The price comparison used to happen in the showroom, across the desk, with a manager's gut as the only instrument. It now happens before the shopper arrives, and the store is rarely the one running it. A store that does not know its own market position by the time the customer sits down is negotiating blind against someone who is not.
This is not a technology story. It is a sequencing story. The comparison exists either way; the only open question is whether the dealer sees it first, and whether the customer sees it only when it helps.
What Changed When Shoppers Started Sending AI to Do the Comparing?
CBT News reported on October 27, 2026 that the CEO of CarEdge sees AI shopping tools raising the stakes for dealer pricing transparency.✓ Oct 28 The detail worth a dealer's attention is not the opinion. It is the volume.
"CarEdge's AI shopping agent has contacted dealerships 160,000 times since July 2025, and customers who use the tool convert at three times the rate of those who don't." Reported by CBT News | #1 Source for Automotive News & Dealership Intelligence
Read that as an operations signal. A shopper who arrives through an agent has already collected quotes. The first conversation is not discovery. It is a comparison being closed, and the store either already knew where it stood or finds out at the desk.
The old defence was information asymmetry: the customer did not know what the store down the road was asking. That defence is gone for any shopper willing to use the tool, and the number above suggests many are.
Why Is a Manager's Gut-Feel Markdown the Wrong Instrument?
A markdown made on instinct has two failure modes, and both cost gross. Price too high and the shopper never sits down, because the agent already moved on. Price too low and the store gives away margin to a customer who would have paid more, and nobody ever records the difference.

The gut read also cannot see the right comparison. The loudest competitor price is rarely the right one. Two stores can advertise the same new vehicle at the same number and be selling very different deals, because one starts from a sticker loaded with add-ons and the other from a bare window sticker. A raw price comparison punishes the honest store and rewards the one that moved its charges around.
That is the second loser in this story: the agency deck. A monthly report that shows impressions, clicks and a cost per lead never shows the customer's actual alternative. It describes the store's own advertising and stays silent on the competing price that decides whether the click turns into a visit. If your marketing partner cannot tell you where your advertised price sits against the same vehicle at stores 50 to 100 miles away, it is reporting on activity, not on position.
The pricing exposure runs the other way too. A number the shopper sees in an ad is a number a regulator reads, a point made in our piece on hidden fees starting on the website. Position and presentation are the same problem seen from two sides.
What Does a Fair Market Read Actually Have to Measure?
Fair means each competing listing is priced off its own window sticker, before dealer fees, with only the rebates every buyer gets. Anything else compares a discount to a different discount. A store that lists a lower final number after stacking fees below the line looks cheaper than it is, and the comparison has to see through that.
It also means the read is a snapshot per deal, not a daily repricing engine. A comparison built for one customer proposal answers one question: given this vehicle, this price and this market today, are we competitive? It does not need to move anything to answer it.
And it means the comparison has a defined verdict, not a vibe. A price is competitive when its discount off sticker is at least the market median and no deeper than the 90th percentile of the dealers compared. Below the median, you are expensive. Beyond the 90th percentile, you are giving money away. Both are findings the store needs, and neither belongs in front of the customer.
Why Should the Customer See the Comparison Only When the Store Wins?
Because a comparison is a sales document and a diagnostic at the same time, and those two audiences need different things. The customer should see proof when the proof is true and favourable. The store should see the finding when it is not.

Showing a shopper a market read in which the store is expensive does the dealer no good and gives the customer a reason to leave. Hiding a market read from the store because it is unflattering does the dealer real damage, because the problem persists. The sane design routes by verdict: competitive goes to the proposal, too high or too low goes to the manager as a recommended adjustment.
There is a trust point inside this. Competitors are never named. The customer is shown where the price sits in the market, not a rival's logo and dealer name. That keeps the document defensible as a statement about the store's own price and avoids turning every proposal into an attack ad.
Nothing in the loop changes a price on its own. A recommended adjustment is a recommendation. The manager owns the number, as the manager should.
What About Used and Certified Vehicles, Which Have No Window Sticker?
A used car has no sticker to price against, so the read changes shape. The price in dollars is compared with comparable pre-owned listings at the vehicle's mileage, inside a mileage band around its odometer, one car per dealer, with the store's own units left out. The search widens from the nearest zone until five dealers answer: 50 miles in a metro, 100 suburban, 250 rural, 500 ultra-rural.
The customer reads the store's price against the average list price of those comparables, while the internal verdict reads the median, with competitive meaning between the 10th percentile and the median. Every pre-owned proposal also carries a rarity tag reflecting how far the search had to reach: a vehicle found at 50 miles is a popular choice, at 100 miles hard to find, at 250 rare, and at 500 or too few dealers even then, super rare. That tag is useful to a customer and to a manager for the same reason: it says how much leverage the shopper really has.
How AUTONOMi Solves This
AUTONOMi Desking packages the options a salesperson desks in Desking, under Present, into one customer proposal: a web page and a US Letter PDF, on the dealer's own domain when it has one.✓ Oct 28 The proposal carries each desked deal with its full disclosure and the price walk from Total MSRP down to the selling price and incentives.✓ Oct 28
For a new vehicle, the market presentation uses MarketCheck listings of the same vehicle at other dealers within 50 to 100 miles, each priced off its own window sticker before dealer fees and with only the rebates every buyer gets.✓ Oct 28 The store's advertised price is compared with them, and the comparison is shown to the customer only when the price is competitive: a discount off sticker at least the market median and no deeper than its 90th percentile.✓ Oct 28 When the price is too high or too low, the finding goes to the store as a recommended adjustment and never to the customer, and competitors are never named.✓ Oct 28
Inventory prices are not changed automatically, and the read is built per proposal rather than as a daily lot repricing.✓ Oct 28 A certainty score reports the share of 1,000 resamples of the compared dealers that return the same verdict.✓ Oct 28 Used and certified cars are read against comparable pre-owned listings at their mileage, widening from 50 to 100, 250 and 500 miles until five dealers answer, with a rarity tag for the zone reached.✓ Oct 28
A second read, shopper demand, comes from the store's own Google Analytics: distinct shoppers on pages titled with the vehicle's year, make and model over the last 28 days, printed only at 50 or more.✓ Oct 28 It is never presented as live viewers. The result is a proposal that answers the comparison before the customer asks, and a manager who sees a recommended adjustment instead of discovering a problem at the desk. The same record-first habit runs through AXIOM's decision ledger, where the aim is that a number can be explained after the fact, not just defended in the moment.
Where Does Pricing Go Once Every Shopper Has an Agent?
The direction is clear. As more shoppers send agents ahead, the store's advertised price stops being a headline and becomes an input to a comparison it does not control. The stores that do well will not be the ones that guess better. They will be the ones that read their own position first and act on it before the customer's tool reports it for them.
That favours stores with a record over stores with a feeling, and it favours honest price stacks over clever ones. The same pressure is pushing enforcement attention toward the advertised number, as we covered in our look at the advertised price being judged first. A store that knows where it sits and can show the comparison when it wins has less to fear from either direction. If you want that read running on your own deals, sign up and see Desking proposals on your live inventory.
Source: CBT News | #1 Source for Automotive News & Dealership Intelligence



