared to management who have been working in the business for the last, say, 20 years. And the answer, ironically, is in our ignorance.”

Dave Lake

“we look very, very coldly at the facts that are on the table, the facts that we uncover through our analysis”

Dave Lake

Dave

Where your partners do not agree

Five places where two partners gave answers that cannot both be a complete description of one firm.

These exist only because six people were read against each other, and they are the most useful material in the report. Four of the five resolve once you see what each partner is actually describing. The first one does not resolve, and it is the one to talk about.

1. Is writing the plan a cost, or is it the thinking?

Peter van den Steen

“Unfortunately, it's the actual writing of the plan. The writing up of the plan has proven itself over the years to take up way too much time. The information and what we want to put into it is available, and then the time to actually put a restructuring plan together has taken way too much time.”

Dave Lake, who holds the pen

“It's actually a funny thing that preparing that information is almost part of the thinking process, the two go hand in hand, and so they get built simultaneously.”

What is actually going on

Two founders, one artefact, opposite descriptions, and both are right. There is a half of that document which is transcription, and Peter has been compressing it with AI for six to eight months already. There is another half which is the reasoning itself, and Dave is telling you it cannot be separated from the writing without losing it. The distinction is not academic. It decides whether “help you produce the plan faster” is a proposition or an insult, and nobody can draw the line except the two of them.

2. How long does the first deep look take?

Denis 1 day to a week, the question set Dave 2 weeks, a rough view, explicitly not firm Peter 6 to 8 weeks, the review complete Ashleigh months, the model, and it never stops

What is actually going on

Four answers spanning two orders of magnitude, and not one of them is wrong. They are four different milestones inside a single envelope, and Dave's two weeks is the first split anybody in the firm has given of Peter's six to eight. Nobody at Metis has said this out loud, which means the firm has no shared language for how far into a mandate it is. The diagram is on the Flow screen.

3. How often does it happen?

Peter van den Steen

“I think deep dive reviews, we as a team in Metis Strategic Advisors, we probably do that half a dozen times a year.”

Dave Lake

“It is absolutely one per engagement. It is one deep investigation per engagement... You do probably a few engagements a year and so it'll be a few investigations per year.”

What is actually going on

Dave rejected the unit before he answered. Asked how many times a week or a month he does this, he said he was not sure he understood the question, and then explained that the period is the wrong denominator entirely. His unit is the engagement, and it is the one that can be planned against. Peter's half a dozen is the team, not the person.

4. Is the firm busy, or is it quiet?

Martin Liebenberg, and separately Josh Cunliffe

“it's actually very quiet where we are right now in our phase, so they have a lot of spare time.”

“our team is quite quiet at the moment, so we are wanting to take on smaller work.”

What the same six interviews contain, in our words not theirs

Peter's 22 face to face meetings in ten weeks, plus a monthly ExCo that runs from half past eight until half past three or four. Ashleigh's weekly and monthly model cycles, on top of an entire portfolio of internal work she says she does not bill for. Josh's own three or four proposals a month, at four to five hours each.

What is actually going on, and it is the point of this report

Both readings are honest. The firm feels quiet because the load that is genuinely full is the load nobody bills, and a firm measures itself on what it bills. That is the same finding as the headline, arrived at from the opposite direction: the biggest block of partner time is invisible to the firm precisely because no invoice is ever raised against it.

5. Does the work repeat?

Josh Cunliffe

“My days very seldom look similar. The nature of our work is that it's not particularly repetitive. Each mandate / project / enagement has different components and therefore, requires different workstreams and thinking.”

Denis Chifunyise

“So that's pretty much a day-to-day and repeats on a continuous basis.”

What is actually going on

Josh is describing the content and Denis is describing the shape, and both are exactly right. No two mandates contain the same facts. Every mandate runs the same sequence, which is why six people who were never shown each other's answers produced one flow diagram. For anybody proposing to systematise anything here, including us, that is the whole test: the content is bespoke and must stay bespoke, the sequence is not and never was.

Single source, carried anyway

What one partner sees that the other five did not mention.

