Reputation management for auto dealers: escaping the CSI trap

In December 2024, the FTC and the Illinois Attorney General took Leader Automotive Group, a Chicago-area group owned by AutoCanada, to court and came away with a $20 million proposed judgment, the largest the FTC has ever secured against an auto dealer. Most of the complaint was about overcharging. But keep reading and you hit the review program: management required employees to post fake positive Google reviews and withheld bonuses from staff who didn’t produce them. At one store, staff refused to return a buyer’s car keys until she posted a five-star review of the dealership.

Nobody at that store invented that behavior. It was trained into them. For decades the factory has graded dealerships on a private satisfaction score, CSI, and attached real money to it, and in doing so it taught an entire industry that a satisfaction metric is something you work: coach the customer, intercept the unhappy ones, chase the perfect score. Then public reviews started deciding which store the customer drives to, dealerships ran the same play on Google, and that play is now a federal violation priced at $53,088 apiece. This post is about running reviews legally at a dealership while the CSI machine keeps shoving you the other way.

The machine that taught your store the wrong habits

If you’ve never worked retail automotive, here’s the mechanism. Two or three days after a sale closes or a repair order cashiers out, the manufacturer emails the customer a satisfaction survey. Sales and service are scored separately, so a store can have a healthy service score and a showroom on fire, or the reverse. A “good” score starts somewhere around 85 depending on the brand, and the factory really wants 90-plus.

And the score is not decorative. CSI feeds dealer bonus payouts, it feeds eligibility for co-op marketing money, and at plenty of brands it influences allocation, meaning the store with the better score gets more of the inventory everyone actually wants. A bad quarter of surveys is a real financial event.

The surveys also grade like a figure-skating judge from a hostile country. On most brand instruments, anything short of top marks counts against you: a 9 out of 10 isn’t an A minus, it’s a miss.

So dealerships built rituals around the survey, and if you’re in the business you can recite the delivery speech from memory: “You’re going to get a survey from the manufacturer in a couple of days. Anything less than a perfect score counts as a fail for me personally. If there’s any reason you can’t give us all top marks, call me first and let me fix it.” Some stores went further: the pre-filled iPad at delivery, the service advisor who “helps” the customer through the email, the F&I office with the survey coaching sheet in the folder next to the GAP brochure.

Now the uncomfortable part: essentially none of that is illegal. The CSI survey is a private instrument between you, the customer, and the factory. Coaching it may annoy your zone rep and it definitely corrupts the data, but the FTC has no opinion about it.

The trouble started when stores pointed the same ritual at the public internet. Screen every customer with a “how did we do?” text and send only the happy ones a Google link, which is gating. Offer a $25 gas card for five stars, which is an incentivized review. Put each salesperson on a five-review monthly quota with a spiff attached. And at the far end of the exact same road: Leader Automotive, where the store just wrote the reviews itself and docked pay until people cooperated.

The law caught up in October 2024

The FTC’s Consumer Reviews Rule took effect October 21, 2024, and it changed the economics of faking. Fake or AI-generated reviews, purchased reviews, reviews from employees or family without disclosure, and suppressing or burying negative reviews are all violations, and for the first time each one carries civil penalties: $53,088 per violation after the 2025 inflation adjustment, a figure that has held through 2026. Per violation means per review. Thirty fabricated five-stars is a theoretical $1.6 million of exposure sitting on your Google profile, findable by anyone with a subpoena and a scroll wheel.

Leader was the demonstration case, and the details matter for every dealer principal reading this. The FTC didn’t fine a rogue salesperson. It went after the group, because management set the quota and controlled the bonuses. If your pay plan or your Monday save-a-deal meeting creates pressure to produce five-star reviews, the store owns whatever your greenest employee does with that pressure at 8:55pm on the last Saturday of the month.

Google closed its side too

Google’s review policy has prohibited gating and incentives for years, and the incentive ban applies regardless of sentiment. A gift card for “an honest review” is still a violation; you’re paying for volume, and Google treats it the same as paying for stars. In April 2026 Google added two rules aimed straight at dealership habits: no staff review quotas, and no prompting customers to name a specific employee in their review.

