What a CMO Evaluation Actually Produces
Sage's 2024-25 study of 3,000 businesses across 11 countries found that the average Australian business spends approximately 81 working days a year, roughly 567 hours, on admin tasks at an annual cost of about AUD 79,000. That figure is the size of the prize an activity-level diagnostic makes visible. It is also the size of the gap most operating-model conversations skip past.
Last week we introduced the CMO Evaluation, our productised Cost Model Optimisation diagnostic for Australian and New Zealand businesses in the 5 million to 50 million dollar revenue band. (Distinct, in case the abbreviation has caused confusion, from a pharmaceutical Contract Manufacturing Organisation review, a Chief Marketing Officer performance audit, a Chief or Career Medical Officer review, a Community Managed Organisation evaluation, and the Context-Mechanism-Outcome realist evaluation framework used in policy research.)
This piece answers the question buyers ask after reading the launch. What does the diagnostic actually produce, and what does an operator do with the output on the Monday after the engagement closes?
What the CMO Evaluation produces for a scaling business in Australia or New Zealand
The CMO Evaluation produces three artefacts. They sit on the table at the end of the engagement, and the operator can act on them without commissioning a follow-on programme.
The first is an activity map: a documented decomposition of the work being done in the business, organised by function, frequency, regulatory sensitivity, and value contribution. The second is a three-layer activity-level allocation that places each captured activity into one of three categories: AI-augmented, offshore-restructured, or locally-retained. The third is an implementation sequence: a staged 90-day plan for moving from the current operating model to the redesigned one, including role profiles, governance cadence, and the order in which activities should change hands.
The artefacts are designed to be implementation-ready. The phrase is a precise claim. Implementation-ready means an output a leadership team can act on in the week the diagnostic closes. The activity map names actual activities. The layer allocation names which activity goes where. The sequence names which activity moves first, and what governance protects the change.
The diagnostic does not run the implementation. It produces the plan an operator runs from. The distinction matters, because the most common failure pattern in operating-model work, documented across Bain and BCG research and visible in our own engagement base, is the diagnostic that produces "opportunities" and stops. Opportunities decay into PowerPoint. Sequences become Monday-morning decisions.
The activity map: what a defensible decomposition actually looks like
Reejig, the Australian-founded workforce intelligence platform backed by Bersin research, published findings in mid-2025 showing that a single role typically decomposes into around 77 tasks and 225 sub-tasks once you map it at the level of granularity an AI-exposure analysis requires. That is the public benchmark for what serious role decomposition looks like in 2026.
Outrun's methodology is grounded in over 90,000 captured activity data points across our concurrent engagement base, with more than 5,000 new activities added each month. That data set is what underwrites the activity map. The diagnostic does not start with a generic taxonomy and impose it on the business. It captures activities the way the business actually performs them, then sorts them against the four axes above.
For a scaling business with 10 to 100 employees, a credible first-pass map documents hundreds to low thousands of distinct activities. The number sounds large until you remember Reejig's per-role figure. A 50-employee business has a lot of work happening, and most of it has never been written down in this form before.
PMVA, an Australian property-management offshore provider, published a customer case in 2024 showing Phil Jones, Principal of Propel Realty in Brisbane, restructuring more than 20 processes covering over 300 daily and monthly tasks across his agency. The case is third-party-published evidence, useful here because it shows the order of magnitude in a real ANZ business at activity level. Once the activities are visible, the next decision is where each one belongs.
The three-layer activity-level allocation: AI-augmented, offshore-restructured, locally-retained
The three layers are not a menu. They are the categories every captured activity is sorted into.
AI-augmented activities are the high-frequency, rule-based, low-judgement work where current AI tooling already performs at or above the level of a human doing the same activity at speed. Reconciliation, document classification, first-pass drafting, scheduling, structured data entry. These activities do not move offshore or stay onshore. They get supported by AI tooling that an offshore or local team member runs, with the human in the loop at the points that need judgement.
