Marcus Cauchi

Saying they will speak to the CFO is an intention.

• Saying afterwards that they did is the buyer's account of something the seller did not see.
• The CFO joining the next meeting is something the seller observed.
• The CFO approving the spend is a decision.
• These are not four versions of the same thing, Cauchi argues, but ordinary handovers flatten them: an account executive updates the record, a manager repeats it in a forecast call, a sales leader shortens it further, and 'I'll speak to the CFO' reappears as 'CFO engaged'.
• Nobody has to lie for this to happen.
• Each person compresses what they were told into something shorter and easier to use, often dropping the uncertainty, so the next person inherits the confidence without the doubt that came with the original evidence.

'I'll speak to the CFO' becomes 'CFO engaged' after a few handovers, even though no one misreported anything.

Later posts apply the same ladder to AI-written CRM updates, asking whether the software can keep the four levels apart.

Read original · Original 2 · Original 3

When deals sit in Commit, people outside sales read the total as revenue the business expects, not revenue salespeople judge likely.

• Delivery reserves capacity, finance commits spend, hiring plans change, and cash timing gets treated as settled.
• Cauchi's example is a £500,000 deal in Commit that is genuinely well evidenced but still has one outstanding approval.
• If finance hires two people against it, the company has made a second decision on the back of the first; a slip now costs the number plus two salaries it might not otherwise have taken on.
• This can happen with an accurate CRM and no forecast inflation, Cauchi argues.
• The question is not whether sales will hit the number, but who else is relying on it.
• Cauchi also separates two things a forecast mixes up: good evidence that a customer will buy is not the same as knowing when they will sign, when you can invoice, and when the money arrives.

Hiring two people against a £500,000 Commit deal turns a sales judgement into a payroll commitment.

In a post revisiting earlier writing, Cauchi says they stopped too soon. The first argument was that CRMs describe the seller better than the buyer; the second asks what everyone else in the business is entitled to assume because a deal is in the forecast.

Read original · Original 2 · Original 3 · Original 4

A dashboard records calls, emails, meetings, demos, proposals, follow-ups and stage changes—all things the seller did.

• It rarely records who the buyer spoke to internally, who joined the decision, what the buyer investigated or agreed to spend, or whether they did what they said they would do.
• Some of that sits in email, chat or a salesperson's memory; some happened inside the customer's business, out of view.
• Cauchi does not ask sales to know everything.
• The problem starts when unknowns are treated as knowns: a proposal, three follow-ups and a manager review produce several recorded events, while in the same week the buyer loses their finance lead's support and finds a security problem that could stop the purchase.
• Activity rises while knowledge of the buying decision stays flat, which is why Cauchi is reluctant to call a busy opportunity a progressing one.

A week of logged follow-ups can coincide with the buyer's internal support collapsing, with none of it in the record.

Read original · Original 2

Evidence created by the seller's own effort vanishes when they stop working the deal, while a buyer who has gone to their own finance team, taken on internal risk or changed their requirements has produced something that stands on its own.

• Cauchi frames the wider failure the same way.
• Forecasts rarely break at quarter end; by then the damage is done.
• They break earlier, as evidence turns into interpretation and interpretation quietly turns into confidence, driving hiring, spending, capacity and board expectations that rest on more certainty than the original observation supported.

Ask whether the deal's evidence would survive the salesperson leaving the account alone for a month.

Read original

Their condition is sequencing.

• If a system is handed statements at different levels of evidence and writes them all as some version of 'economic buyer engaged', it makes the record tidier and less useful—and does so faster than a person could.
• The same collapse shows up in dates and approvals: 'we're hoping to get this done in October' is not a close date, 'send me the paperwork' is not approval, 'that sounds reasonable' is not commercial agreement, and 'I'll take this to the board' is not a board decision.
• Cauchi wants the distinctions chosen before capture is automated: taking ambiguity already buried in sales data and making it beautifully consistent leaves you with consistent rubbish.

Four CFO statements fed to an auto-updating CRM, with the test being whether the resulting fields keep intention, report, observation and decision apart.

Read original · Original 2

A salesperson tells a forecast meeting that a £500,000 deal is weaker than they thought, and the response is a burst of questions: why, what changed, can we save it, who else can we involve, what is the recovery plan, and where is the replacement £500,000 coming from.

• If they had said nothing and left the deal where it was, almost nothing would have happened that day.
• A business that says it wants bad news early can make delivering it considerably harder than leaving yesterday's answer alone.
• Nobody needs to manipulate the CRM for this to affect the data: people learn which answers create grief and which let them get on with their day.
• Cauchi checks what happens when someone makes the number smaller before blaming record-keeping discipline.