These are carried at one mark each and labelled as such. A single account is not weak evidence, it is narrow evidence, and the difference matters when the one person who said it is the only person who does that job.

Martin Liebenberg, and only Martin

The end of a mandate is a matrix, not a decision: bring the filings current and deregister, or abandon it the way a liquidator would, or apply for liquidation. He owns that call. One group he described has over a hundred companies and is in year six.

Nobody else described the wind-down at all. Stage 6 of the flow rests entirely on this one account.

Josh Cunliffe, and only Josh

Two live creditor disputes running to arbitration, with the statement of defence built out of historic documents and former employees. Metis is also opening an investing arm and taking equity positions in underperforming businesses, which he raised without being asked.

The investing direction is a second front and no other partner mentioned it.

Ashleigh Cohen, and only Ashleigh

An entire portfolio of firm work that is never billed: presentations, lecturing and slide content for SARIPA and INSOL, event and marketing organisation, AI tooling, prospect research. She named seven categories in one answer and no hours against any of them.

This is the third unbilled bucket. It is the only one in this report with no number attached, because nobody has ever counted it.

Dave Lake, and only Dave

The rebuild of the Stefanutti cash flow, in sequence: build our own model, fact check every assumption in theirs, restate it conservatively with the reasoning written down, then run the restructuring off ours. It explains from the founder's side why Ashleigh's collate, query, reconcile, review cycle exists at all.

Two partners, four days apart, describing the same artefact from opposite ends.

03 · Flow

The mandate, end to end

Six stages, assembled from six separate accounts. Nobody was asked to describe the whole thing.

Each partner was walked through their own corner. The sequence below is what those six corners make when they are laid against each other, and the fact that they fit is itself a result. The coral bar at the top of a card means that stage waits on somebody outside the firm.

The question with four right answers

How long does the first deep look take?

Four partners answered this and gave four durations spanning two orders of magnitude. Read as a disagreement it is alarming. Read properly it is a decomposition, and it is the single most useful thing six interviews produced that one interview could not. Dave's two weeks is the first split anyone in the firm has given of Peter's six to eight.

“small businesses might take maybe a day or two. Large complex listed businesses, it could take even a week.”

Denis Chifunyise, on completing the question set

“within two weeks of working with an organization, you've got a rough view of where things might go. It's definitely not a firm view at that point, and it would definitely change or be refined.”

Dave Lake, the split nobody else has given

“it must have been six or eight weeks, something like that”

Peter van den Steen, on the complete deep dive review

“The initial production and collation could take months depending on the size and complexity of the client.”

Ashleigh Cohen, on the first build of the model

Why this matters more than it looks

A firm with no shared language for how far into a mandate it is cannot tell a client, a lender or a new partner where the work stands, and cannot tell itself either. Dave was careful to say his two week view is “definitely not a firm view at that point”. That caveat is the useful part. It marks a real checkpoint, roughly a quarter of the way through the envelope, that already exists in the work and is not named anywhere.

Stage by stage, with who described it and what each one waits on.

Stage 1 · the referral

A bank, an attorney, a lender or a guarantor calls.

Four partners described the intake and none described going out to find work. Peter's Stefanutti mandate began with a phone call from a construction guarantee provider, which put Metis in a room with the banks before it was ever in a room with the company.

Conversion about 90% once a bank or lawyer brings it Frequency large mandates, at least one a year hoped for Value tens of millions large, R300k to R700k small Held by whoever owns the relationship, individually

Peter · Denis · Ashleigh · Josh

Stage 2 · scoping and the proposal

NDA, a meeting to understand the need, a tailored proposal, an engagement letter.

Josh gave the only hours on this step and the only volume: four to five hours a proposal, three or four out the door in the last month. He also said plainly that not many convert, which is the honest counterweight to the 90% figure above. The two are not in conflict. Ashleigh's 90% is a bank or lawyer introduction. Josh's proposals include the colder end.

Effort 4 to 5 hours per proposal, personally Volume 3 to 4 in the last month Unbilled every hour of it is pre-mandate

Ashleigh · Josh · Denis

Stage 3 · the deep investigation

Organogram, cash flow, the question set, site visits, and management at every level.