AI in accounting is at scale in this region already. Wolters Kluwer's 2025 Future Ready Accountant report, drawing on responses from 2,768 tax and accounting professionals across 14 countries, found that 81 percent of APAC accounting firms use AI weekly. The APAC sample combines Australia, New Zealand, Malaysia, and Singapore, with Southeast Asia running ahead of Australia and New Zealand on the headline figure. The diagnostic meets that adoption where it is.
Offshore-restructured activities are the structured, recurring, documentable activities that benefit from dedicated specialist attention but do not require local presence or local market judgement. Bookkeeping reconciliation, claims pre-processing, customer service ticketing on documented workflows, transaction-level financial reporting. These move to dedicated offshore team members in the Philippines on a zero-markup employment model.
Locally-retained activities are the judgement-heavy, regulated, relationship-driven, or strategically-sensitive activities that a local team member is best positioned to perform. Client conversations, regulatory sign-off, deal structuring, strategic prioritisation, anything that requires being in the room.
Most roles split across all three layers. That is the point. A bookkeeper's role is not one layer or another. It is a stack of activities, each of which has a natural home.
External validation for this structure is published and accessible. The Australian accounting industry has converged on a hybrid staffing convention of around 30 to 40 percent onshore and 60 to 70 percent offshore, with reported staffing-cost savings of 40 to 60 percent at the whole-firm level. That convention appears in independent commentary from Accritic, Credfino, Future Firm, and Outsourced Accountants. It is industry consensus across independent commentators.
For comparison at the enterprise tier, PwC Australia's October 2025 Organisation Design for the Digital Age service, delivered with Orgvue, applies AI and automation impact analysis to role-level organisation redesign for global-scale clients. The CMO Evaluation operates at activity level, at mid-market scale, on a productised timeline.
A note on cost-savings figures, because three different ones live in the public conversation and they are not interchangeable. Outrun's 60 to 75 percent labour-cost-savings band is measured against the cost of hiring the same role locally in Australia or New Zealand. The accounting hybrid-staffing convention's 40 to 60 percent figure is measured against the on-staff cost of a fully retained local team. The PMVA Phil Jones case reports a 60 to 70 percent role-level saving against an Australian assistant property manager salary. Each is defensible inside its own comparison basis. None of them should be read as the others.
From activity map to implementation: the sequence that makes the diagnostic actionable
Harvard Business Review's 2009 study of 5,400 new leaders found that the most common executive failure mode is focusing too heavily on quick wins at the expense of overall direction. The Quick Wins Paradox is one of the better-replicated findings in transformation research. It is also the reason a diagnostic that produces "opportunities" without a sequence quietly underperforms.
McKinsey's transformation database supports the corollary. Top-quartile companies capture 74 percent of transformation value within the first 12 months. The average transformation realises just over half of its expected value within 18 months. The difference is sequencing, not speed.
The CMO Evaluation closes with a 90-day implementation sequence in this order:
Activity capture and categorisation. The activity map is finalised. Activities are sorted against the four axes (type, frequency, regulatory sensitivity, value contribution).
Layer allocation. Each activity is placed in AI-augmented, offshore-restructured, or locally-retained.
Process documentation for the activities that are moving. Activities that change hands need to be documented before they move. Skipping this step is one of the four failure patterns we see most often. Documentation is what allows the work to transfer cleanly and the AI augmentation to be applied consistently.
AI augmentation of the well-bounded routine activities. The augmentation rolls out where the tooling is proven, against the activities the business has already documented.
Offshore restructure of the documented recurring activities. Dedicated offshore team members come on at this point, against documented processes and against AI-augmented workflows.
The McKinsey "back on track" article from August 2025 puts realistic numbers on the leadership commitment. Effective operating-model transformation typically requires roughly four hours per week from each top-team member, sustained across a six-to-twelve-month main phase. The CMO Evaluation produces the plan inside the diagnostic window. The implementation runs over the months that follow, on the operator's clock and against their priorities.
Local re-allocation is the sixth move, slightly outside the 90-day sequence. The hours released by the first five steps go back into the activities the diagnostic flagged as undersourced: client work, strategic priorities, the activities that grow the business. That re-allocation is where the cost rebase becomes a growth lever rather than a cost-cut.
How the diagnostic compares to a Big 4 transformation engagement
A reasonable buyer question, particularly at the board level, is whether a productised mid-market diagnostic substitutes for a Big 4 engagement.