Reporting a deal as weaker draws a recovery interrogation that staying quiet would have avoided.

Read original

A company buys a CRM: sales liked the demo, the business case worked, the board approved it and procurement closed it.

• Then the conditions show up.
• The data was never cleaned, managers still run pipeline from spreadsheets, salespeople treat the system as admin, marketing and customer success use different fields, and the sponsor has moved on.
• Three months later, everyone is unhappy and the supplier is blamed, though the software was not necessarily the problem.
• Cauchi links this to ghosting, which they read as the buyer seeing the risk before signing, and buyer's remorse, which is the buyer finding it afterwards.
• They propose that sales test more than interest, pain and budget, and ask who will own the change, what has to change, who will resist it, and what happens once the excitement of signing has gone.

A CRM that closes well but is never adopted leaves customer success to inherit the fallout.

Read original

A buyer asking for pricing might be showing intent, comparison shopping, doing internal research, performing a procurement process, checking a budget or simply being curious.

• Rather than decide which story is true, they offer a fill-in frame: the claim you are tempted to make, the evidence you have, the evidence you are missing, what you should not trust yet, a choice between advancing, fixing or exiting, and a three-day reality check.
• The structure is the point: it forces the gap between the story and the evidence into writing before anyone acts on it.

The frame's closing question, 'what are you about to trust this week that has not earned it yet?', is what the exercise is built to answer.

Read original · Original 2

This post supplies the vocabulary the rest of Marcus Cauchi's writing builds on — the difference between an intention, a reported action, an observed event and a decision.

Cauchi takes one sentence a buyer might say — "I'll speak to the CFO" — and shows four versions of it, each carrying different weight.

• The buyer saying they will speak to the CFO is an intention.
• The buyer saying afterwards that they did is a claim about something you did not see.
• The CFO joining the next meeting is something you observed yourself.
• The CFO approving the spend is a decision.
• Cauchi then follows those four items through an ordinary sales organisation: the account executive updates the record, a manager repeats it in a forecast call, a sales leader summarises it, finance includes it, the board sees it.
• Nothing dishonest has to happen for "I'll speak to the CFO" to end up recorded as "CFO engaged".
• Each person compresses what they were told into something shorter, and the uncertainty is usually what gets dropped.
• Cauchi's suggested response is practical rather than philosophical: take ten important opportunities and trace claims such as "economic buyer engaged" or "budget approved" back to whatever actually happened that earned the label.

For brands planning a campaign

This is the strongest opening piece for a webinar or article series aimed at revenue operations and sales leadership. It sets up a simple exercise audiences can run on their own pipeline that same week, giving a sponsor a natural follow-up: a session where attendees trace their own records and report what they found.

Useful for anyone weighing an AI sales tool.

Cauchi starts from the same four levels of evidence, then applies them to automation.

• If a stated intention, a reported conversation, an observed meeting and an actual approval all get written up as some version of "economic buyer engaged", the software has made the record tidier and less useful at the same time.
• The post lists comparable collapses in plain language: "We're hoping to get this done in October" is not a close date, "Send me the paperwork" is not approval, "That sounds reasonable" is not commercial agreement, and "I'll take this to the board" is not a board decision.
• Cauchi is not arguing against the tools.
• The argument is about sequencing: decide which distinctions matter before you automate their capture, or you will scale the ambiguity already sitting in the data.
• The closing line — that beautifully consistent rubbish is still rubbish — makes the point in one sentence.

For brands planning a campaign

For a revenue operations or CRM software sponsor, this is a caution rather than a sales pitch.

• It works better as a candid expert segment — "what to check before automating capture" — than as a product endorsement.
• The practical checklist can become a short assessment tool for audiences to complete before a webinar.

The post asks readers to compare this quarter's Commit figure with what happened to comparable Commit at the same point in earlier quarters, and offers a hypothetical: if 75% of that value normally closed, that is a useful figure to have.

• Cauchi says the £4m will not simply become £3m, because deals differ and one large deal can change the result.
• But ignoring the history does not make the full £4m more likely either.
• The problem is not the sales team's judgement on individual deals.
• It is what happens when those judgements are added up and other departments read the total as revenue expected rather than revenue judged likely.
• Delivery reserves capacity, finance commits spend, and cash timing — when the customer signs, when you can invoice, when the money arrives — is treated as settled by a figure that only addresses whether the customer will buy.
• The post ends with a question: ask not whether sales will hit the number, but who else is relying on it.

