How we researched this.
ClearLogic’s construction research draws on publicly posted customer reviews from NoCowboys, the same major New Zealand tradesperson-review platform used in our trades and property maintenance research — a separate corpus, crawled the same way. This page states exactly how we collected, scored, and filtered that data, so any figure we publish can be checked rather than taken on faith. (Our trades research methodology and property maintenance research methodology are separate pages, each covering a separate corpus.)
Publicly posted reviews, crawled politely.
We crawled NoCowboys respecting the site’s published robots.txt,
including its explicit allowance for AI agents and its 5-second crawl delay. We did not access anything
not publicly visible on the platform, and we did not create an account or interact with the site beyond
reading published pages.
The corpus covers two construction categories, national scope, no geographic pre-filtering:
| Category | Businesses | Reviews |
|---|---|---|
| Builders | 166 | 5,772 |
| Outdoor & civil construction | 162 | 7,795 |
| Total | 328 | 13,567 |
Category construction, stated plainly: builders is a base corpus topped up with businesses found genuinely new in a construction-project-managers scrape — the rest of that category’s listings were exact-duplicate NoCowboys listings already in the builders data, excluded as double-counting. Outdoor & civil construction merges three originally-separate platform categories — concrete services, earthworks & retaining walls, and fencing contractors — because they overlapped each other by 44–59% (the signature of the same generalist businesses listed under multiple categories), materially higher than any of them overlapped the builders base (20–41%). Reporting them separately would have meant counting the same businesses multiple times, so they were deduplicated by business identity and combined into one category instead.
This is a cross-sectional snapshot of each business’s full public review history as it stood at crawl time (12 August 2026). It is not a rolling monitor and does not update automatically — if we republish an updated version later, we’ll date it separately from this one.
Same crawl method as our other research.
Consistent with our trades and property maintenance research: a cross-sectional snapshot of each business’s full public review history as it stood at crawl time, not a rolling monitor.
How reviews are scored, and how we coded them.
The scoring system
Each review on the platform carries four sub-scores — communication, quality, reliability, and value — plus an overall score, all on a 0–100 scale, self-reported by the reviewer at the time they left it. We didn’t create or adjust this scoring system; we report it as published. Same system, same method across all our research.
Complaint & praise coding
We defined a negative review as one scoring below 50/100 on any dimension, and a positive review as one scoring 80/100 or higher overall. Within those groups, we coded the reason for the score against a fixed taxonomy — 6 complaint categories, 13 praise categories — using an LLM classifier (Claude Haiku 4.5, run in batches) rather than keyword matching. The taxonomy is fixed in advance of coding, not built from the results.
A statistic only gets published if it clears two thresholds.
- 01At least 30 reviews behind any single figure (a category × region, category × business-size, or similar breakdown), so a figure isn’t resting on a handful of opinions.
- 02At least 5 different businesses contributing to that figure — our own addition, on top of the review-count threshold, because a stat built from 1–4 businesses’ reviews can be traced back to roughly which businesses it describes even with no name attached. Breakdowns that don’t clear 5 businesses are dropped from what we publish, not estimated or smoothed over to fill the gap.
Three specific breakdowns were dropped for this reason:
| Category & breakdown | Businesses | Reviews |
|---|---|---|
| Builders, Canterbury region | 1 | 30 |
| Builders, Northland region | 3 | 185 |
| Outdoor & civil construction, Bay of Plenty region | 4 | 146 |
All three clear the 30-review floor on their own — Northland’s 185 reviews would look robust in isolation — but fail the business-count threshold, exactly the failure mode that threshold exists to catch: review volume can look solid while the business count behind it is small enough to identify. These three cells remain in our internal data but are excluded from anything published here or in the study.
No names, ever.
No business name, owner name, staff name, or individual review appears in anything we publish. Every figure we report is an aggregate across many reviews and many businesses. No verbatim review text is published or linkable from our construction research.
Three studies, three genuinely different patterns.
Each of our three studies so far has found a different relationship between business size and review outcomes, and we think that difference is worth stating plainly rather than forcing a single comparison across all three. Our trades research (electricians, plumbers, painters) found a small negative correlation between business size and approval rating for two of the three trades (r = −0.32 plumbers, r = −0.23 electricians). Our property maintenance research found no equivalent correlation in any of its three categories (r = 0.04–0.15). This construction corpus shows a third pattern, different in kind from a simple correlation: the spread between best- and worst-reviewed businesses is far wider among outdoor & civil construction contractors (92 points) than among builders (37 points), and that spread sits almost entirely in the small-business tier (92-point spread small vs. 14.6 mid, 4.9 large) — but praise data for pricing and quote accuracy barely moves with size, so it reads as a tail of a minority of poorly-reviewed small operators, not a size effect on the average. We flag this here rather than in any one study, because the three corpora describe different service models — one-off callouts, recurring client relationships, planned multi-stage projects — and reading any one as a rebuttal of another would overstate what any single corpus can show.
Read the numbers with these in mind.
- —This is public reviews, not a random sample of every job done. People with an unremarkable, fine experience are less likely to leave a review at all than people who were thrilled or angry, so the corpus skews toward strong opinions in both directions.
- —The outdoor & civil construction category merges three originally-separate review categories because they overlapped each other too heavily to report separately without double-counting the same businesses. Treat it as one broader category, not three distinct sub-trades.
- —Small figures on small bases are directional, not precise. Builders’ negative sample (~23 reviews) was too thin to categorise into complaint themes at all; outdoor & civil construction’s (48 reviews) clears our floor by a hair, stated plainly rather than estimated around.
- —Correlation and spread are descriptions of this corpus, not tested mechanisms for why some businesses are poorly reviewed.
Figures drawn from this corpus are treated internally as candidate findings until reviewed and approved for publication. This page will be updated if the underlying data is refreshed or re-crawled. Last updated: 12 August 2026.
