Methodology. How we score, and what we can't.

Each trade benchmark covers five independent measures. Together they answer the question: what does good look like for this trade?

1. Google Review Bar — the median review count in the Google Map Pack for this trade, by metro. The number your competitors are at.
2. Website Scorecard — a 45-point score across 9 dimensions. How good the average site is for this trade, and where it leaks.
3. GBP Category & Setup — the primary Google Business Profile category for this trade and whether operators are setting it correctly.
4. Ops Baseline — average job value, close rate, speed to lead, review behaviour, seasonality. The business-side numbers.
5. Buyer Voice — the actual questions homeowners ask about this trade, clustered from real research, with quotes and source URLs.

For each trade, we audit the Google Map Pack in 17 Australian metros (Sydney, Melbourne, Brisbane, Perth, Adelaide, Gold Coast, Canberra, Hobart, Newcastle, Wollongong, Sunshine Coast, Geelong, Townsville, Cairns, Darwin, Toowoomba, Launceston). The review bar is the median review count across the top-ranking businesses.

We also report P25 and P75 (the 25th and 75th percentile) so you can see the spread, plus the max review count observed.

Per-metro data: the benchmark pages show the review bar broken down by metro, so you can compare your city specifically. Where a metro has fewer than 3 results, we note the small sample.

What it tells you: not how many reviews you should have, but how many the businesses currently winning the Map Pack do have. It's the bar, not the target — you need to be in the range to compete.

We score websites across 9 dimensions, 5 points each, 45 total. The scoring is manual — a human reviewer assesses each site against defined criteria. The market median across all 40 trades (253 AU-Primary sites scored) is 31/45.

The 9 dimensions:

1

3-Second Clarity

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

2

Visual Polish

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

3

Trust Above Fold

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

4

CTA Clarity

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

5

Local Relevance

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

6

Service Architecture

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

7

Proof Depth

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

8

SEO/AI Usefulness

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

9

Mobile/Speed Confidence

Scored 1–5. Does the site deliver on this dimension within 3 seconds of landing?

Each trade page shows the per-dimension breakdown and highlights the two weakest dimensions — the highest-leverage fix areas.

AU-Primary only: the per-trade averages use Australian-primary sites only (253 of the 289 scored). National/US reference sites are excluded from the per-trade rollup to keep the benchmark Australian.

Gold exemplar: each trade names the highest-scoring site we found. It's not an endorsement — it's a reference point for what a strong site looks like in that trade.

The Google Business Profile category determines which searches your listing appears for. We report the modal (most common) primary category for each trade and flag when a significant number of operators have it set incorrectly.

A wrong category doesn't just lose you ranking — it means you're competing in the wrong race entirely. The benchmark pages call this out per-trade.

The operations baseline covers the business-side numbers:

• Average Job Value (AJV) — sourced from industry data where available, modelled where not. The chain is shown on each trade page.
• Close Rate — typical vs benchmark (with follow-up). Most trades sit at 20–30% typical; proper follow-up lifts this to 30–40%.
• Speed to Lead — how fast operators respond to enquiries. Sourced from industry surveys and Google data.
• Review Behaviour — whether operators in this trade actively ask for reviews, and what the response rate looks like.
• Seasonality — peak and off-peak patterns for this trade.

The leak calculation: on trade pages that have both AJV and close-rate data, we calculate a "monthly leak" figure — the revenue difference between typical and benchmark close rates at 10 enquiries/month. This is modelled, not measured, and the chain is always shown beneath the number.

The buyer-voice section draws from more than 4,300 distinct homeowner accounts across 8 Australian states and 40 trades — real quotes, with source URLs, clustered into the questions homeowners actually ask about this trade.

These aren't survey responses to our questions. They're quotes from homeowners talking among themselves — forums, review sites, community groups — where nobody was selling to them.

Each trade page shows the top question clusters with representative quotes. The point: these are the questions your website should answer before the buyer picks up the phone.

When we can't find a reliable figure for a measure in a specific trade, we label it NOT FOUND — not blank, not estimated, not replaced with a cross-trade average.

NOT FOUND means: we looked, we couldn't find a primary source that meets our standard, and we'd rather show the gap than fill it with a guess. It's a feature, not a bug.

As better data becomes available — from industry bodies, government surveys, or our own field research — NOT FOUND entries get replaced with sourced figures and the page is re-run.

Every number on a benchmark page carries one of three labels:

Sourced

We have a URL, a publication, or a dataset. The source label and access date are shown.

Modelled

Derived from other figures with the calculation chain shown (e.g. "AJV × close-rate gap × 10 enquiries/month"). You can check the arithmetic.

Not Found

No reliable source exists. The gap is visible, not hidden.

This is enforced by the build pipeline — the benchmark data schema requires a provenance label on every figure. A bare number cannot ship.

Each trade benchmark page shows a "Field data · [date]" stamp in the header — that's when the data was last audited.

The library is updated on a rolling basis, not on a fixed schedule. When we re-audit a trade's Map Pack, re-score its sites, or get new industry data, the page is regenerated from the database and redeployed. The date stamp updates with each run.

We don't backdate. If a figure is 6 months old, the date says so.

The benchmarks are a yardstick, not a diagnosis. Limitations you should know about:

• Small sample per trade. Some trades have as few as 5–8 scored sites. The per-trade average is directional, not statistically robust. The cross-trade market median (253 sites) is more stable.
• Metro selection. We cover 17 Australian metros. If you're in a regional area outside these metros, your local market may differ.
• Snapshot, not real-time. Map Pack rankings and review counts change daily. Our data is a snapshot at the audit date.
• Website scoring is manual. A human reviewer assessed each site against the 9 dimensions. This means some subjectivity, offset by a consistent rubric across all 40 trades.
• Ops data varies. AJV, close rates, and speed-to-lead figures are sourced from industry reports and surveys where available. Coverage varies by trade — some have robust data, others are modelled from adjacent trades.
• No causal claims. The benchmarks show what good operators have; they don't prove that having those things caused the success. Correlation, not causation.