Loading
Economics · Demand Forecasting
Build plans around the drivers of demand, not a single optimistic number.
For leadership teams wary of repeated forecast misses and false precision, but still needing a basis for capacity, inventory, hiring and investment. We combine historical patterns, pipeline, conversion, cohorts, pricing, capacity and external drivers into transparent forecasts, with ranges, documented assumptions and reforecast triggers.
- Drivers and ranges, not point guesses
- Back-tested before trusted
- Reforecast triggers built in
Trusted by established businesses
The constraint
Situations where the forecast is the problem.
Forecasts fail for structural reasons, not bad luck. Select a situation to see what's usually behind it.
An untrusted forecast doesn't get ignored. It gets replaced by hallway numbers. Each function plans on private assumptions, and capacity, stock, hiring and cash quietly drift out of alignment until the gap surfaces as a crisis nobody forecast.
The approach
What a decision-ready forecast contains
A forecast is decision-ready when seven things are explicit. The unit and horizon: what's being forecast, over what period, for which decision. The drivers: the levers and conditions that actually generate the demand. The assumptions: written down, each with a note on how much the answer moves if it's wrong. Ranges: base, upside and downside trajectories instead of one number pretending to certainty; a single point is false precision, and false precision is how forecasts lose trust. Validation: back-testing, which simply means replaying the method over past periods to see how it would have performed before anyone bets on it. Ownership: a named person who maintains it. And reforecast triggers: the pre-agreed conditions that force an update, so the model is revised on evidence rather than on whoever shouts loudest.
What we analyse
Forecast methods, matched to the decision, not the fashion
Decisions
Decisions this supports
The forecast exists for the commitment it informs. That's how its unit, horizon and method get chosen.
Hiring & staffing
Rosters and recruitment timed against demand ranges, not last quarter's panic or optimism.
Capacity commitments
Volume outlooks that scale and capacity decisions can be honestly tested against.
Inventory & purchasing
Stock positions balanced against the cost of stock-outs and the cost of capital, under stated ranges.
Cash planning
Revenue trajectories with downside cases, so cash decisions see the scenario that hurts.
Investment & market entry
Demand evidence for the business case, with assumption criticality visible to the people signing it.
Budget allocation
Budgets set against driver-level evidence of where demand actually comes from.
Deliverables
What you walk away with
- ✓A documented forecast model: drivers, assumptions and logic visible, owned by a named person
- ✓Demand and revenue scenarios: base, upside and downside, each with its assumptions
- ✓Back-testing results showing how the method would have performed on your own history
- ✓A bias analysis of past forecasts: whether they leaned optimistic or pessimistic, and by how much
- ✓A leading-indicator set: the early signals that say which trajectory is unfolding
- ✓Decision implications: what the ranges mean for the capacity, inventory, hiring or investment call at hand
- ✓Forecast governance: ownership, update cadence and the reforecast triggers that force a revision
How we work
From decision to governed forecast
The intake conversation asks four things: the decision the forecast serves, its unit and horizon, what history exists, and the drivers you already know matter.
- 01
Define the decision
Agree what the forecast is for, the unit and horizon that decision needs, and what accuracy is actually worth to it.
GateA named decision and horizon, not ‘a forecast’ in the abstract
- 02
Prepare the evidence
Demand history, events, pricing, marketing activity, pipeline, retention, capacity records, launch plans and external conditions, assembled and quality-checked.
- 03
Map the drivers
Build the driver tree connecting the levers you control and the conditions you don't to the demand being forecast.
GateA driver map the leadership team recognises as their business
- 04
Build and back-test
Fit candidate methods, replay them over past periods, and compare their errors honestly, including where each method fails.
GateBack-test results shared openly: method chosen on evidence, not preference
- 05
Compare the scenarios
Base, upside and downside trajectories with assumption criticality: which assumptions move the answer most, and how the pending decision fares across the range.
- 06
Hand over with governance
The model, its documentation, the leading indicators and reforecast triggers, plus the cadence and ownership that keep it alive.