All six partners described this stage, which no other stage can claim. It is the firm's product. Dave frames it as one deep investigation per engagement rather than a recurring task, and gives the reason the digging cannot be skipped even when the numbers are good: the banks have to be able to trust the work, so the unpacking is as much about third party credibility as it is about accuracy.

Dave a rough view at 2 weeks, not firm Peter 6 to 8 weeks complete, half a dozen a year as a team Denis 1 day to a week for the question set Ashleigh months for the first model build Reliability 80 to 90% of what management gives you Waits on client finance and operations teams

Dave · Peter · Martin · Denis · Ashleigh · Josh

Stage 4 · the plan

The group structure dictates the plan. Then somebody has to write it.

Martin explains why the structure comes first: on Murray and Roberts the holding company stood behind guarantees large enough that the group could not be sold whole, so the plan had to be built around selling several levels down. Dave gives the shape of the document that results, and the house rule that governs it. Peter gives the cost of producing it, and the two founders do not agree about what that cost is.

The pen Dave Lake, named by Peter Versions about eighteen over four years on one plan Already compressed AI on the writing for 6 to 8 months Waits on board, guarantors and lender approval

The number this report is missing

We asked Dave how many hours he personally spends building the assessment document. It is the pen holder's own cost for the artefact the firm is named for, and it is the one figure that would settle the disagreement in stage 4. Our interviewer asked the question and then closed the interview without waiting for the answer. That is our defect, it is written up in Method, and the figure is still outstanding.

Dave · Peter · Martin

Stage 5 · implementation

Fire hydrant first, then a model cycle that runs for years.

Martin's account of the opening weeks is the most vivid thing in the corpus: three to four thousand employees, retention agreements, every operational site visited, creditors calmed, and C-suite roles that sometimes have to be stepped into because the incumbents walked. Then it settles into Ashleigh's cycle, which is weekly or monthly forever. Peter's Stefanutti plan ran four years through roughly eighteen versions.

Weekly cycle 3 to 5 hours a week, per client Query loop usually closes inside 24 hours One rescue item can take a full week, all of it Sign-off Ashleigh, personally, with no cover Waits on spreadsheets, workbooks, paper statements

Dave · Peter · Martin · Ashleigh · Josh

Stage 6 · wind-down

Abandon, deregister or liquidate. Then hand the company over clean.

One partner described this stage and nobody else touched it. Martin walks the matrix: if too much is owed and the cost to wind down is too high you abandon it the way a liquidator would, if there is something worth keeping you bring every filing current and deregister, and if neither works you apply for liquidation. A construction group might have opened two hundred companies and be trading in twenty three.

Per country 6 months to 2 years to bring filings current One live group 100+ companies, now in year six The decision the business rescue practitioner, alone Waits on the tax authorities, in each jurisdiction

Martin only

Coverage, stated honestly

Stage 3 is the only one every partner reached. Stage 6 rests on a single account.

Stage 3 is the only stage every one of the six touched, five of them in detail, and it is the most solid thing in this report. Stage 6 is described by one partner and mentioned by one other. That is not a criticism of the interviews, it is a fact about who does what, and it is the reason the two outstanding partners matter: Trevor Murgatroyd and Gerhard Albertyn have not been interviewed, and until they are, any stage they own is invisible here.

04 · Evidence

The ledger

Every number in this report, with the sentence it came out of.

Thirty three rows. Click any sentence to open the exact exchange it came from, the question we asked and the answer given. Exactly four of them can be turned into a monthly number, because in those four one partner gave both a duration and a frequency in their own words. The other twenty nine are listed and left alone, because summing across units nobody stated is how a report stops being true.

The missing input

This report has no price in it, and it is not going to guess one.

Nobody gave us a charge-out rate, so nothing here is converted into money by us. Type a rate and every hour figure below prices itself, live. The presets are illustrative round numbers, not an estimate of what Metis charges.

R per hour

Nothing is stored and nothing is sent anywhere. This runs entirely in your browser.

Unbilled, per month

36 to 54 hours a month

Peter's networking plus Josh's proposals. Two partners, one activity each, neither billable.