That second one quietly killed a beloved dealership CRM feature. Plenty of the industry’s review tools counted per-salesperson name mentions so stores could pay spiffs on them, which is why every dealership review for a decade read “ask for Tony!” That whole loop, the prompt, the mention, the spiff, is now against policy on Google’s side and evidence of an undisclosed incentive program on the FTC’s.

And Google’s enforcement doesn’t involve a hearing. Detected gating or incentives gets reviews bulk-removed, and repeat offenses can freeze new reviews on the profile entirely. For a store whose profile is its front door, that’s a worse outcome than most fines.

Stop treating the review like a survey

Here’s the reframe that actually fixes this, and it cuts against every instinct CSI installed: the OEM survey and the Google review are different instruments, and the fastest way to get a store in trouble is treating the review ask like a CSI ask.

The survey is a private diagnostic that the factory grades you on. The review is public marketing that your next customer grades you on. Different owner, different audience, different rules. Pre-selling the score, filtering who gets asked, coaching the wording: on the survey side that’s merely data-corrupting. On the review side it’s the specific conduct the Consumer Reviews Rule was written to punish.

It also doesn’t even serve the goal. Shoppers don’t trust a 5.0 with 400 identical-sounding raves; they trust a 4.6 with visible negatives and composed responses, because that distribution looks like reality. BrightLocal’s 2026 local consumer review survey found 97% of consumers read reviews for local businesses, and recency now weighs about as heavily as the rating itself, with 74% wanting reviews from the last three months. A gated trickle of coached five-stars loses on the two dimensions that matter, volume and freshness, to an unfiltered stream of honest ones.

So run them as separate programs. Keep whatever survey ritual your zone rep tolerates on the survey. Strip every trace of it from the review ask. The legal review ask is uniform and boring: every customer, same message, same timing, no preview of sentiment, no reward, no quota, no “mention Dana.” Boring is the compliance feature.

The ask, sales side and service side

The two departments need different plays, same rules.

Sales. The ask goes out by text the morning after delivery, with a direct link to the Google profile. Not at the delivery itself, and never in the same conversation as the survey speech. The OEM survey lands 48-72 hours later, and if your review ask arrives wrapped in the “all top marks or I fail” framing, the customer merges the two instruments in their head and you’ve just coached a public review. Put daylight between them: review ask on day one, let the factory’s survey be the factory’s problem on day three.

Service. This is the volume engine, and most stores ignore it. A showroom moving 80 units a month generates 80 review opportunities. The service drive behind it closes 900 repair orders in the same month. The industry average is about 11.7 new Google reviews per store per month, while the top-performing groups clear 100 or more, and they get there on the drive, not the showroom. Automate a text at RO close, every customer, warranty work and oil changes included. The customer who only came in for a recall still counts, and per BrightLocal, her March review is worth more in June than your best January one.

As for where to point the asks: Google holds about 79% of a dealership’s reviews per location, so the Google Business Profile for each rooftop is the main arena. But this vertical has a second venue worth feeding: DealerRater reaches 14 million-plus monthly readers across 41,000-plus dealerships, and the same 2018 DealerRater/Versta research found 86% of car shoppers doing their homework online before they ever touch your lot. Shopper behavior has only moved further online since then. Alternate destination links if you want both profiles growing; just never condition the destination on how happy the customer seemed at the cashier window.

The wider mechanics of asking, timing, wording, follow-up cadence, and where the FTC lines sit, are covered in our guide to getting more reviews without breaking the rules. If you’re building the whole program from zero, start with the small-business reputation management guide and layer the dealer specifics from this post on top.

Reviews with a salesperson’s name on them

Counting reviews per salesperson is not just a reporting choice at a dealership. DealerRater and several other dealer platforms attribute reviews to an individual, complete with their own profile and their own star average, and Google reviews name people constantly even without formal attribution. That creates a political problem no other vertical has to solve.

A few ways it goes wrong. Top producers accumulate a personal review profile that follows them to the store across the street, so an asset you paid to build partly belongs to an employee. New hires with four reviews and one bad one carry a 3.5 average that is statistically meaningless and visible to every shopper. And when reviews get tied to spiffs, you have created a private incentive to solicit selectively, which is the gating problem again wearing a different hat.

The rule that holds up: pay on review volume, never on review score. Volume is under the salesperson’s control and asking everyone is exactly what you want. Score is not under their control, and paying on it funds the behavior the rest of this post is about stopping.