Scope
What this is, and what it isn't
What's included
- A transparent, documented model your team can explain and maintain
- Ranges with assumption criticality: which inputs move the answer most
- Back-testing and bias analysis before the model is trusted with a decision
- Reforecast triggers and governance, so updates happen on evidence
What this is not
- A guaranteed number: no honest forecast eliminates uncertainty, and we won't pretend otherwise
- Single-point predictions presented without ranges or assumptions
- Black-box models nobody in your business can explain or challenge
- Target setting: a forecast estimates what's likely; what you choose to aim for remains a leadership call
- Leadership teams burnt by forecast misses but still needing a planning basis
- Businesses facing capacity, inventory, hiring or investment commitments
- Teams whose marketing, sales and operations each plan on a different number
- Operators launching offers or locations who need structured assumptions, not blind extrapolation
Probably not the right fit
- −You want a precise number that ends all argument: ranges and assumptions are the honest product here
- −You need a big number for a pitch deck, not a planning tool
- −Nobody will own or update the model: a forecast is a process, and an orphaned one decays fast
Where this fits
The demand outlook other decisions are built on
Demand Forecasting owns the volume outlook. The services around it consume its ranges, widen its stress cases, or build the data foundations under it.
Demand Forecasting & Commercial ModellingThe hub: owns the demand outlook
Owns: forecast units and horizons · drivers · assumptions · ranges · back-testing · reforecast triggers.
Demand Forecasting & Commercial ModellingScale, Capacity & MES
When it comes into play
The forecast feeds a capacity commitment.
What it owns
Capacity thresholds, utilisation, investment triggers.
Relationship to Demand Forecasting & Commercial Modelling
Consumes the demand ranges to time fixed-cost steps.
Proof
Selected work.
Payment Processing Cost Reduction. An ecommerce retailer was losing a significant percentage of revenue to payment processing and invoice platform fees. Web Lifter redesigned the entire sales and payment workflow, replacing Stripe and Paycove with a direct Westpac PayWay integration and a custom-built invoicing platform. The new architecture reduced transaction costs, streamlined operations, and delivered immediate profit improvements without requiring any increase in sales volume.
Read the caseWhat clients say
“We can't recommend Web Lifter highly enough … a digital partner who could understand our operations, connect the dots between marketing and backend systems, and deliver real results.”
Questions
Common questions
Scope
No single number will be, and any forecast sold that way should worry you. The honest product is a transparent model with ranges, documented assumptions and triggers that tell you early when reality is leaving the range. That's what lets management act quickly instead of defending a number.
No. A target is what leadership chooses to aim for; a forecast is what the drivers say is likely. This work produces the forecast, and where a target exists, it frames the gap between the two so the stretch is explicit rather than hidden inside the number.
No. Method follows the decision: driver-based and pipeline models for management planning, cohort models for retention-shaped revenue, time-series where history is strong, econometrics only where the data supports it. We deliberately avoid promoting complexity: the simplest method that answers the decision wins.
Method & uncertainty
Replaying a forecasting method over your own past periods (pretending we're standing in an earlier month and forecasting forward) then comparing what it would have said against what actually happened. It's how a method earns trust before any real decision rides on it.
Three ways, all visible: ranges or intervals instead of single points; assumption criticality: which inputs move the answer most if they're wrong; and reforecast triggers: the pre-agreed conditions that force an update. Uncertainty presented clearly is usable; uncertainty hidden inside one confident number is a trap.
Only when the decision needs them and the data can carry them, typically when demand is genuinely driven by measurable external factors and there's enough clean history to estimate the relationships. The feasibility check comes first; where the evidence is thin, we say so and use simpler structures.
Inputs
Less than you might fear, but the answer changes shape. Rich history supports time-series and cohort methods; thin or no history (new offers, new locations) moves the work to driver-based structures, comparable evidence and pipeline data, with wider ranges and faster reforecast cycles while the early signal comes in.
Four things: the decision the forecast serves, the unit and horizon it needs, what history is available, and the drivers you already know matter. No account, no waitlist: the answers are enough to scope the engagement honestly.
Pathways
Capacity commitments route to Scale, Capacity & MES; market-entry cases to Growth & Expansion; external-shock exposure to Macroeconomic Resilience; downside limits to Sensitivity, Scenario & Stress Testing; and automating the model's data feeds to Data Engineering.
Plan on drivers