What Do Carvana Reviews Say About Buying vs Selling a Car?
In 5,380 public Carvana reviews, people selling a car were far happier than people buying one on Trustpilot. For buyers, the worst ratings came after delivery, when a problem was never resolved.
Carvana runs two businesses under one name. It sells used cars online, and it buys cars from people who want to sell theirs. One star rating covers both, but a seller and a buyer have very different experiences with the company. A seller gets an offer, a pickup and a payment. A buyer gets a car that then has to arrive on time, work, get registered and be covered by a warranty.
To see what each side praises and complains about, we read 5,380 public reviews of Carvana from five sites, written between January 2024 and June 2026, and tagged each one with what the writer was doing and what went right or wrong.
The data: 5,380 reviews from five sites
Trustpilot (1,788 reviews), the Better Business Bureau (1,629), Google Play (1,194), the App Store (567) and Reddit (202). Dimension Labs tagged every review with 25 fields, among them whether the writer was buying, selling or trading in, the stage of the purchase they describe, the main problem, the main praise and whether the problem was resolved. 5,377 of the 5,380 reviews were tagged. Below, an unhappy review means a one- or two-star review.
Dimension Labs analysis
Selling goes well, buying goes wrong after delivery, and an answered problem is forgiven
From 5,380 public reviews of Carvana, January 2024 to June 2026.
Sellers are happier than buyers, mostly on one site
On Trustpilot, 10.9% of reviews from people selling a car to Carvana are one or two stars. For people buying a car, it is 76.7%, seven times as high. Reviews that describe the seller pickup on Trustpilot are unhappy 8.3% of the time; reviews there about owning the car after delivery, 85.1%.
The other sites do not repeat that gap. On the App Store and Google Play, the few seller reviews are less happy than buyer reviews: 66.7% of 27 seller reviews on the App Store are one or two stars, against 31.9% of 298 buyer reviews, and on Google Play it is 50.0% of 24 against 27.2% of 394. On the Better Business Bureau both sides are almost all unhappy.
Dimension Labs analysis
On Trustpilot, sellers are far happier than buyers. On the app stores, the few seller reviews are less happy
Share of reviews that are one or two stars, by what the writer was doing and by review site, January 2024 to June 2026. Under each site: how many reviews.
Selling a car to CarvanaBuying a car from Carvana
So the cheerful seller picture comes mostly from Trustpilot, which holds 84% of the seller reviews on these four sites, while nearly half of the buyer reviews (48%) are on the Better Business Bureau, where people go to complain. Added up across sites, 21.9% of seller reviews and 73.0% of buyer reviews are one or two stars. That pooled gap of 51 points mixes very different pictures: 66 points in the sellers' favor on Trustpilot, and the reverse on the app stores. The reviews alone cannot say why sellers gather on Trustpilot. If a business asks one side of its marketplace for reviews more often than the other, the ratings will reflect who was asked, so check your own review invitations before comparing the two sides.
What sellers praise and what they complain about
The praise in Carvana reviews is mostly about the process. An easy, seamless process is the main praise in 1,304 reviews, 85% of them five stars, and a good price or value in another 305, 86% five stars. Seller reviews on Trustpilot read like this: "The easiest experience I’ve ever had!" and "The best quick way to sell a car at a very great price."
From the report: Carvana review brief

When selling goes wrong, the offer is the usual reason. A dispute over a low offer or the price is the main problem in 204 reviews, and 79.9% of those are one or two stars. The rest of the seller complaints describe the steps after the offer: a pickup that did not happen, or money that did not arrive. One seller wrote, "I couldn’t get my check when I sold my car." Another: "They never showed up." A third said that a month after the sale Carvana had not paid off their loan. Seller payments also appear in the public record: Connecticut's attorney general settled with Carvana for $1.5 million in January 2025 over complaints that included delayed titles and registrations and delayed payments to sellers.
What goes wrong for buyers
For buyers, the trouble starts after delivery. Across all sites, about nine in ten of the 1,027 reviews about owning the car are one or two stars. Reviews about browsing and shopping are unhappy 18.0% of the time. The rate has not improved over the period: 93.2% in 2024, 87.0% in 2025 and 89.0% in the first half of 2026, while the share of reviews about ownership fell from 24.3% to 15.5%.
The problems buyers name most often all come with very high unhappy rates: a mechanical defect (568 reviews, 96.5% one or two stars), a car that arrived dirty or damaged (368, 92.1%), a late or rescheduled delivery (353, 86.7%), support they could not reach (292, 88.0%), a refund refused or delayed (234, 97.0%) and a title or registration delay (138, 91.3%).
A high unhappy rate does not tell you which problem does the damage, because complaints pile up on the complaint sites. These effects use broader topic tags than the primary-problem counts above, so a review can carry several; the tag counts are in the figure below. The brief estimated each problem's effect on the chance of an adverse review, meaning a one- or two-star rating or a BBB complaint, while holding the site, the year, the quarter and the review length fixed. A mechanical defect raises it by 23.5 points (95% interval 19.3 to 27.9), unreachable support by 17.2 and a late delivery by 12.8. A refund or billing dispute adds 4.5. A title or registration mention comes out at minus 4.4 once the site is held fixed: its raw link to bad ratings comes from where those complaints are posted.
From the report: Carvana review brief

