The car Jerry can’t see

Three questions Jerry answers wrong every day

Every automotive conversation eventually lands on one vehicle. When it does, the customer asks three things, and Jerry cannot currently answer any of them with confidence.

None of these is a conversation problem. They are all inventory problems that surface in the conversation. Jerry is downstream of listing feeds, website scrapes, and third-party syndication, and every one of those sources is a delayed copy of a copy. The marketplaces that send Jerry its leads are working from the same stale feed. When the feed is wrong, Jerry is confidently wrong, and the customer who drove in for a car that was sold on Tuesday blames the store, not the software.

There is exactly one place in a dealership where inventory is always true: the DMS. Everything Jerry knows about a car should start there.

Put an inventory truth layer under Jerry

VINTheorem becomes the inventory management solution under Jerry. It takes the DMS feed as the source of record, adds the two things the DMS does not track well, and hands Jerry a single answer per VIN. Each layer solves one of the three questions, in order, and each one is buildable on its own.

Layer 1: What is actually in stock

This is the plumbing, and it is the part nobody wants to build because it is not glamorous. The DMS sends inventory to VINTheorem every day; VINTheorem reconciles what sold and what was acquired since yesterday, normalizes it, enriches it, and serves it to Jerry. This alone removes the sold-on-Tuesday appointment.

The same sync brings in the sales data. Every delivered deal, with the unit, the gross, the days it took, and the trade that came with it, lands in VINTheorem next to the inventory. That is what turns an inventory list into inventory management: the store’s own history says which models, trims, colors, and price points actually turn here, and every pricing and buying recommendation is built on that instead of on a national average.

The same layer owns the feed going out. Today the dealer’s inventory management vendor sends the inventory to Autotrader, Cars.com, CarGurus, and the dealer’s own website, on its schedule and with its errors. VINTheorem does that job from the DMS record, with photos, descriptions, pricing, and status already reconciled, so every marketplace shows the same car Jerry sees. The dealer replaces the inventory management solution they already pay for with VINTheorem, and gets a better Jerry in the same move.

Layer 2: Where the car physically is

New cars spend weeks in transit. Used cars spend days at vendors. The DMS records a stock-in date and nothing else. A recon board in VINTheorem, shared with the vendors and the service drive, gives every unit a location and an expected front-line date. Jerry can then say “it lands Thursday, want to be first on it?” instead of promising a test drive on a car that is on a truck in Nebraska.

This matters most for the customer who researches online and buys in the store. That customer has already decided on the unit before they walk in. If Jerry brought them in for a car that was not there, the store did not lose a lead, it lost a sale it had already made.

Layer 3: Whether it can still be sold

Status is the hardest of the three because every store has its own philosophy. One dealer holds a car on a verbal. Another needs a deposit in hand. A third will sell out from under a deposit if the second buyer is stronger. There is no universal rule, so VINTheorem should not impose one.

What Podium uniquely has is the conversation. When a vehicle of interest is attached to a customer name and the conversation turns to a deposit, that is a signal no DMS ever sees. Jerry can raise the status in real time, the store can confirm or overrule it under its own rules, and every other customer asking about that VIN gets the right answer while the first deal is still being worked. The dealer sets the philosophy. Jerry applies it consistently.

How do we get the data out of the DMS?

Everything above depends on one answer. There are three ways into a DMS, and the one Podium can reach today decides whether layer 1 is a six-week project or a six-month one. The feed we want is wider than inventory: vehicle status, sales, internal repair orders, and sublet purchase orders, pulled daily, because those records are what let the recon board move on its own.

The practical path is to start every rooftop on the export it already has, ask Podium which certified interfaces it can already reach, and use the headless route only to prove out status checks with a willing pilot store while the certified path is being built.

Why inventory management is a profitable win for Podium

Better data for Jerry is the reason Podium wants this. It is not the reason a dealer writes the check. The dealer writes the check because they already pay five vendors, four of them Cox, to do pieces of this job badly and separately. Podium does not have to create a budget line. It takes one that already exists, consolidates it, and gets something none of those five can offer: every module also makes Jerry smarter.

