The 1940s Waste Audit That Still Diagnoses Your 2026 Revenue Org

A document that has no business still being right

Somebody sat down, probably during or shortly after the Second World War, and wrote out twenty-four questions under a single heading: Wastes of Productive Time.

The questions were written for a factory floor. They ask whether skilled workers are grinding their own tools. Whether machine capacities have been determined and recorded. Whether there is a continuous record of wasted machine time — and whether anyone has made any effort to sell it. Whether tools are kept properly sharpened. Whether the plant has a well-systematized method of scheduling work ahead of each department, work area, and worker, so that nobody sits idle waiting on material that hasn’t arrived.

There is no software in this document. No dashboards, no cloud, no AI. It predates the transistor.

And it diagnoses your revenue organization more precisely than the last four vendor assessments you sat through.

That should be uncomfortable. It means the operational discipline we apply to manufacturing — measured capacity, recorded downtime, fixed responsibility for maintenance, standardized methods, inspection after every operation — has simply never been applied to the function that actually generates the money. We industrialized production. We left revenue as a craft.

Here is what happens when you run the 1940s audit against a 2026 go-to-market org.

Are skilled workers required to do jobs which could be performed by less expensive labor?

The original sub-questions are grinding and sharpening of tools and machine setups. The premise is that a machinist paid for judgment should not spend the morning on preparation any competent helper could do.

Now price your account executive. Fully loaded, a mid-market AE runs six figures. Then look at how the week is actually spent.

Salesforce’s research across thousands of sales professionals found reps spend less than a third of their time on direct selling, with the balance consumed by administrative work, internal meetings, manual research, and CRM upkeep. The same research found that roughly seven in ten reps report being overwhelmed by the sheer number of tools in front of them, and that the overwhelming majority of sales organizations planned to consolidate their stacks in response.

Read that back through the 1940s lens. The skilled worker is grinding his own tools. He is also doing his own machine setup, his own material handling, and — because the CRM is a bad warehouse — his own inventory search. And the plant’s response has been to buy him more tools.

The audit’s implicit answer was not train the machinist to grind faster. It was stop having the machinist grind.

That is the entire argument for autonomous agents in revenue operations, arrived at eighty years early by someone who never met a software vendor.

Idle time, scheduling, and the distance between operations

The audit asks how much productive time is lost waiting for materials, waiting for parts, waiting for tools. It then asks — in the longest single item in the entire document — whether the plant has a well-systematized method of continuously scheduling and recording work ahead of each department, work area, and worker, to prevent idle time and coordinate needs in material, tools, and facilities.

Continuously. Ahead of. Coordinate.

Your pipeline coverage problem is this question. Not a headcount problem — a scheduling problem. When an SDR runs out of qualified accounts on a Tuesday afternoon, that is not a motivation failure. That is material arriving late to the workstation. When an AE’s calendar has a two-hour gap because the meeting that should have filled it was never booked, that is idle time with a fixed hourly cost, and no one is recording it.

It asks, too, whether waste time between operations could be cut by better transportation within the plant. Translate: how long does a lead sit between marketing qualification and SDR touch? Between SDR handoff and AE first call? Between verbal commitment and paperwork? Every one of those gaps is intra-plant transportation, and in most mid-market revenue orgs nobody owns the forklift.

Capacity, and the astonishing question about selling waste time

The audit asks whether machine capacities have been determined and recorded — then whether machines are being worked to capacity, and whether capacity could be raised by minor adjustments.

Ask a mid-market CRO for their revenue organization’s determined and recorded capacity. Not quota. Not headcount. Capacity — the maximum throughput of qualified conversations the current configuration can produce, measured, written down, and compared against actual output.

Very few can answer. Quota is a target assigned downward. Capacity is a property measured upward. They are not the same number, and confusing them is why so much of the market is running plans that were never physically achievable.

The Ebsta x Pavilion benchmark study, drawn from a large pipeline dataset and a survey of thousands of revenue leaders, found missed-quota rates climbing year over year while the spread between top and bottom performers widened — a small fraction of sellers now generating the large majority of revenue, with an order-of-magnitude gap in velocity between the best and the rest.