Titling is still a real problem for Carvana. The brief notes that it drew the 2022 Illinois license suspension and the 2025 Connecticut settlement. It just does little to move the chance of an adverse review.
The worst ratings come from problems nobody closed
The strongest pattern in the data is about what happens after something goes wrong. When the writer says the problem was resolved well, 5.1% of reviews are one or two stars (311 reviews). When it was never resolved, 98.1% are (1,395 reviews). Refunds owed but not paid (364 reviews) and warranty claims that were denied (157) sit at about 98% too.
From the report: Carvana review brief

Problems also stack. On the same adverse measure, reviews that name no problem come out adverse 34% of the time; reviews naming one, 82%, and two, 88%. Most of the jump comes with the first problem.
How to analyze reviews of a two-sided marketplace
Tag which side each review comes from. Buyer, seller, trade-in, financing only. A single rating for the whole business hides that the two sides can be moving in opposite directions.
Compare the sides site by site. Each review site draws a different crowd at a different moment. Here the seller advantage was large on Trustpilot and reversed on the app stores, so a pooled number would have overstated it. Comparing restaurant locations with stores like them follows the same rule.
Tag the stage, not just the topic. Shopping, pickup, delivery, ownership, support. For Carvana the split between happy and unhappy reviews follows the stage the writer is describing.
Rank problems by effect as well as count. Hold the site, the time period and the review length fixed and estimate what each problem does to the chance of a bad review. The most-mentioned problem here, money owed back (825 mentions), had one of the smallest effects. Our post on which complaints cause bad reviews found the same thing for hotels, gyms and cafés, and causal inference covers the method.
Track resolution as its own field. Whether the problem got closed was the clearest single split in this data. It is also the part a support team controls.
Join reviews to your own records. Public reviews show what customers say. Offer acceptance, pickup completion, payment timing, delivery dates, warranty claims and support tickets show what happened. Joined by customer or by week, they tell you which complaints move sales and repeat business.
The limits: reviews come from people who chose to post, often right after a strong experience, so read the rates as comparisons, not as the share of all transactions that went badly. The effects above are estimated from observational review text and survived placebo and subset checks; they suggest a ranking but are not an experiment, and the ranking depends on the estimator used. Seller numbers on the app stores rest on 51 reviews. The full analysis, with every exhibit, is in the Carvana brief.
FAQ
What do online car buyers complain about most?
In these Carvana reviews, the most common buyer problems were a mechanical defect, a car that arrived dirty or damaged, a late delivery, support they could not reach, a delayed refund and a title or registration delay. A mechanical defect had the largest measured effect on a one-star review.
Are people who sell their car to Carvana happier than people who buy one?
On Trustpilot, yes, by a wide margin: 10.9% of seller reviews are one or two stars against 76.7% of buyer reviews. On the App Store and Google Play the few seller reviews are less happy than buyer reviews, so the answer depends on where you look.
Why not use the average star rating?
It mixes the two sides of the business and the different crowds on each review site. Split by side, by site and by stage before reading anything into it.
Dimension Labs reads every review, ticket, chat and survey a business has, tags each one with fields built for that business, such as which side of the marketplace the customer is on and how far the problem got, and joins them to orders, payments and support records to show what moves revenue and repeat business. The agent then keeps the reports and dashboards up to date on a schedule. See automotive and review intelligence, or start with one file of reviews.