What it is worth to Podium

• One line item instead of five. The dealer’s existing inventory spend consolidates onto the Podium contract. Podium prices below the sum of the five and still grows revenue per rooftop.
• Daily-use stickiness. The used-car manager opens the recon board and the buy list every morning. Inventory tools are the hardest software in a dealership to rip out, which is why Cox built its moat there.
• Cox loses the data seat. Whoever holds inventory, pricing, and trade data holds the dealership. Today that is Cox. With VINTheorem it is Podium, and it is the same company that already holds the conversation.
• Payroll the dealer can see. Unbiased numbers on every deal let a store run with half the desk managers. Ranked buy lists let a group run a buying team of three to five instead of eight. Those are savings a dealer principal can count, which makes the renewal conversation short.
• A second product with its own price. Self-checkout, where Jerry desks the deal and presents the F&I menu, sells separately on top of the inventory layer. Every module here is a prerequisite for it. Podium cannot get there without the inventory layer, and no competitor can get there without the conversation.

Don’t lose your identity to Cox. Use AI to tell your inventory what you value, and let it build the lot you would build if you had the time.

One feed out, and every surface shows the same car

A dealer’s inventory reaches the public through a feed, and today that feed is the most neglected system in the store. A syndication vendor picks up a file from the DMS or the lot-photo service on a schedule, reshapes it for each destination, and pushes it to the website provider, the marketplaces, the OEM’s certified program, and the social listings. Each destination gets its own version on its own cadence. Sold cars linger for a day. Prices on the website and on Autotrader disagree because they were pulled at different hours. Photos arrive late, descriptions are boilerplate, and Jerry is reading yet another copy. Nobody in the building owns it, and everybody pays for it.

The feed is the truth layer, published

VINTheorem already holds the reconciled record for every unit: the DMS facts, the recon stage, the price from the store’s own model, the status, the photos, and the decoded build with its options. Syndication is that record pushed outward, as it changes, to every destination the dealer sells through. There is no second file and no second schedule. When a car sells in the DMS, it leaves the website, the marketplaces, and Jerry in the same sync. When the price changes at the desk, it changes everywhere at once.

Recon and status decide what is published

Because the recon board and the feed share a record, the dealer sets a publishing rule instead of a person deciding car by car. A common one: publish a used car to the marketplaces only when it is front-line ready, but show it on the dealer’s own website as coming soon with the target date from the board, so Jerry can take a deposit on it. A car on deposit is marked pending or pulled, according to the store’s own status philosophy from layer 3. A new car in transit is listed with its arrival date. The rule is written once and applied to every unit, every day.

Photos and descriptions without a vendor

Photos are the usual reason a car is invisible for its first three days. Podium already moves photos every hour of the day. A salesperson or a lot attendant sends a set from the lot in a text thread, VINTheorem attaches them to the stock number, checks them against the build, and the car is live. Descriptions are written by the AI from the decoded VIN, the confirmed options, the condition evidence, and the store’s own voice, so every listing is specific to that car instead of a template with the trim name swapped in.

The feed reports back

Sales data closes the loop. Every delivered deal carries its lead source, so the store learns which destinations produce sales and which produce clicks. A marketplace that has not sold a car in six months at a monthly fee shows up in the same report as a car that has sat for sixty days, and for the same reason: it costs money every day and returns nothing. Most dealers have never seen that number for their syndication, because the vendor sending the feed is the vendor selling the placements.

The website, the marketplaces, and Jerry should never disagree about a car. They only can when they are fed separately.

Price every car from the market it is actually in

Pricing tools today rank a car against a national pool and hand the store a percentage of market to aim at. That fails in two directions. It treats a store in Bountiful like a store in Phoenix, and it has no idea what the car is costing to keep while the rank is being chased. VINTheorem prices from the comps that matter, the store’s own history, and the store’s own economics, and then explains the number in words a manager can repeat to a customer.

The market, as this store sees it

The recommendation, and the reason

Every unit carries a recommended price, the expected days to sell at that price, and the day it crosses from earning gross to losing it. The recommendation is re-run daily against fresh comps and fresh sales, and when it moves, the manager sees why: a competitor undercut, day supply changed, the unit crossed an aging threshold, or a comp sold and reset the market. A price change is a proposal with a reason attached, approved with one tap or left alone. Nothing changes on the website without a person choosing it.

Price the car the way you see the market

vAuto gives every store the same model and lets them nudge it. VINTheorem starts from the other end. The dealer describes, in plain language, how they see the market and how they want to price cars, and the AI writes the pricing model to match. The dealer reads it back in words they used, corrects it, and the model updates. No pricing philosophy has to be squeezed into someone else’s settings screen.

