We Read 38,379 BBB Ratings. 92% Hold an A+ — Except in the Categories You'd Expect.

The Better Business Bureau's letter grade is one of the oldest trust signals on the consumer web — a single letter next to a business name, implying decades of accumulated judgment. We pulled BBB's own public data for 38,379 rated business profiles and asked the obvious question: does the grade actually discriminate between good and bad businesses, or is it mostly the same letter for everyone?

The headline number says the second. The real answer is more interesting than either.

92% hold an A+ — but that number hides a real, 27.8-point gradient

Across every rated BBB profile we pulled, 92.0% hold an A+ (35,335 of 38,379). F ratings are rare — just 1.9% (743). Read in isolation, that looks like a grading system with almost no discriminating power: nearly everyone gets the top grade, so the grade barely tells you anything.

That's true in aggregate, but it stops being true the moment you split by category. We pulled the A+ and F rate specifically within 11 categories, chosen to span the range from classically complaint-prone trades to licensed professions:

Collections Agencies68.7%n=150

Towing Company78.2%n=142

Used Car Dealers84.5%n=414

Moving Companies86.5%n=385

Financial Services91.4%n=151

Auto Repairs91.7%n=689

Roofing Contractors93.1%n=1,571

Veterinarian93.2%n=191

Dentist94.4%n=390

Mortgage Broker95.6%n=158

Lawyers96.5%n=487

The pattern runs exactly the direction you'd predict from real-world complaint reputations. Collections Agencies — an industry built on chasing people for money they may not even owe — sits lowest at 68.7% A+, with the highest F-adjacent risk of any category we checked. Towing (78.2%), Used Car Dealers (84.5%), and Moving Companies (86.5%) — the classic "read the reviews before you sign anything" trades — all sit meaningfully below the 92% overall average. Licensed professions sit at the top: Lawyers hold a 96.5% A+ rate and, across 487 profiles, not a single F.

So the grading system isn't meaningless. It's just discriminating from a very compressed, very high floor — even the worst category we tested is still more than two-thirds A+, and its F-rate (4.0%) is a long way from what "the most complained-about BBB category" might imply on its own.

Paying for accreditation moves the F-rate far more than the A+ rate

BBB accreditation is a paid program — businesses apply and pay dues to display the BBB seal. It's reasonable to wonder whether that's just a rating-inflation mechanism in disguise. The data says: not exactly.

F rate — non-accredited3.86%

744 of 19,291 non-accredited profiles

F rate — accredited0.005%

1 of 19,099 accredited profiles

Accredited businesses are 96.3% A+ (18,392 of 19,099) versus 87.9% A+ for non-accredited (16,949 of 19,291) — a real but modest +8.4-point gap. The F-rate gap is the striking part: only 1 out of 19,099 accredited profiles holds an F (0.005%), versus 744 of 19,291 non-accredited (3.86%) — roughly 770x more common outside accreditation. Paying for the seal doesn't buy much extra shine at the top; it correlates with an almost total absence of the bottom grade, plausibly because BBB's own accreditation standards revoke accreditation from businesses that slide to an F rather than accreditation itself lifting anyone's grade.

A+ doesn't mean complaint-free

The letter grade is a single number. The complaint history underneath it is not, and the two don't always agree the way you'd assume. We pulled the full complaint record for two A+-rated, accredited businesses in the lowest-scoring category, Collections Agencies:

Both hold the exact same letter grade. One has generated roughly 11x the complaint volume of the other over the same three-year window. If you're using an A+ as a shortcut for "this business won't generate friction," the letter alone can't tell you which of these two you're looking at — you have to read the complaint history yourself.

The companion picture: what people actually report to BBB as scams

BBB also runs Scam Tracker, a separate self-reported database of consumer scam incidents — not tied to individual business ratings, but a useful adjacent picture of what's actually costing people money right now. Pulled across all US states and Canadian provinces over a trailing 90-day window:

• "Online Purchase" is the #1 reported scam type in nearly every state and province — usually 35–65% of all reports, ahead of phishing, counterfeit products, and employment scams.
• Alaska, Nevada, and Wyoming show reported losses-per-capita far above what their report volume alone would suggest — a handful of large-dollar reports in a small population skews the per-capita rate; worth reading as "a few costly incidents," not "this state is uniquely scam-heavy."
• Delaware's pct_reporting_loss (35.4%) is a low outlier against most states' 50–57% — meaning a smaller share of Delaware's scam reports involve an actual confirmed dollar loss, for reasons the data doesn't explain on its own.

We also checked whether a state's BBB rating profile tracks its Scam Tracker risk — it doesn't cleanly. Delaware's BBB business population is actually more A+-heavy (93.5%, 217 of 232 rated profiles) than California's (84.9%, 6,711 of 7,901), despite Delaware being the pct_reporting_loss outlier above. That's not really a contradiction: BBB rates businesses that registered with BBB, while Scam Tracker logs consumer-reported incidents that are disproportionately unregistered, anonymous bad actors — different populations measuring different things. We're reporting the negative result rather than forcing a link the data doesn't support.

What this shows, and what it doesn't

What holds up: BBB's rating system is not simply broken or meaningless — it produces a real, directionally sensible gradient once you look at the category level, and accreditation has a genuine (if narrow) statistical relationship to the bottom of the grade distribution. The overall "92% A+" number is real, but citing it alone flattens a more interesting, more useful pattern underneath.

What limits it:

• This is a self-selected, BBB-registered population, not a random sample of American businesses. Businesses that register with BBB (and especially those that pay for accreditation) are already selecting themselves into a system with reputational incentives — this isn't evidence about the broader universe of unrated businesses.
• The category facet is a top-N breakdown, not a guaranteed full partition of the ~53,000-business catalog — the 11 categories checked here are real, individually-queried subsets, not derived from an unverified aggregate.
• "Rated" (38,379) is itself a subset of all crawled BBB profiles — a meaningful number of businesses in BBB's index have no computed letter grade at all; every percentage above describes the rated subset, not BBB's full catalog.
• Complaint counts are BBB's own self-reported "closed in the last N months" figures, pulled directly from each business's public complaints page, not independently audited by us or anyone else.
• The Scam Tracker geographic comparison is a negative result on a small sample (Delaware's n=232) — treat it as "no clean link found," not as proof no link could ever exist with a larger sample.

Sources

Methodology

Every number above was pulled live on 2026-08-25 and 2026-08-26 from BBB's own public business-profile and Scam Tracker data. The overall 92.0% A+ figure and the accreditation F-rate comparison cover all 38,379 rated profiles in the dataset. Category-level A+/F rates were computed by filtering to each category individually and reading its own rating breakdown — not derived from the top-level category list. The two complaint case studies were selected from A+-rated, accredited Collections Agencies profiles and their full complaint histories pulled directly from BBB's own complaints pages. Scam Tracker figures cover a trailing 90-day aggregation window across US states and Canadian provinces. The BBB-rating-vs-Scam-Tracker state comparison used Delaware and California as a contrasting pair; we did not compute this for all 50+ states/provinces, so treat it as a spot-check, not an exhaustive analysis.

Want to pull this data yourself? Pricing has the free tier to get started.