For brands planning a campaign

Lead with this for a finance, board or chief revenue officer audience, framing it around the forecast's downstream consumers rather than its accuracy. It also suits a roundtable: the questions at the end are designed for people in different functions to answer.

Cauchi contrasts what a sales dashboard records well — calls, emails, meetings, demos, proposals, follow-ups, stage changes and forecast updates — with what it rarely records: who the buyer spoke to internally, who became involved, what the buyer investigated or agreed to spend, what changed their requirements, and whether they did what they said they would do.

• Some of that sits in email, some in chat, some only in the salesperson's memory, and some happened inside the customer's business entirely out of view.
• Cauchi says that is acceptable; the trouble starts when unknowns are treated as knowns.
• A proposal sent, three follow-ups and a manager review generate several recorded events, while the buyer could be losing support from their finance lead and discovering a security problem that could stop the purchase — with nothing recorded.
• Activity rises while knowledge of the buying decision stays flat, which is why Cauchi is reluctant to call a busy opportunity a progressing one.

For brands planning a campaign

Pairs naturally with the facts-ladder post as a two-part session: one half on what sellers record, one half on what buyers actually do. For a sponsor, the useful angle is not more activity tracking but which buyer-side facts are worth capturing and how.

Cauchi sets up a scene in a forecast meeting.

• A salesperson says a £500,000 deal is weaker than they thought.
• The response is a burst of questions: why, what changed, can we save it, who else can we involve, what is the recovery plan, where is the replacement £500,000 coming from?
• Then they ask readers to imagine the salesperson had said nothing and left the deal where it was, pointing out that today almost nothing would have happened.
• Cauchi's conclusion: a business that says it wants bad news early can make delivering that news far harder than leaving yesterday's answer alone.
• Nobody needs to manipulate the record for this to affect the data; people simply learn which answers create grief.
• Management then looks at a pipeline and assumes it is seeing an accurate picture.
• Cauchi's instruction is to examine what happens when someone makes the number smaller before blaming record-keeping discipline.

For brands planning a campaign

Aimed at sales managers and revenue leaders rather than tooling buyers, this works best as a workshop or podcast segment about the forecast call itself. A software sponsor should be careful: the piece argues the problem is behavioural, and a campaign that reduces it to a product feature will look like it missed the point.

This is the clearest short statement of the position a newcomer should understand before reading anything else.

Cauchi revisits an earlier article, saying they still agree with it but stopped too soon.

• The earlier argument was that most CRM systems tell you far more about what the seller has done than about what is happening on the customer's side.
• The addition is a second question: what is everyone else in the business entitled to assume because a deal sits in the forecast?
• The example is a £500,000 deal in Commit where the customer wants to proceed, the commercial case is agreed and one internal approval is outstanding.
• That may be a fair judgement.
• But if finance hires two people against the expected revenue, the company has made a second decision on the back of the first.
• If the deal slips, it misses the number and carries two salaries it might not have taken on.
• Cauchi stresses this can happen with an accurate system and no forecast inflation.
• They also note that the seller's confidence and the uncertainties they still carry do not transfer with the record, and that AI summarising an already-blurred record may only produce a cleaner version of the same judgement.
• The line to keep is Cauchi's: a deal can be worth pursuing without being safe to hire against.

For brands planning a campaign

This is the best single piece for a sponsor to build an article series around: it names the problem and its extension in one place. A webinar can pick up the second question directly and ask the audience what their own Commit label entitles other departments to assume.

The most immediately usable piece for someone who wants to apply the approach.

Cauchi takes a familiar moment — a buyer asks for pricing — and lists what it might mean: genuine intent, comparison shopping, internal research, procurement theatre, a budget check or simple curiosity.

• Rather than resolve which, the post offers a fill-in frame: the claim you are tempted to make, the facts you have, the facts you are missing, what you should not trust yet, a choice between advancing, fixing or exiting, and a three-day reality check.
• The structure forces the gap between the story and the facts into writing before anyone acts on it.
• A near-identical version appears elsewhere here, suggesting this is a repeated format rather than a one-off reflection.

For brands planning a campaign

The obvious campaign asset. The frame can become a downloadable worksheet or slide deck, giving a sponsor something portable that audiences can use after the event.

Extends the topic beyond forecasting to whether the purchase holds.

Cauchi describes a familiar sequence: a company buys a new CRM; sales liked the demo, the business case made sense, the board approved it and procurement closed it.

