Methodology: how DevPartner calculates, and what it will not claim

A feasibility is only as trustworthy as the discipline behind it. This page sets out how DevPartner computes, what is user-stated versus engine-derived, how runs and report versions work, and the limitations we would rather state plainly than hide in a footnote.

1. The engine is deterministic

Given the same inputs and the same engine version, DevPartner returns the same outputs — every time, for every user. There is no sampling, no model inference and no hidden adjustment inside the calculation chain. That property is what makes a result defensible: two people can reopen the same run months apart and see identical numbers.

inputs (all user-stated)
  -> planning capacity   : buildable and saleable area, yield, parking
  -> cost build-up       : construction, fees, statutory, external, contingency
  -> revenue             : mix x rate, phased by absorption
  -> programme           : approvals, construction, sales
  -> cash flow           : peak funding and finance cost
  -> structure           : outright / area share / revenue share on one base
  -> decision            : verdict against your stated criteria
outputs (each traceable to the inputs above + an engine version)

2. What is stated, what is derived

DevPartner supports structured intake and commercial modelling for sites across all 28 states and 8 union territories. Its governed regulatory coverage is narrower and explicit: a planning value can be rule-derived only when the competent jurisdiction is confirmed and the exact rule-pack version is centrally verified for deterministic use. Otherwise the analysis reports a gap or labels a professionally confirmed value as user supplied. It never substitutes a national average, a neighbouring state's rule or a model bye-law.

3. Runs, staleness and frozen versions

• A run is immutable. It records the inputs it consumed, the engine version that computed them and the outputs produced.
• Changing an assumption marks the run stale. Nothing is recomputed behind a report you have already circulated.
• A report version is frozen. Once created it reopens unchanged, which is what makes it usable in an investment committee or a negotiation months later.
• Every metric can be traced. The trace shows the inputs and the arithmetic that produced the figure on screen.

4. Ranges, not single points

A single-point answer implies a precision no development appraisal has. DevPartner presents scenarios and a sensitivity grid across the two drivers that move the result most — typically the achievable sales rate and total development cost — so the decision is made against a range with named drivers.

5. The free calculators are deliberately simpler

The public tools — FSI / FAR, development potential, construction cost, residual land value and joint development — are single-page models, separate from the workspace engine. They do not phase a programme, model drawdown or compare structures across a cash flow. Their starting values are visible placeholders you are expected to replace, never a statement of what applies to your site.

6. Known limitations

• No universal statutory coverage: unsupported, stale, conflicting or unverified jurisdictions fail closed and require professional review.
• No market data feed; rates and absorption are yours to evidence.
• No title, encumbrance, litigation or environmental due diligence.
• No tax, GST or stamp-duty computation beyond the figures you enter.
• Outputs are decision support, not a valuation or an approval.