Unbilled, per year

432 to 648 hours a year

The same two activities, annualised on their own stated volumes.

Everything measured, per month

50 to 77 hours a month

Adds Ashleigh's weekly model cycle and Peter's ExCo pre-reading, which are client work.

Everything measured, per year

600 to 924 hours a year

Three partners, four activities. Two of your eight partners are not in this number at all.

Enter a rate to compare this against a mandate

Read this as a floor, not a total. It covers four activities from three partners. It excludes the entire portfolio of internal firm work Ashleigh described and does not bill for, because she gave seven categories and no hours, and nobody at Metis has ever counted them.

Filter by partner. Click any sentence to open the transcript.

Partner	What it measures	Figure	Their own words	How we read it
Peternetworking	Face to face meetings that produce work	24 to 34 h / month	“since the 20th of July until the end of September, I had 22 physical face-to-face engagements of a networking nature”	costable 22 meetings at his own 2.5 to 3.5 hours, across the ten week window he named
Joshproposals	Drafting the principles document	12 to 20 h / month	“It probably took me 4-5 hours to put together the proposal. I've done 3-4 in the last month.”	costable his own duration times his own volume, in one sentence
Ashleighmodel cycle	Weekly reconciliation and review	13 to 22 h / month	“Weekly takes a bit quicker as less to reconcile at once time, usually 3-5 hours per week”	costable her weekly figure at 4.33 weeks in a month
PeterExCo	Pre-reading before a monthly ExCo	1 to 2 h / month	“It could take me between one and two hours.”	costable one ExCo a month, on one mandate
PeterExCo	The ExCo meeting itself	7 to 8 h, monthly	“The ex-co meeting would typically last from about half past eight in the morning until half past three, half past four in the afternoon.”	Client facing, so not in the unbilled figure. The pack runs to four or five hundred pages and he scans rather than reads it.
Peternetworking	One meeting, door to door	4 hours	“I left home at 6am, I only got back at about 10am, so it's four hours out of my day.”	Above his own stated 2.5 to 3.5 range, so the range used in the model is the conservative one
Ashleighmodel cycle	Monthly reconciliation	8 to 15 hours	“monthly reviews take longer to reconcile the whole month - usually 8-15 hours per week also client dependent”	flagged written as per week inside a description of a monthly cycle. Deliberately excluded rather than guessed.
Ashleighbusiness dev	Research and analysis on one prospect	6 to 8 hours	“6-8 hours. We normally dont do outreach ourself, we usually get approached by banks and lawyers”	Duration given, frequency explicitly not monthly, so no monthly figure
Joshdisputes	One working session with counsel	1 to 2 hours	“We have had several sessions. Sessions are usually an hour or two.”	Several is not a number, so this cannot be converted
Martinclose-out	One close-out task on a winding rescue	5 min to 1 hour	“Each one can be completely different. One could be five minutes, one could be an hour.”	Volume per day not stated, though he says four or five meetings are blocked out in his mail
Martinclose-out	Chasing one sale agreement to close	every 2nd or 3rd day	“Depends on the size and depends on the input that you receive from the people, but it shouldn't take longer than a month.”	He gave both halves in consecutive answers, so the touch count is real: roughly ten to fifteen per agreement. Time per touch was not asked.
Martinpeak load	One rescue item at the opening of a mandate	a full week	“one of those specific items could take you a week, just 100% of your time”	Peak load, not steady state. His phrase for the opening weeks is drinking water from a fire hydrant.
Davethe document	Hours spent building the assessment	not answered	“On the Stefanidi case, thinking of the last time you built the assessment of the options, roughly how many hours did you personally spend building it?”	our defect the question was asked and the interview closed over it before he replied. Written up in Method.
Denisreview record	Producing the initial review document	48 hours	“I guess it depends how transparent an open the interviewee has been. So yeah, not too long, but 48 hours at most.”	flagged reads as turnaround, not effort. Our interviewer then restated it as work time and called the artefact a Word document. He said neither. Excluded.
Denismodelling	Cash flow scenario model	2 days to a week	“I have a cash flow from about two days, not a day, two days, to a week.”	He corrected us at the read-back and the corrected figure is the one used. This is the read-back working.
