Start with the drivers

Two businesses with the same revenue and profit can be worth very different amounts, and a generic model template will not tell you why. The reason is usually sector-specific: e-commerce value sits in cohort retention and acquisition economics; construction value in the quality of the work-in-progress book; property value in lease expiry and capitalisation rates.

The same core skeleton, a revenue build, a cost build, working capital, debt and cash flow, applies everywhere. What changes is what you model first, which risk factors belong in your assumptions, and which metrics to watch. This guide covers nine sectors common among Australian SMEs, for owners preparing for a sale or finance application and advisors building models on their behalf. The examples are generic; the numbers illustrate the method, not any particular business.

E-commerce

E-commerce valuations live or die on unit economics. Revenue is easy to grow with paid traffic; the question is whether each cohort pays back its acquisition cost.

Model revenue by cohort, not by total

Average revenue per customer is an average of wildly different behaviours. Model cohorts by acquisition month, tracking retention, repeat purchase rate and average order value (AOV). Retention decays quickly, with the steepest drop in the first 60 to 90 days. Flattening this into one blended churn number overstates long-run revenue.

Compare lifetime value with acquisition cost by channel

Customer acquisition cost (CAC) should be modelled per channel (paid social, search, affiliates, organic); a blended number hides a deteriorating mix. Compare lifetime value (LTV) with CAC on a contribution margin basis, not revenue. A brand acquiring at an LTV to CAC ratio below 2.0 has no headroom if auction prices rise. Subscription businesses follow the same discipline; see our guide to SaaS business valuation.

Put inventory obsolescence in the P&L

Markdowns and write-offs are a recurring cost of the category, not a one-off. Model a provision as a percentage of stock on hand, age it by stock turn, and stress-test with a clearance scenario.

Attribute marketing spend to the channel that earned it

Model return on ad spend (ROAS) by channel and marketing efficiency ratio (MER) at the blended level. A 7-day click window understates brand campaigns that convert weeks later, which misleads the CAC assumption.

A worked example shows why cohort modelling changes the answer. A direct-to-consumer brand acquires 1,000 customers in one month at an average CAC of $45. The table builds revenue and contribution margin from observed repeat behaviour: 34% of the cohort returns in month 2, then decays slowly; AOV is held at $75 ex GST.

CAC on the cohort is $45,000, so LTV to CAC is 1.3 to 1 against a commonly quoted healthy band of 3 to 1. At this ratio, scaling ad spend grows revenue and destroys value at the margin; the fix is retention, and the model should test retention improvements against CAC reductions before any budget increase. The cohort build in Excel is a simple sum of monthly contributions:

=SUMPRODUCT(active_customers_range, orders_per_customer_range, aov_range) * margin_rate

Building & Construction

Model revenue recognition and WIP

Under AASB 15, construction revenue is recognised over time, usually on a percentage-of-completion basis. The model needs a work-in-progress schedule for every contract: original contract value, costs incurred, costs to complete, revenue recognised, and the resulting contract asset or liability. Over-claimed WIP is the classic failure mode: jobs booked at 80% complete that are really at 60% hide the variance until completion. Build a provision for expected losses on loss-making contracts.

Watch contract tenure and pipeline

A residential fit-out business may carry three weeks of booked work; a commercial builder may carry two years. Model backlog coverage, booked revenue divided by monthly capacity, and new work won versus quoted. A shrinking backlog against a full cost base is a leading indicator of trouble. Tenure also drives the forecast horizon: short-tenure businesses need a strong new-business assumption; long-tenure businesses need a strong cost-to-complete assumption.

Stress project concentration

Model the top three projects as a share of revenue, and EBIT if the largest project runs 10% over its cost estimate. One problem contract can wipe out the margin on five good ones. For the valuation side of this sector, see our dedicated guide to construction business valuation.

Allow for cyclical labour costs

Tradie shortages push subcontractor and labour rates up faster than fixed-price contracts adjust. Model labour as a percentage of revenue with a scenario for rate escalation, and check contracts for escalation clauses. In tight supply, margin on fixed-price work compresses even as revenue grows.