A factory with that distribution of output across identical machines would not conclude it had hired the wrong operators. It would conclude it had never determined capacity, never recorded downtime, and never standardized method — which is exactly what the capacity, method, and standardization questions are for.

Then there is the item that should stop you. It asks whether the plant keeps a continuous record of wasted machine time. And then, almost as an afterthought:

Any effort to sell waste machine time?

Somebody in 1940-something looked at an idle machine and asked whether the idle capacity itself was a sellable asset. That is a more sophisticated thought about capacity economics than most software pricing models contain today.

Because here is what seat-based SaaS actually does: it bills you for capacity whether or not it is used. You buy twenty seats. Fourteen are logged into regularly. Six are ghosts — offboarded reps, a manager who checks quarterly, a seat bought for a hire that never closed. You are paying full price for waste machine time and you have no continuous record of it.

The 1940s auditor would have flagged that on the first walkthrough.

Preventive maintenance and tool control

The audit asks whether there is an adequate system of preventive maintenance, and immediately drills into the part everyone skips: responsibility fixed. Then periodic inspection and repair. Then proper reporting. Then the hours lost to breakdowns.

It then does the same for tools across nine sub-items — responsibility fixed, time loss analyzed, tool room equipped, tool room neat, tools sharpened, tools standardized, tools assigned properly, record of tools available.

Now consider the average mid-market GTM stack. Somewhere between fifteen and forty tools. Ask the four questions the audit asks:

1. Is responsibility fixed?Who owns the enrichment vendor’s data quality? Name the person.
2. Are tools kept properly sharpened? When was the prompt library last reviewed? The scoring model retrained? The ICP definition updated against actual closed-won data?
3. Is there a record of tools available? Can you produce a complete inventory of every system touching a prospect record, with owner, cost, and renewal date?
4. Is the tool room neat and orderly? Or does every rep have a personal shadow stack of browser extensions nobody approved?

Preventive maintenance is the discipline autonomous systems most obviously need and least often get. An agent that worked well in March will drift by September — the market moves, the ICP shifts, the messaging that converted stops converting. Without periodic inspection, fixed responsibility, and proper reporting, you have not deployed a system. You have deployed a machine with no maintenance schedule and walked away.

Storage, indexing, and complete material lists

The audit asks whether material storage conditions have been analyzed for waste — better location, excess handling, ease of movement, properly indexed for ease of finding, authority fixed for issue of material, and movement to the workplace properly timed and controlled.

That is a description of your CRM, and your CRM fails most of it.

Properly indexed for ease of finding — can a rep locate every prior touch on an account in under thirty seconds? Excess handling — how many times is a single lead touched, re-scored, re-routed, and re-enriched before it dies of old age? Authority fixed for issue of material — who is permitted to pull a lead from the pool, and is that rule enforced or social?

Another item goes to engineering: complete drawings of all products, complete material lists, better standardization of parts, elimination of odd shapes and sizes, better interchangeability.

The go-to-market equivalent is documentation almost nobody maintains. A complete, current, written specification of what you sell, to whom, against which alternatives, with which proof points, and which objections resolved how. Most mid-market companies have this distributed across the heads of three tenured reps and a deck last updated two funding rounds ago.

You cannot automate against undocumented specifications. The single most common reason AI agent deployments underperform is not model quality. It is that the material list was never written down.

Inspection, and the infrared lamp

The audit asks what time is wasted for lack of inspection — specifically, inspection after every operation where salvage could be effected and processing time saved, and whether there are adequate methods of reporting the machine and man-hours already spent on parts that were later rejected.

Applied to revenue: how much human effort is expended on pipeline that was never qualified in the first place? Not lost deals — deals that should have been rejected at the first inspection point and instead consumed six weeks of AE time before dying. The audit’s demand is that you report those hours, so the waste is visible as a number rather than absorbed as bad luck.

For AI-generated outbound, inspection-after-every-operation is not optional hygiene. It is the control that keeps an automated system from scaling a defect. A human writing three hundred bad emails is a coaching conversation. An unsupervised agent writing three hundred thousand is a brand event.

One item dates the document most charmingly and ages the best. It asks whether you have analyzed new timesaving technological developments for feasible application — and offers an example: infrared lamps for drying paint.