It does not. They are different products built for different scales. A Big 4 operating-model engagement (PwC Australia + Orgvue at the enterprise tier, KPMG Powered Enterprise, Deloitte Operating Model Transformation) is partner-led, scoped open-ended, priced from the high six figures upward, and runs 12 to 24 months. It is the right product for an ASX 100 organisation reorganising at scale.
The CMO Evaluation is a productised diagnostic for a 5 million to 50 million dollar business that needs an activity-level view of its operating model and a sequenced plan. It is scoped, time-bound, and priced inside the band a mid-market CFO can sign off without a board paper.
The honest framing the research base supports is this. McKinsey's 2025 survey of 2,000 executives found that 63 percent of operating-model redesigns now meet most of their objectives, up from 21 percent a decade ago, with completion rates lifting from 51 percent to 79 percent over the same period. The discipline has matured. Bain's research across 470 companies finds that only about 12 percent of cost-transformation results are sustained at full value. BCG's May 2025 work documents 10 to 20 percent leakage between projected and realised savings on most operating-model programmes.
The realisation gap is real. The CMO Evaluation is designed against it. The activity-level evidence base sharpens the diagnosis. The documentation step in the sequence reduces the rework that drives leakage. The 180-day re-recruitment guarantee covers the talent-fit dimension of the realisation gap, which is the dimension Outrun has the most direct control over. If an offshore placement is not the right fit inside the first six months, we re-recruit at no additional fee.
From the artefact to the path: how to start
Most readers of this piece are not yet ready to commission a paid diagnostic. That is fine, and the path is staged accordingly.
The first step, free, is the Outrun Calculator. It estimates indicative savings from a small number of inputs about the business. It produces a starting figure, useful for orientation.
The second step, also free, is an Activity Analysis Session. It is a 45 to 60 minute conversation with an Outrun consultant. The output is a one-page activity-level read on the business and a candid view on whether a CMO Evaluation is the right next step. There is no obligation either way.
The third step is the CMO Evaluation itself. The entry-tier diagnostic starts at AUD 750 and runs across about two weeks. The full engagement is scenario-priced based on business size and scope, typically running four to ten weeks. Pricing and timeline are confirmed at scoping, before any work starts.
The fourth step is implementation, which the diagnostic produces the plan for and the operator runs.
Frequently asked questions
What does an operating-model audit actually deliver for a mid-market Australian or New Zealand business?
A structured operating-model diagnostic produces an activity-level inventory, a three-layer allocation map (AI-augmented, offshore-restructured, locally-retained), a cost-rebase model, an implementation sequence, and a 90-day execution plan. The output is implementation-ready: a sequenced plan a leadership team can act on the week the diagnostic closes.
How long does an operating-model assessment take?
The free Activity Analysis Session is a 45 to 60 minute conversation. The entry-tier diagnostic runs around two weeks. A full CMO Evaluation typically runs four to ten weeks, depending on business size and scope.
How do you decide what work goes onshore, offshore, or to AI?
At activity level, not role level. Each captured activity is allocated against four axes: type, frequency, regulatory sensitivity, and value contribution. AI-augmented for high-frequency rule-based activities, offshore-restructured for structured recurring work, locally-retained for judgement-heavy or regulated activities. Most roles end up split across all three layers.
What is activity-based costing in workforce planning?
Allocating cost to discrete activities the business performs, rather than to roles. It surfaces the cost patterns that role-level budgeting hides. For a scaling business, the question shifts from "how do we cut headcount" to "where does each unit of work belong".
How is this different from a financial audit or a process audit?
A financial audit looks backward at records. A process audit looks at how specific workflows run. An operating-model diagnostic looks forward at how activities should be allocated across AI, offshore, and local layers before the next phase of scaling. Three different questions, three different moments in the business lifecycle.
The activity map is the artefact. The three-layer allocation is the decision. The implementation sequence is the path. They are produced together, inside a window measured in weeks, by a diagnostic priced for a mid-market CFO. If the question on the table is whether to commission one, the free Activity Analysis Session is the first conversation.