Banking & Finance

Track regulatory capital as the binding constraint

For authorised deposit-taking institutions the constraint is APRA capital adequacy; for non-ADI lenders, finance brokers and advice firms it is the capital requirement attached to their ASIC licence. Either way, the model must track the capital buffer, not just the P&L. Growth in the loan book consumes capital, so the constraint on growth is often capital, not demand.

Model interest rate sensitivity explicitly

Net interest margin (NIM) moves with the cash rate and the funding curve. Model the repricing gap: how much of the book reprices within 90 days, and margin compression if the cash rate falls 100 basis points while funding costs lag. A lending business is effectively a geared duration position; present it with a rate scenario block rather than a single margin assumption.

Provision against the loan book

AASB 9 requires expected credit losses: stage 1 at 12-month expected loss, stage 2 where credit risk has increased significantly, stage 3 for impaired exposures. Drive provisions from arrears and from probability of default (PD) and loss given default (LGD) assumptions, then stress the book: PD doubles, LGD rises, security values fall. Model and monitor NIM, cost-to-income, arrears over 90 days, provision coverage and the capital adequacy ratio.

Manufacturing

Model utilisation and absorption

Fixed overheads are absorbed into unit cost at a budgeted utilisation rate. When utilisation falls, every unit carries more overhead and margin disappears faster than revenue. Model break-even utilisation rather than break-even revenue:

=annual_fixed_overheads / (contribution_per_unit * units_per_hour) / annual_capacity_hours

A plant needing 72% utilisation to break even that runs at 68% burns cash at the margin even if total revenue looks stable.

Build supply chain resilience into assumptions

Model single-source suppliers, lead times and imported inputs. Test a scenario where lead times double or freight stays elevated, plus an FX scenario for importers. Inventory is the balancing item: too lean and a disruption stops production; too fat and working capital swallows the cash flow.

Schedule fixed asset replacement

Depreciation is an accounting approximation; replacement is a cash event. Build the capex schedule from the actual asset register: purchase dates, expected lives and replacement cost at today's prices. A plant running assets past their design life is a liability the P&L does not show.

Real Estate

Start with WALE and the lease expiry schedule

Weighted average lease expiry (WALE) is the single most informative number for a commercial asset: a 6.5-year WALE with staggered expiries supports a stable valuation; a 2.8-year WALE with three leases expiring in the same year does not. Model the expiry schedule year by year with renewal probabilities and incentives. Reversion risk, the gap between passing rent and market rent at expiry, belongs in the cash flow, not in a single growth rate.

Test the capitalisation rate both ways

Capitalisation rate is net operating income divided by value. In a low-rate environment buyers accepted cap rate compression; the model should show what happens as rates normalise. Use the reversionary view as the cross-check: value at market rents, not just passing rents.

Run a granular vacancy and cap rate sensitivity

Value is a range, not a point. For a generic suburban office asset with $1.4 million gross rental income and $290,000 of fixed operating costs, the two-way table below shows the swing from vacancy and cap rate assumptions alone:

The same property is worth between $18.9 million and $13.7 million on defensible assumptions, a 28% swing. A single number without this table invites the buyer to choose the assumption that suits their offer.

Professional Services

Quantify key-person risk

A firm where one principal originates 60% of revenue is not worth a multiple of current revenue; it is worth a multiple of what revenue looks like after that person leaves. Model a key-person scenario: a 30% to 50% revenue haircut in year 1, a gradual rebuild, and a replacement hire at market rates. Retention incentives and non-compete provisions are valuation adjustments, not footnotes.

Build revenue from utilisation and leverage

Build revenue from headcount times utilisation times effective rate, not from a revenue growth percentage. Utilisation is billable hours divided by available hours; effective rate is revenue divided by billable hours. Leverage, the mix of partners, seniors and juniors, drives margin: more juniors per partner lift margin only if they stay billable.

Test billable hour stability and client concentration

Model the mix of retainer versus ad hoc work, top client share of revenue and pipeline coverage. A firm living on two large clients has value only if they are contracted; a model that grows the top client's spend indefinitely without a signed basis is assumption drift.

