What 7,571 NZ accounting and bookkeeping reviews actually show.
A study of every publicly posted customer review across 484 New Zealand accounting and bookkeeping firms. The typical experience is excellent — and one region, and one in five firms, don’t fit that pattern.
Are NZ accounting and bookkeeping firms reliable?
Across 7,571 publicly posted reviews of 484 NZ accounting and bookkeeping firms, the typical experience is excellent — 92.6% of reviews are 5-star, mean rating 4.79 out of 5. But the picture isn’t uniform: Wellington firms rate measurably lower and more inconsistently than every other region, and 22.7% of firms sampled have no public review at all, meaning there’s often no review signal to check in the first place.
Overwhelmingly positive, genuinely consistent.
Mean star rating 4.79 out of 5 across 7,571 reviews. 92.6% give 5 stars; only 4.3% rate 2 stars or fewer. Standard deviation is 0.82 on a 5-point scale — tight. This holds at every region and every firm size: no group in this corpus drops below a 4.6 mean.
Review count and rating don’t move together. Across the 374 firms with at least one review, review count and average rating correlate at r = 0.02 — essentially zero. A firm with 80 reviews is no more or less likely to be well-rated than a firm with 3. Busier doesn’t mean better or worse here.
It’s regional, and it’s Wellington.
Averaged nationally the picture is flat, but one region stands apart on every measure at once:
| Region | Mean stars | 5-star share | 2-star-and-below share | Rating spread (lowest–highest firm) |
|---|---|---|---|---|
| Bay of Plenty | 4.93 | 97.3% | 1.3% | 1.6 pts (3.4–5.0) |
| Canterbury | 4.83 | 92.0% | 2.8% | 2.0 pts (3.0–5.0) |
| Auckland | 4.80 | 93.6% | 4.4% | 4.0 pts (1.0–5.0) |
| Waikato | 4.75 | 91.9% | 5.8% | 1.7 pts (3.3–5.0) |
| Wellington | 4.65 | 87.1% | 7.4% | 4.0 pts (1.0–5.0) |
Wellington has the lowest mean rating, the lowest 5-star share, the highest share of poor ratings, and — tied with Auckland, on roughly a third of the firm count — the widest full-range spread of any region. It isn’t one outlier dragging the number down across an otherwise-tight market: Wellington’s complaint mix also runs hotter than the national average (10.2% of its negative reviews mention being treated rudely or unprofessionally, against 7.9% nationally, on a sample too thin to call a leading theme but consistent with the rest of Wellington’s numbers rather than contradicting them).
This isn’t simply “more firms, wider range.” Canterbury has more firms in this corpus than Wellington (85 vs 51) and a far tighter spread (2.0 points vs 4.0) — if spread were just a function of sample size, Canterbury’s should be wider, not narrower. Something specific to the Wellington market is producing both a lower average and a wider range, not sampling noise. Flagged as a real regional pattern, not proof of a cause.
Mid-sized firms are the tightest group, not small or large.
| Firm size (by review count) | Mean stars | Rating spread (lowest–highest firm) |
|---|---|---|
| Small | 4.79 | 4.0 pts (1.0–5.0) |
| Mid | 4.63 | 2.7 pts (2.3–5.0) |
| Large | 4.81 | 3.7 pts (1.3–5.0) |
No gradient: mid-sized firms have the tightest spread, not small or large. The regional pattern above is the real structure in this corpus, not a size effect in disguise. (“Size” here is public review count split into three equal groups — a proxy for public footprint, not headcount; see the methodology page.)
A fifth of firms have no public review at all.
22.7% of the 484 firms sampled (110 of 484) have zero public reviews. For roughly one in five NZ accounting and bookkeeping firms, there is no public review signal to check before choosing one — not a bad rating, no rating. That’s a different kind of risk from “could pick a badly-reviewed one”: here, a meaningful share of the category can’t be screened by reviews either way. A firm with no reviews isn’t a bad firm; it’s an unmeasured one.