Denisquestion set	Completing the consulting question set	1 day to a week	“small businesses might take maybe a day or two. Large complex listed businesses, it could take even a week.”	Elapsed time on one instrument, not the whole review
Deniswaiting	Waiting on the client's financial reports	up to a few months	“In the worst case, it's taken even a few months.”	Worst case, explicitly. Normal is after the NDA, from the client.
Daveinvestigation	First rough view of a business	2 weeks	“within two weeks of working with an organization, you've got a rough view of where things might go”	He was careful that this is not a firm view. It is the first split anyone has given of Peter's six to eight weeks.
Daveinvestigation	Reliability of management information	80 to 90%	“on average, I would say most companies are 80 to 90% reliable”	The first number anybody has put on the information chase. The gap is optimism in the cash flow, not concealment.
Daveinvestigation	Deep investigations per year	a few	“It is absolutely one per engagement. It is one deep investigation per engagement.”	He rejected the weekly and monthly framing before answering. His unit is the engagement.
Peterinvestigation	The deep dive business review	6 to 8 weeks	“It was a deep dive review. My recollection, because it's a number of years ago now, is that that review probably took about, it must have been six or eight weeks, something like that.”	One mandate, recalled from several years ago, and he flags the recall himself
Peterinvestigation	Deep dive reviews per year	about six	“I think deep dive reviews, we as a team in Metis Strategic Advisors, we probably do that half a dozen times a year.”	The team, not the person. Dave counts the same work per engagement instead.
Peterthe plan	One plan through its revisions	4 years, about 18 versions	“there was probably a dozen and a half versions of the plan every time asking for revisions or suggesting revisions to the plan”	Stefanutti Stocks, through COVID and successive funding requirements
Peterthe plan	How long AI has been in the drafting	6 to 8 months	“over the last six to eight months, employed artificial intelligence to assist us, and it has significantly cut down the time that we need to put these plans together”	Volunteered, not asked. Already in production against the step Dave calls the thinking.
Ashleighmodel cycle	Query back to an information provider	usually 24 hours	“Differs per client but usually within 24 hours”	When it drags she escalates it to a meeting rather than chasing again
Ashleighmodel build	Building a model for the first time	months	“The initial production and collation could take months depending on the size and complexity of the client.”	After that it settles into the weekly or monthly cycle above
Ashleighintake	Conversion once a bank or lawyer introduces	about 90%	“If work comes through in this way, almost 90% it will land a mandate”	Warm introductions only. Josh's colder proposal flow converts far lower and he says so.
Ashleighmandates	Value of a large mandate	tens of millions	“The large mandates are worth tens of millions, depending on the remuneration agreed with client”	Not used in the model. Hoped for at least one a year.
Joshmandates	Value of a smaller mandate	R300k to R700k	“A large mandate will be in the R millions but smaller ones will be in the R300 to R700 thousand range.”	Used only as the comparison line under the rate model, in his own range
Joshdisputes	Live work remaining on the contested JV	4 months	“I can't see that they will be able to get a ruling on the disputed termination of the JV before the live work is complete - there is only 4 months left.”	Which is why he frames that dispute as risk mitigation rather than an existential one
Martinpeak load	Employees in scope on a live rescue	3 000 to 4 000	“I'm dealing with potentially three to four thousand employees, where those employees might be being transferred out, being retrenched, or just to keep them calm”	Peak of a large mandate, with three or four Metis people on it
Martinwind-down	Bringing one country's filings current	6 months to 2 years	“some countries, you know, it could be six months, some countries it could be two years”	Not Metis effort. It sits with the tax authority once handed over.
Martinwind-down	One group, still winding down	year six	“I can tell you now in a company with over 100 companies, we are now year six and we're still dealing with them”	Elapsed. A construction group might open two hundred companies and trade in twenty three.