• Then the problems appear.
• Nobody has cleaned the data, managers still run pipeline from spreadsheets, salespeople treat the system as admin, marketing and customer success use different fields, and the executive sponsor has moved on.
• Three months later, everyone is unhappy and the supplier takes the blame, though the software may not be the problem.
• Cauchi links this to two behaviours: ghosting, which they read as the buyer seeing the risk before signing, and buyer's remorse, which is the buyer discovering it afterwards.
• Their proposal is that sales test more than interest, pain and budget: whether the buyer can actually make this work, who owns it, what has to change, who will resist, and what happens once the excitement of signing fades.
• If sales leaves those questions untested, customer success inherits the mess.

For brands planning a campaign

Shifts a campaign idea downstream, toward customer success and implementation audiences. It also gives a sponsor a defensible reason to talk about onboarding rather than acquisition, a less crowded subject in this space.

Their examples show how those distinctions can blur as information moves from a salesperson to a manager, forecast, finance team, or board—even when nobody is being dishonest.

Compare specific buyer statements and actions, then ask what evidence supports each CRM field. They also propose tracing important claims in ten opportunities back to their source.

For brands planning a campaign

A CRM or revenue operations brand could build a practical webinar or article series on preserving the source and certainty of sales claims. Use examples such as “I’ll speak to the CFO” and “the CFO approved the spend” to show what each statement does—and does not—support.

Examples: Source 1 · Source 2 · Source 3

A deal can be worth pursuing or included in Commit without being safe to treat as certain. They recommend weighing comparable historical outcomes alongside the evidence and remaining steps for each deal.

Ask not only whether Sales expects a deal to close, but who relies on the forecast, what is established with the customer, what remains unresolved, and what decision the forecast supports.

For brands planning a campaign

For a forecast-integrity campaign, show how to connect deal-level evidence and historical outcomes to decisions by Finance, delivery, and leadership. Make clear that a historical close rate is useful context, not a guaranteed prediction for a different set of deals.

Examples: Source 1 · Source 2 · Source 3 · Source 4

A busy opportunity may show more activity without stronger evidence that the buyer is progressing.

Reconstruct two histories for an opportunity: what the seller did and what the buyer demonstrably did. Compare them with the evidence available at each stage or forecast change.

For brands planning a campaign

A sales analytics or CRM company could use this distinction to explain why activity dashboards alone may not establish deal progress. A useful demonstration would place the seller’s timeline beside buyer actions that can be evidenced, while leaving unknowns visible.

Examples: Source 1 · Source 2

Their examples distinguish an intention to speak with the CFO, a reported conversation, the CFO joining a meeting, and the CFO approving expenditure. They make the same point about hoped-for dates, paperwork, commercial agreement, and board decisions.

Before enabling automatic updates, give the system statements representing different levels of evidence and check whether its resulting fields preserve those distinctions.

For brands planning a campaign

An AI-for-sales campaign could demonstrate how a system captures the source of a claim and keeps reported, observed, and decided events distinct. This is relevant to CRM automation because it turns an abstract data-quality concern into a test buyers can run on real statements.

Examples: Source 1 · Source 2

Testing interest, pain, and budget is not enough if the customer cannot make the solution work under their actual conditions.

Ask who will own the change, what must change, who may resist, and what could get in the way after the excitement of signing fades.

For brands planning a campaign

A customer success, implementation, or revenue operations brand could use this perspective to discuss qualification beyond purchase intent. The strongest angle would show how pre-sale questions about ownership, change, and likely obstacles can reduce the risk of a deal closing but failing to hold.

Examples: Source 1

If a salesperson faces blame or pressure after reporting that a deal weakened, they may learn that leaving an earlier, more confident record is easier. They look beyond data-entry rules to the management response that shapes what people report.

Observe what happens when a salesperson reduces a deal’s value because the evidence changed, then ask whether the business makes early correction safer than leaving the record untouched.

For brands planning a campaign

A sales leadership or revenue operations campaign could pair data-quality practices with a discussion of how forecast meetings respond to bad news. This adds a people-and-process angle to a technology-led message about CRM accuracy.

Examples: Source 1

They offer a compact prompt that asks what claim is being made, what evidence exists, what is missing, what not to trust yet, and what to do next.

Record the claim, available evidence, missing evidence, what not to trust yet, whether to advance, fix, or exit, and a short reality check.

For brands planning a campaign

This format could support a practical sales enablement post, workshop exercise, or downloadable checklist. It gives the audience a concrete way to apply the broader evidence-versus-inference idea to a live opportunity.

