Build the buy list from what your store already sold.

What a sold-history buy list is

Take every retail deal the store closed in the last 30 days. Group them by year, make, model, trim and mileage band. For each group, write down three numbers from the DMS: how many you sold, the median days to sell, and the front and back gross. That table is the store's real demand. It is not an opinion and it is not a national average. It is what walked out of your door, at your price, in your market.

A buy list built from that table has one rule: a car on the block only makes the list if the store has sold that cohort recently and made money on it. Everything else is somebody else's car.

Sold Cars and Replacements

Why the national book cannot do this

A book value is a national clearing price. It cannot know that your store sells Tacomas in nine days and Tahoes in sixty, or that your 2018 Equinox buyers come from one zip code with one lender. Buying from the book puts you in a bidding war with every store that reads the same book. Buying from your own history puts you in a lane where you already know the ending.

• The book says $19,400. Your history says you sold four of these at $18,990 in 24 days with $2,100 front. That is a buy up to a number you can compute, not a feeling.
• The book says $27,000. Your history says the last two sat 80 and 94 days and went to the wholesale truck. That is a pass at any price.
• The book has no mileage band. Your history shows the 60k-to-80k version turns and the 100k-plus version does not.

The method, by hand

1. Export the last 30 days of retail deals from the DMS with stock number, VIN, sale date, sale price, cost, front gross, back gross and days in stock.
2. Group by year, make, model, trim and a 20,000-mile band. Keep groups with two or more sales, or one sale with strong gross.
3. For each group, note median days and median total gross. Cohorts under 35 days and over $1,500 total gross are your A list.
4. Open the run list for this week's sales and match it against the A list by year, make and model, then filter by mileage band and condition grade.
5. For each match, set a max bid from the cohort's median sale price minus your target gross, fees, transport and typical recon.

Done by hand, this takes a used-car manager most of a Sunday. It is also exactly why most stores do not do it.

How DealerCortex does it every morning

The Sold Cars and Replacements board builds the cohort table from the store's own DMS history automatically, then finds the replacements running at Manheim this week, lane by lane, with the light, the condition grade and the MMR. The Pre-Sale Buy List scores every car on the run against that model set and the store's buying rules, strips the cars a buyer should never touch, and is ready before 7 AM Eastern on every sale day. On one Illinois franchise store the method matched 958 cars out of a 24,000-car run to proven cohorts after the filters removed roughly 14,000 cars with no condition report, branded titles, mileage over the cap, grades under the floor, as-is flags and mechanical announcements.

What changes in 30 days

The sales history loads on day one. By the first sale day the replacement board is live. The wholesale truck leaves lighter within the first month because the cars that used to be bought on a feeling are not bought at all.

Questions dealers ask

How much sold history do you need to build a buy list?

Thirty days gives a usable cohort table for a store selling 60 or more used cars a month. Twelve to twenty-four months gives seasonality and a much larger model set. DealerCortex loads the full DMS history at onboarding.

What if my DMS export does not have gross?

Sale price and cost are enough to compute front gross. Back gross comes from the F&I columns when the DMS has them; without them the board uses front gross and days to turn.

Does the buy list work for OVE as well as Simulcast?

Yes. The OVE Buy Agent scans every live OVE car daily and the Pre-Sale Buy List covers the Simulcast run for each sale day. Both match against the same sold cohorts.