Four rows are marked costable and only those four feed the model. Three carry a warning: one because a unit is ambiguous in the participant's own writing, one because our interviewer put words into an answer, and one because our interviewer asked the best question in the whole exercise and then closed the interview before the answer arrived. All three are excluded rather than resolved, and all three are named in Method.

05 · Method

How this was made

How this was made, what it does not cover, and the five places our instrument got something wrong.

A report that only tells you where the client is weak is half a report. This screen is the other half. Everything our own instrument did badly is written down here, including the one number we asked for and lost.

What happened: Each partner opened a link and had a conversation with an AI interviewer. Four spoke, two typed. Nobody saw anybody else's answers, at any point, and no two interviews ran on the same script because the questions follow whatever the partner has just said.
When: 3 to 7 September 2026. Peter first, then Josh and Ashleigh, then Denis, Martin and Dave on the morning of the seventh.
How much of your time: Two hours and thirteen minutes in total, across six partners, measured from opening the link to finishing. The longest was Dave at nineteen minutes. No consultant spent a day on site and no diary was coordinated.
What was analysed: 272 conversational turns, of which 142 are your partners answering, totalling 7 786 words in their own voices. Nothing was summarised before analysis. Every quotation and every figure in this report opens the exact exchange it came from, the question we asked and the answer given.
What is not here: Trevor Murgatroyd and Gerhard Albertyn have not been interviewed. Any part of the firm that only they can describe is invisible in this report, and no finding here should be read as covering it. Every count is out of six.
Quoting: Written answers are reproduced exactly as typed, including typing slips, because smoothing somebody's writing is the first step towards smoothing their meaning. Where the transcript split one continuous thought across consecutive turns by the same speaker, those turns are joined. No word is added, changed, reordered or removed inside any quotation, and every quotation is cut at a sentence or clause boundary rather than mid-phrase.
What we refused to do: No figure is annualised, extrapolated or converted into money unless one partner gave both the duration and the frequency in their own words. Where a unit was ambiguous the figure is flagged and excluded rather than resolved by us. There is no benchmark in this report, no industry comparison and no recommendation.

Coverage

Which partner described which part of the work, and how thoroughly.

Nobody was asked to cover the whole mandate. This is what each of them reached on their own, and it is the honest picture of how much weight each stage of the flow can bear.

Stage of the mandate	Dave	Peter	Martin	Denis	Ashleigh	Josh	Trevor	Gerhard	Detailed
1 · The referral arrivesbank, attorney, lender, guarantor									3
2 · Scoping and the proposalNDA, meeting, principles document									2
3 · The deep investigationorganogram, cash flow, question set									5
4 · Building the planstructure, options, the document									3
5 · Implementationmodel cycle, disputes, years of it									4
6 · Wind-downabandon, deregister, liquidate									1
Running underneath · keeping people calmstaff, lenders, suppliers, customers									2
Running underneath · tracking what is openthe gap this report is about									2

described in detail, with specifics mentioned in passing not raised by this partner not yet interviewed

Stage 3 is the only one with five detailed accounts, which is why it carries the most weight in this report. Stage 6 has one. Two whole columns are empty and will stay empty until Trevor and Gerhard complete.

Where our instrument got it wrong

Five defects on our side, written down rather than tidied away.

Each of these changed something, and one of them cost the most valuable figure in the whole exercise. They are here because a review that hides its own instrument's failures is asking you to take the rest on faith.

We asked the pen holder what his own document costs him, and then closed the interview over the answer.

The last question put to Dave was how many hours he personally spends building the assessment of the options. That is the pen holder's own cost for the artefact this firm is known for, and it is the figure that would settle whether writing the plan is a bottleneck or the thinking itself. Our interviewer emitted its closing thank-you twice, both times truncated mid-word, without waiting. The figure is gone and it has to be asked again.

Consequence: the disagreement between Peter and Dave in Findings cannot be resolved from this data. Logged on our side as a product defect, not a note.
We put two things into Denis's answers that he never said.

Asked how long the initial review record takes, Denis said “48 hours at most”, which reads naturally as turnaround. Our interviewer then restated it back to him as “48 hours of work time” and separately described the artefact as “a Word document”. He said neither of those things. Both are excluded from every figure in this report and both are marked in the ledger.