The last column is what makes the model honest. Once the store’s financials are in, every car carries a real cost-to-keep per day, and the aging curve stops being a guess. The desk sees the same number Jerry sees, and both know the exact day a unit becomes cheaper to sell at a loss than to keep.

What the manager sees every morning

• Units priced above market with no reason on file. Either the store meant it, and says so once, or the price moves.
• Units priced below market leaving gross on the table, which is the mistake nobody reports because the car sold.
• Units crossing the cost-to-keep line this week, with the number, so the decision to hold is a decision.
• Comps that undercut a unit overnight, and which store did it.
• Holes and surpluses, from the same market read, feeding the buy list.

What the market knows that the store does not

The same analysis runs across the whole 300-mile market, not only the cars the store owns. It surfaces the models and trims that are turning fastest nearby that the store has never stocked, the stores that consistently price beneath this one on the same units, and the price bands where demand outruns supply. That is the intelligence a buying team of eight used to gather by driving lots and reading run lists, delivered as a page.

What Jerry gets

Jerry gets the price, the floor the desk will actually take, and the comps behind both. When a customer says they saw the same truck cheaper across town, Jerry can answer with the mileage, the trim, and the accident history of that truck, because it is one of the comps. That conversation used to require a manager. Now it requires the data the manager was already looking at.

A national price rank tells a store where it stands in a market it is not in. Its own comps, its own sales, and its own costs tell it what the car is worth here, today.

Every unit has a place, a stage, a vendor, and a clock

Recon is where a used car’s profit is decided and where most stores have the least visibility. A unit lands, gets a stock number, and disappears for eleven days into a shop, a body vendor, a detail bay, and a photographer’s backlog. Nobody can say where it is without walking the lot, and every one of those days costs the flooring and fixed expense the pricing model already knows about. The recon board makes the eleven days visible and then makes them shorter.

Each unit sits in exactly one stage with a location, an owner, days in that stage, cost so far against the approved estimate, and a target front-line date. The board is the used-car manager’s morning screen: what is stuck, who has it, and what it is costing. New cars run the same board with different stages, from allocation and in-transit through PDI and accessory installation, so “where is it” has one answer for the whole lot.

Let the DMS move the cards

Most of what the board needs is already being typed into the DMS by people who do not think of it as recon data. A dealership that uses the DMS status field marks a new car in transit and then in stock. The shop opens an internal repair order when a used car goes into mechanical and closes it when the work is done. Parts or the office cuts a purchase order when a car goes out to a body shop or a detail vendor and closes it against the invoice when the car comes back. Every one of those is a stage change with a timestamp.

If the daily DMS sync carries vehicle status, repair orders, and purchase orders alongside inventory and sales, the board moves itself. A card goes to mechanical when the RO opens and leaves when it closes. It goes to a named vendor when the PO is issued and comes back when the PO closes. Nobody walks the lot and nobody updates a spreadsheet. The vendor text and the photo become the confirmation and the estimated return date, which is the part the DMS never knows, instead of the only source of truth.

The open question is the same one as the DMS feed itself: whether status, RO, and PO records come through the certified interface or the daily export for each DMS the store runs. That is the first thing to confirm with a pilot store, because the answer decides how much of the board is automatic on day one and how much is filled by the vendor texts below.

Vendors are on the board too

The stages a store cannot see are the ones outside the building: the body shop, paintless dent, wheel repair, glass, upholstery, the transport company. VINTheorem tracks every sublet unit by vendor, with the promised date, the actual date, the cost, and whether it came back right. The vendor does not need a login. Podium already knows how to reach anyone by text, so a vendor gets a message when a unit is sent, replies with a status or a photo, and the board updates. The store stops calling around on Friday afternoon to find out what is coming back Monday.

What the board feeds

• Jerry. Stage and target date become the answer to “where is it and when can I drive it?” A car in detail is “ready Thursday.” A car waiting on a body vendor is not offered for a test drive tomorrow.
• The pricing model. Recon cost accrues into the unit’s cost basis as it happens, so the recommended price and the walk-down are built on what the car actually cost, not the number the buyer wrote at the auction.
• The service department. Internal recon work is scheduled and costed like customer work, so service knows what recon is worth to them and used cars knows what service is charging.
• The buyer. Time-to-line and recon cost by model and by source flow back into the buy list. A car that always needs paint from one auction lane stops being a good buy there.

Time to line is the one recon number that matters, and the board is how a store gets it from eleven days to five without hiring anyone.