That was the frontier technology of the moment. The question is not about infrared lamps. It is about whether the organization has a standing process for evaluating genuinely new capability against its own operations, rather than adopting on hype or refusing on reflex.

Most mid-market companies still do not have that process for AI. They have a pilot, a champion, and a hope.

What is your total estimated waste?

The audit closes by demanding a number. Idle time of workers. Idle time of machines. Deliberate curtailment of production. Other factors enumerated in this section.

This is the question that turns an interesting exercise into a budget conversation, and it is the one to bring to your CFO. Not should we buy AI. Instead: here is our determined capacity, here is our actual output, here is the recorded gap, here is what the gap costs, and here is the mechanism to close it.

How PrescientIQ™ maps to the audit

We built PrescientIQ™ as an autonomous multi-agent revenue operating system — twelve agents across three stacks, coordinated by a single orchestrator. The architecture was not derived from this checklist, but it answers the checklist item by item, because the underlying problem has not changed in eighty years.

Coverage — the answer to idle time and scheduling

The audit's idle-time items are coverage questions: material arriving late, work not scheduled ahead of the worker, time lost in transit between operations. The Coverage stack runs continuous account research, ICP qualification, signal monitoring, and sequencing so the workstation is never starved and nothing sits in transit between stages.

Constraint — the answer to capacity and maintenance

Its capacity and maintenance items are constraint questions: what is the machine’s determined capacity, is it running at capacity, what is its downtime, who maintains it. The Constraint stack instruments throughput, records idle capacity continuously, and applies fixed-responsibility maintenance to the agents themselves — periodic inspection, drift detection, reporting.

Consideration — the answer to inspection and specification

Its inspection and specification items are consideration questions: is the work inspected, is the specification complete, is new capability evaluated. The Consideration stack holds the material list — ICP definitions, proof points, objection handling, disclosure requirements — and inspects output before it reaches a buyer.

And on that remarkable prompt about selling waste machine time: this is precisely why PrescientIQ is priced as Labor-as-a-Service rather than per seat. You are not buying capacity and hoping to use it. You are buying completed work. Idle capacity is our problem, not a line on your invoice.

One question the 1940s could not ask

The original document has no item for disclosure. It never occurred to the author that a machine’s output might need to announce itself as machine-made, because no machine in 1945 could produce something a customer would mistake for a person.

That gap closed on 2 August 2026, when the transparency obligations under Article 50 of the EU AI Act (Regulation (EU) 2024/1689) became applicable. The European Commission adopted interpretive guidelines on 20 July 2026. Among the obligations: systems that interact directly with people must disclose that fact, and certain generated or manipulated content must be marked. Legal analyses of the regime note that non-compliance can carry penalties reaching into the millions of euros or a percentage of worldwide annual turnover, and that the reach extends to organizations outside the EU whose AI output is used within it.

Read against the 1940s audit, Article 50 is the inspection item with legal force. It is inspection-after-every-operation, made mandatory, with a reporting requirement attached.

If you are running autonomous outbound into European buyers, disclosure is now an operating step in the workflow — not a legal footnote. That is a full pillar in its own right, and we treat it as one.

The uncomfortable conclusion

The 1940s auditor was not smarter than you. They had a smaller problem and a better habit.

The habit was this: assume waste exists, go find it, write down the number. Not assume the team is trying hard. Not assume the tools are working because you bought them. Assume waste, quantify it, fix the largest item, repeat.

Twenty-four questions. Most mid-market revenue organizations cannot answer twelve of them with a number.

Start there. The technology conversation gets much simpler once you know what the waste is worth.

Sources

• Salesforce, State of Sales research on seller time allocation and tool overload — salesforce.com
• Ebsta x Pavilion, 2025 GTM Benchmarks Report — press release · full report PDF
• Regulation (EU) 2024/1689 (AI Act), Article 50 — EUR-Lex
• Cooley LLP, analysis of Article 50 applicability from 2 August 2026 and penalty exposure — cooley.com
• European Commission, Article 50 transparency guidelines adopted 20 July 2026 — digital-strategy.ec.europa.eu