Consequence: a restatement is not a quotation, and ours was not marked as one at the time. Forty eight hours of partner effort is a very different number from a two day turnaround, and only one of them is his.
The read-back is a partial control, because people correct selectively.

We read our understanding back to three partners. Martin said “Go back a step” and corrected us. Peter confirmed ours was right. Denis corrected the cash flow figure, from a day to two days, and in the same breath let the 48 hours restatement stand unchallenged. So the read-back caught one error and missed another in the same answer, from the same person.

Consequence: a confirmation at the end of an interview is evidence that something was heard, not evidence that everything was checked.
Josh's microphone never reached us, and the fault was ours.

His voice session recorded zero words from him. He told us so inside the interview: “I can hear you but you are not picking up what I am saying through my microphone”. He then completed the entire interview by typing, and it is one of the two richest written accounts we have. He has since said he believes typing cost the openness of his answers, and he is probably right.

Consequence: his interview is used on the same footing as the others because the content earns it. The defect is ours and is being fixed.
The opening was emitted three times to Denis, and Dave's first words came back as Danish.

Denis's first two replies were mis-transcribed as fragments and the introduction repeated. Dave's opening reply was transcribed as a sentence in Danish. Neither affected a single finding, because both happened before any substantive question, neither is hidden from the exchanges this report opens, and neither is going to be quietly deleted.

Consequence: cosmetic here. Not cosmetic if it happened mid-answer, which is why it is written down.

One thing our instrument got right, and it matters

We pushed Peter three times for a concrete instance of a follow-up slipping. He said “I don't know”, then “I don't think anybody does”, then “I can't think of any right now”. A human interviewer under time pressure would very likely have moved on and left it out of the write-up as an unproductive stretch. It is the single most load-bearing exchange in this report, because a firm cannot produce an example from a system that keeps no record, and three refusals in a row are what that looks like from the outside.

06 · The fix

Deliberately held

This report stops at description on purpose.

Everything up to here is what six partners said, counted and traced. What follows would be what we think you should do about it, and we have not earned that yet. Two of your partners have not spoken, one number we need is missing because of our own defect, and nobody has given us a rate. So this screen lists what is held and why, rather than pretending the questions are answered.

Held until the cohort is complete

Four questions this report raises and deliberately does not answer.

01
Which half of the plan is transcription, and which half is the thinking?
Peter has been compressing his half with AI for six to eight months. Dave says his half cannot be separated from the writing. Only the two of them can draw that line, and the number that would inform it is the one our interviewer lost. This is the most commercially loaded question in the document and we are not going to answer it from the outside.
02
What does the untracked network actually cost?
Peter gave 22 meetings in ten weeks at two and a half to three and a half hours each, and could not name a single instance of a follow-up slipping, because nothing records them. The cost is not zero and it is not knowable from here. It becomes knowable the moment anything at all starts recording what came out of each conversation.
03
How much of Ashleigh's unbilled portfolio is firm-building, and how much is just unbilled?
She listed seven categories of internal work she does not bill for and gave hours against none of them. It is the only bucket in this report with no number. Some of it is genuine firm-building and should never be billed. Nobody currently knows the split, including her.
04
Is the firm quiet, or is it fully loaded on work it does not invoice?
Two partners said quiet. The same six interviews contain a monthly all-day ExCo, 22 networking meetings, weekly model cycles and a portfolio of unbilled internal work. Both readings are honest. Which one is true is a question about how Metis measures itself, and it is yours.

What would unlock it

Three things we would put a clock on next. Each one is a measurement, not a proposal.

Interview Trevor and Gerhard, and re-ask Dave the one question we dropped.

Two empty columns in the coverage matrix and one missing figure. Both are cheap to close: the interviews are about fifteen to twenty minutes each and the re-ask is a single question.

What it changes: stage 6 stops resting on one account, any stage Trevor or Gerhard owns becomes visible, and the disagreement between the two founders in Findings becomes resolvable.
Give us a charge-out rate, or three of them.