When someone leaves, resist the urge to purge. A glowing review naming a salesperson who quit in March is not a liability, it is evidence the store hires well, and scrubbing it looks worse than leaving it. The trickier case is a negative review naming someone who is no longer there, which has its own handling rules when a review names a former employee. The short version for dealers: do not make “that person no longer works here” your entire reply. Every shopper reads it as the store throwing an ex-employee under a bus to avoid a systems problem.

One angry customer, five listings

Dealership reviews do not live in one place, and this is the operational detail that separates auto retail from a restaurant or a dental practice. A single unhappy service customer can post on Google, DealerRater, Cars.com, and Edmunds in the same evening, and several of those platforms syndicate or mirror content between each other and onto your own website’s review widget. Reviews on manufacturer-affiliated sites can also feed the brand scorecard, so the same complaint can hit your public rating and your allocation.

Three things follow from that.

Every surface needs its own reply, written separately. Copy and pasting the same paragraph to four sites is visible, because aggregators surface all four to the same shopper. It also reads as automated at exactly the moment you need to sound like a person. Change the opening line and the specific commitment on each one.

Reply on the platform with the worst distribution first. Not the one with the most reviews. If a Cars.com review is being mirrored onto your own site and into a third-party shopping feed, it is doing more damage per word than a Google review buried on page three.

Do not chase removal across all of them. Each platform has its own policy and its own appeal path, and a complaint that violates nothing on Google may violate a dealer-platform rule about naming employees, or the reverse. Read the actual policy before you flag. Google’s prohibited and restricted content policy (current as of 2026) is narrower than most dealers assume. “It is not accurate” is not a removal ground anywhere.

A worked example

Brennan Toyota in Fort Wayne catches this two-star on a Tuesday:

“Dropped my Highlander off at 7am for a recall and an oil change. Nobody called all day. When I finally reached a human at 4:30 they told me the recall part wasn’t even in stock and they’d known that since morning. Burned a vacation day for an oil change. Two stars because the shuttle driver was nice.”

The CSI-brain response is to treat this like a survey save: apologize hard, promise the world, get the score changed. The better response reads like a service manager who actually fixed something:

“You’re right, and I won’t dress it up: we should have called you by 9am when the parts screen showed that component on backorder, not left you to find out at 4:30. That’s a process failure, not a busy day. As of this week, any repair order with an unavailable part triggers a call before 10am, and I check the exceptions myself. When your part lands I’d like to handle the recall personally, with a loaner so you don’t lose another day. Ask for Dana Whitfield, service manager.”

No begging, no coupon, no request to revise the rating, which keeps it clean under both the FTC rule and Google policy. The customer came back for the recall and updated to four stars eleven days later on her own. And the response keeps working after she’s gone: every shopper who scrolls the profile’s negatives sees a named manager describing a specific fix, which does more for close rates than three anonymous five-stars. The structural rules for responses like this one are in the negative review response playbook, and if you want a starting point already tuned to service-lane and showroom complaints, there’s a template set written for dealerships that handles the common patterns: the parts-delay review, the payment-packing accusation, the “they sold my car out from under me” deposit dispute.

That last category, by the way, is one generic reputation advice never covers: dealership negatives routinely contain allegations with legal weight, deposit disputes, spot-delivery unwinds, as-is warranty fights. When a review alleges something your state AG would care about, the response still gets written within 24 hours, but it gets read by whoever handles your legal exposure before it gets posted.

The part nobody tells you

Gating was a bad trade even before it was illegal.

Every customer your filter screened out was a review you didn’t get, and at an industry average of 11.7 reviews a month, volume is the battleground. The store that asks everyone and eats the occasional two-star outruns the store that pre-screens, because recency and volume compound while a curated average just sits there. The gate never protected your reputation. It capped it.

And one more thing your twenty group probably hasn’t mentioned: the factory reads your Google profile too. Zone reps walk into planning meetings with your public rating in the deck next to your CSI. The private score and the public one are converging into a single question, is this store worth giving cars to, and only one of the two is visible to every shopper, every employee you’re recruiting, and every regulator with a browser.

CSI taught your store that a score is something you manage. The FTC just repriced that lesson at $53,088 per violation. Ask everyone, gate no one, answer everything, and let the distribution be what it is. It will be better than you expect, and unlike the coached version, it will hold up when someone looks closely.