Examples: Source 1 · Source 2

The audit post follows a signed deal that nobody can reconstruct afterwards — which quote was final, who approved the discount, when terms changed, whether the CRM matches what the customer signed.

• The post on first RevOps hires argues that definitions, stage criteria and ownership of each field must come before reporting.
• That is Cauchi's concern about records that drift from what happened, approached through governance and audit trails.

Spira tends to write shorter, more promotional posts that lead into a named product, with arguments about process design and the sequence of RevOps work.

• Cauchi focuses on what a buyer's sentence supports and who else in the business is about to act on it.
• They also ask about the behaviour that shapes data — what happens when someone reports a deal as weaker — while Spira focuses on systems that preserve the trail.

Read original · Original 2

His example is a sales leader with 80 deals scanning fields, activity, notes and calls to work out where to get involved. Like Cauchi, Remedios wants human judgement spent only where it changes the outcome, and the decision, action and outcome recorded so the system can learn from them.

Remedios argues for rebuilding GTM around AI from the start — state, memory, skills, decision traces — so the system understands the operation continuously.

• Cauchi is more cautious about what automation conceals.
• Their fix is a discipline rather than an architecture: keep intention, report, observation and decision apart before anything is summarised, and choose those distinctions before capture is automated.
• They also write in a plainer, non-product register.

Read original · Original 2

A chief executive with $14m of qualified opportunities is left with $4m once every open deal is tested against real milestones: was the economic buyer identified and engaged, were decision criteria documented beyond price, was the outcome tied to a measurable result, and were all decision makers mapped? Cauchi's test is similar in spirit: ask what a deal's evidence would support rather than whether the field has been filled in.

Hines reaches the point through a consulting case study with before-and-after numbers and a defined sales process to coach to.

• Cauchi works through short reframings of everyday statements, without case-study metrics or process-selling language.
• They are more interested in who outside sales is relying on the number than in the milestones themselves.

Read original · Original 2

Spira writes directly to RevOps practitioners.

• One post explains how a signed deal should be reconstructable from what was proposed, what changed, who approved it, when, and what was contracted, then ends with the audit question that exposes a missing trail.
• Another sets out an order of operations for a first RevOps hire—definitions, then architecture, then process, then measurement, then leverage—and offers a concrete interview test around stage definitions.
• If the deliverable is a working session on deal governance and audit-ready records, this is closer to the operational detail a RevOps audience expects.

The material stays on governance, not the judgement being governed.

• Neither post examines why a confident forecast can rest on thin evidence, the gap between what a buyer said and did, or how uncertainty shifts from seller to forecast call to board number—the substance Marcus Cauchi's campaign is built on.
• Both posts close with direct company promotion: one names DealHub.ai as building the governed path it describes; the other carries the writer's own consultancy framing.
• For an educational campaign run by a software sponsor, that may read more like a vendor message than the sponsor wants.

Examples: Source 1 · Source 2

Remedios argues that AI in GTM should start with the work itself—why the buyer needs to act, who cares, how the decision will be made, where the risk sits—and work back to what a system needs to hold.

• Their posts set out an operating layer of state, memory, skills, agents and decision traces, using a VP Sales staring at 80 deals to show how scarce human judgement can be routed only to the decisions that matter.
• For a sponsor pitching an AI-native operating model, that architecture framing is the more direct fit.

The argument makes AI the destination.

• It calls for rebuilding the stack around state and memory so the system can decide and act, assuming the underlying evidence is worth acting on—the question Marcus Cauchi keeps asking, and the one that determines whether an AI capture layer makes records better or just faster.
• The writing is also tied to a named platform and leans on architecture vocabulary.
• Readers looking for a self-runnable test of a CRM claim, or a close reading of what a Commit label entitles the rest of the business to assume, will find less of that here.

Examples: Source 1 · Source 2

One post tells of a mid-market manufacturer whose pipeline fell from $14M to $4M when every open deal was held against real milestones, followed by reported forecast accuracy moving from 62% to 89%, a shorter cycle and a higher win rate.

• The other contrasts Microsoft and Meta on AI spending to make a leadership point about clarity over hope.
• For sales leaders seeking pipeline-discipline coaching and a metrics-led turnaround narrative, this offers more than Marcus Cauchi's material does.

The proposed mechanism is defined milestones and manager coaching—a process fix—rather than the distinction between documented facts and inference at the heart of Marcus Cauchi's campaign.

• The figures come from a recounted client story.
• Ask for supporting detail before treating them as a basis for a decision.
• The posts also close on the writer's own firm, pulling a neutral educational asset toward a consultancy pitch.

Examples: Source 1 · Source 2