The evidence table is deliberately unpriced and the rate box will price the whole thing live the moment there is a number in it. One rate is enough to start. Different rates for different seats is better.

What it changes: every hour figure in this report becomes a rand figure that the partners can argue with, which is the only form in which any of this becomes a decision.
Record what comes out of the next twenty networking conversations. Anything at all.

Not a system, not a CRM selection, not a project. A shared list with a date, a name and what was agreed. Twenty entries is enough to see whether anything is being dropped, and it is the only way to turn Peter's “I don't know” into a number.

What it changes: the largest measured block of partner time in this firm becomes visible to the firm for the first time. Everything else is downstream of that.

Where this leaves it

Six partners described the same sequence. No two of them run it the same way, and nothing outside their heads knows where any of it stands.

That is the whole finding, and it took two hours and thirteen minutes of your partners' time to produce, with nobody on site. Two partners are still to speak, and this report will be reissued when they do. Until then everything in it is counted out of six, every figure traces to a sentence one of your partners actually said, and everywhere our own instrument fell over is written down in Method.

6 of 8 complete Trevor Murgatroyd and Gerhard Albertyn outstanding One figure outstanding on our side

The largest block of partner time we could measure is the one thing the firm does not track, does not bill, and cannot cost.

Eight things came up. Ranked by how many of the six said each one.

Six of eight partners are in.

Six things the work produces. Five land on a single point, and the sixth lands nowhere.

Each line, in the words it came from.

What each partner does, what each one costs, and the thing only they told us.

Five answers from four partners, none of whom was asked about the others. Read them in order.

Josh is paying, on a live file, for a record nobody kept. Dave explains why the people who remember are gone.

From five minutes to six years.

Eight things came up. Here is exactly how many of the six raised each one.

Every engagement opens by unpacking someone else's business by hand.

Waiting on somebody else's information is the dominant delay, and Dave puts the first number on it.

Nothing outside an individual's inbox knows where the work stands.

Work arrives. The firm does not go and get it.

The same four groups have to be kept calm, and two partners named them identically.

AI is already inside the firm, brought in by partners themselves.

Concision is a house rule, and volume is the named anti-pattern.

The firm's advantage over management is that it does not know the business.

Five places where two partners gave answers that cannot both be a complete description of one firm.

1. Is writing the plan a cost, or is it the thinking?

2. How long does the first deep look take?

3. How often does it happen?

4. Is the firm busy, or is it quiet?

5. Does the work repeat?

What one partner sees that the other five did not mention.

Six stages, assembled from six separate accounts. Nobody was asked to describe the whole thing.

How long does the first deep look take?

Stage by stage, with who described it and what each one waits on.

A bank, an attorney, a lender or a guarantor calls.

NDA, a meeting to understand the need, a tailored proposal, an engagement letter.

Organogram, cash flow, the question set, site visits, and management at every level.

The group structure dictates the plan. Then somebody has to write it.

Fire hydrant first, then a model cycle that runs for years.

Abandon, deregister or liquidate. Then hand the company over clean.

Stage 3 is the only one every partner reached. Stage 6 rests on a single account.

Every number in this report, with the sentence it came out of.

This report has no price in it, and it is not going to guess one.

Filter by partner. Click any sentence to open the transcript.

How this was made, what it does not cover, and the five places our instrument got something wrong.

Which partner described which part of the work, and how thoroughly.

Five defects on our side, written down rather than tidied away.

We asked the pen holder what his own document costs him, and then closed the interview over the answer.

We put two things into Denis's answers that he never said.

The read-back is a partial control, because people correct selectively.

Josh's microphone never reached us, and the fault was ours.

The opening was emitted three times to Denis, and Dave's first words came back as Danish.

This report stops at description on purpose.

Four questions this report raises and deliberately does not answer.

Three things we would put a clock on next. Each one is a measurement, not a proposal.

Interview Trevor and Gerhard, and re-ask Dave the one question we dropped.

Give us a charge-out rate, or three of them.

Record what comes out of the next twenty networking conversations. Anything at all.

Six partners described the same sequence. No two of them run it the same way, and nothing outside their heads knows where any of it stands.