Education

Model the enrolment funnel and yield

Build the funnel explicitly: enquiries to applications, applications to offers, offers to enrolments, and continuing students between years. Conversion at each stage, plus attrition, drives revenue with a lag. An enrolment shortfall this year shows up in revenue next year; flag the leading indicators, not just revenue.

Treat regulatory funding as a scenario, not a given

Much of Australian education revenue depends on government funding: state school funding, VET funding, childcare subsidies and HELP for higher education. Funding per student is a policy variable, not a market variable. Model the conditions attached to the funding and a scenario where funding rates change or eligibility tightens.

Respect the fixed overhead base

Campuses, equipment and qualified staff do not flex with enrolment. Model break-even enrolment and the economics of a small cohort: a class of 12 in a room built for 30 loses money that average cost per student hides.

Cost the digital delivery transition

Moving to hybrid or online delivery is a capital project: learning management systems, content production, platform fees. Model it as a transition cost with a payback period, not as a line item that grows quietly in overheads.

Retail

Separate comparable store sales from portfolio effects

Like-for-like growth is the underlying health metric; new store sales and closures are separate effects. Model the portfolio store by store: mature stores at like-for-like growth, new stores on a ramp-up curve, closures for stores that fail the occupancy cost test. Aggregating same-store and new-store growth into one rate hides the mix.

Model lease commitments at the store level

AASB 16 brought leases onto the balance sheet as right-of-use assets and liabilities. Model the lease schedule: expiry dates, rent review mechanisms and occupancy cost ratio, rent as a percentage of sales. Judge each store against its own lease: a store that clears occupancy cost but not full cost is consuming the business slowly.

Track inventory turnover and markdowns

Model stock turns, weeks of cover, shrinkage and the markdown provision. Slow stock is a cash flow problem before it is a P&L problem; connect inventory days to the funding line, not just cost of goods sold.

Model omni-channel conversion by channel

In-store, online and click-and-collect have different contribution margins. Model conversion and margin per channel and the shift of sales between them. A dollar that moves online does not cost the same to serve, and a blended margin misprices the shift.

Logistics

Drive the model from fleet utilisation

Revenue per truck per week, empty running percentage and backhaul utilisation are the core drivers. A truck covering its fixed costs at 70% utilisation adds most of the revenue from the next 10 points almost straight to the bottom line. Model utilisation as a scenario variable to show the operating leverage.

Model fuel volatility with the contract mechanism

Fuel is a large share of operating cost and moves quickly. Model the fuel levy or surcharge mechanism in customer contracts and the lag between diesel price moves and rate adjustments. Stress-test with diesel up 20% against a 50% pass-through lag: the gap is a real margin call, surfaced in the quarter it happens, not as an annual average.

Distinguish contract duration

Spot work is a rate assumption; contract work is visible revenue with renewal risk. Model the contract book with renewal probabilities and rate indexation, treating spot revenue as the residual that absorbs utilisation swings.

Recognise the capital intensity of the network

Fleet replacement cycles, maintenance capex, depots and facilities make capital per dollar of revenue high. The model needs an explicit capex schedule and a return on invested capital view; revenue growth funded by a worn-out fleet is deferred capex, not profit. Buyers discount for deferred maintenance aggressively.

Conclusion

Across nine very different sectors the same pattern repeats. Value is created by a small number of industry-specific drivers and destroyed by concentration, underinvestment and assumption drift. The discipline that holds everywhere is the same: build revenue from the actual unit of activity, whether a cohort, a contract, a lease, an enrolment or a truck hour; model the balance sheet consequences, WIP, provisions, leases and replacement capex; and stress the assumptions that move the number most. A sector-aware model is the same core, built around the right drivers, with the risk factors that actually apply. That is what separates a number you can defend from one you can only quote.

Related reading

• SME Valuation Multiples by Industry in Australia
• Business Valuation: The Complete Guide for Australian SMEs
• Financial Modelling in Excel: The Complete Guide