More reviews every year, same rating.
Review volume is climbing — 1,004 reviews logged in 2024, 1,483 in 2025, 1,687 in 2026 (a partial year at the time of writing) — more people are reviewing accounting and bookkeeping firms publicly each year. Mean rating has barely moved: 4.78 (2025) to 4.82 (2026), +0.04. Flat, not a trend worth a claim either way.
Politeness and referral language lead, nearly tied.
Politeness (55.8%) and referral language (55.4%) lead nationally — reviewers explicitly saying they’d recommend the firm or use it again. Accessibility (42.5%) and proactive advice (42.2%) follow closely — reviewers specifically noting the firm reaches out, explains things unprompted, or is easy to get hold of. Those are relationship traits, consistent with accounting being an ongoing advisory relationship rather than a single job.
The three least-praised traits, nationally: deadline reliability (3.1%), value for fees (7.0%), accuracy (8.9%). All three hold at roughly this level across every firm size — deadline reliability in particular is flat at 2.9–3.1% whether the firm is small, mid or large, which rules out a size-driven read.
Accuracy needs a specific caveat the other two don’t. The complaint-side data shows errors or mistakes mentioned in just 0.3% of negative reviews (1 of 328) — vanishingly rare. Read against that, accuracy’s low praise rate most likely means getting the numbers right is table stakes and goes unremarked when it holds, not that it often fails. Deadline reliability and value for fees don’t have that same complaint-side reassurance — missed-deadline mentions didn’t clear the reporting threshold in any group — so those two stay the more defensible candidates for “quietly under-delivered, not specifically complained about” below.
Two traits clients almost never confirm.
- 01Deadline reliability is almost never specifically praised, at any size of firm: 3.1% nationally, flat at 2.9–3.1% across small, mid and large firms. Tax and compliance work runs on fixed external deadlines (IRD dates, not negotiable client preference), which makes this a plausible candidate for a quietly-expected-but-rarely-confirmed trait. A client whose filing is done on time with no proactive “it’s done, here’s confirmation” message has no reason to mention deadlines in a review either way.
- 02Value for fees is the second-least-praised trait, and declines somewhat with firm size (14.7% small, 9.0% mid, 6.8% large, though the small-firm figure rests on only 34 reviews and should be read as directional). Clients aren’t volunteering that fees felt fair or well-explained relative to the work. A gap in communicating value, not necessarily in pricing itself, since there’s no corresponding spike in overcharging complaints (1.8% nationally, 6 reviews).
Worth saying plainly: this is a thinner, more cautious finding than a study with a strong complaint-side signal could make. The complaint data here is too sparse to corroborate either gap directly. Both points above are read from the praise-side pattern alone, flagged accordingly, not asserted as proven process failures.
A thin sample, said plainly.
Negative reviews are rare (4.3% of the corpus, 328 reviews) and the complaint categories within that slice are mostly too thin to name a leading theme: overcharging or fee disputes (1.8%, 6 reviews), errors or mistakes (0.3%, 1 review), unclear advice (0.3%, 1 review) are all noise-level. The one category with enough signal to mention is rude or unprofessional conduct (7.9%, 26 reviews) — still a small base, but the only complaint theme that clears single digits in review count, and it’s the theme that runs hottest in Wellington specifically (10.2%) and in mid-sized firms (16.4%, 10 of 61 negative reviews — also thin, flagged not claimed).
Full corpus, collection method, scoring, coding, thresholds and platform name are stated on the methodology page. This is public reviews, not a random sample of every client relationship; the corpus skews strongly positive and the negative-review sample is thin, so complaint-theme percentages are kept qualitative, not presented as precise measurement; the Wellington finding is a description of this corpus, not a tested explanation of why; the value-for-fees size trend rests on a small-firm base of 34 reviews and should be read as directional.