Buy the cars this store sells, from wherever they are

A dealership group runs a buying team of eight because buying is manual. A buyer’s day is looking up VINs one at a time, checking MMR, walking lanes on a hunch, and hoping the used-car manager agrees with the number afterward. The inventory is whatever that buyer happened to see. VINTheorem turns buying into a list the store builds once and works every day, from every source, with the same math.

The buy list comes from the store’s own history

Every morning the DMS sync updates what sold and what was acquired, and the buy list rebuilds. It knows which models, trims, colors, and price bands turn at this store and how fast, and it knows what is moving quickly within 300 miles that the store has never stocked. It compares that against what is on the lot and in recon today and produces the holes: the cars the store should own and does not, with a target count, a target price band, and a target days-to-turn for each. That list is the buyer’s assignment, and it changes as inventory changes.

Every source runs through the same appraisal

Trades from the desk, cars in the service lane, physical and online auctions, wholesaler emails, dealer-to-dealer offerings, off-lease and rental de-fleets. Today each of those is a different workflow with a different person and a different number. In VINTheorem they are one queue. A VIN comes in from anywhere, gets appraised the same way against the same buy list, and comes back ranked buy or pass with a max bid.

And one thing the dealership world has never had

Every used-car manager gets wholesale lists in their inbox all day: dealer-to-dealer offerings, rental and fleet de-fleets, off-lease remarketing, wholesaler spreadsheets, auction run lists. Each one is a list of VINs with a price, and each one is a race. Today a manager either ignores the list, or hands it to someone who looks up VINs one at a time until the good cars are gone.

VINTheorem already does bulk appraisal. Point it at the inbox. A wholesale list lands in email, is parsed, and every VIN on it is appraised against the store’s market, its own history, and its current holes, ranked buy or pass with a max bid, within a couple of minutes of arriving. The manager opens the list already knowing which six cars to call on. No vendor in the industry offers this, and it is the feature a used-car manager will describe to another used-car manager unprompted.

The auction works the same way, without waiting for anyone’s API. The buyer downloads the run list as a CSV from whichever auction house they use, Manheim, ADESA, ACV, or a local independent, and drops it into VINTheorem. The program matches every lane against two things: what this store has already proven it can sell, from the imported sales data, and what is moving quickly within a 300-mile radius of the rooftop. That second list is the one buyers miss. A model that turns in eleven days two hundred miles away, that this store has never stocked, shows up on the buy list with the evidence next to it. The buyer walks the lanes already knowing which units to bid on and how high.

How the max bid is built

The number the buyer bids to is not a book value with a discount. It is worked backward from what the car will actually do at this store.

1. 1Expected retail from the store’s own pricing model, for this trim, mileage, and color combination, in this market.
2. 2Less recon, estimated from what this model from this source has actually cost on the recon board, not a flat allowance.
3. 3Less transport from that auction or that wholesaler to this rooftop.
4. 4Less holding cost at the store’s flooring and fixed expense for the target days-to-turn.
5. 5Less the target gross the dealer set in the pricing philosophy.
6. 6Max bid. The buyer can go to it with confidence and stop at it without an argument.

Every purchase is scored

A bought car is tracked from the bid through recon to the sale. Did it turn in the target days? Did recon come in at the estimate? Did it make the gross? Those answers roll up two ways. By buyer, so the group knows who is buying well. By source, so the group knows which auction lanes, which wholesalers, and which dealer partners deliver cars that turn and which ones deliver cars that sit. A lane that always needs paint stops being a good lane. That scorecard is why a group can run a buying team of three to five instead of eight: the list does the finding, the appraisal does the pricing, and the buyers do the buying.

Jerry gets a list of cars that are about to exist. A customer waiting for the right truck can be held for a unit the buyer is bidding on this morning.

Once Jerry knows the car, the rest of the store opens up

The MMR question

Cox owns the number every buyer trusts. We do not have to replace it on day one. We present it alongside ours.

Give dealers the option and let them decide which number is closer to what they paid and what they sold for. At some point Cox will pull MMR access. By then the dealer has months of side-by-side evidence, and the question of which number to trust has already answered itself.

An unbiased number on every deal and every trade

The desk is where a dealership makes or loses its money, and it runs on the least data of any department. A manager’s number comes from what they think the car is worth, what mood they are in, and which salesperson is standing in front of them. Two managers give two numbers. The cost of the car sitting another week is not on the screen, so it is not in the decision. VINTheorem puts it there.

What the desk sees on every deal

The manager’s job becomes deciding rather than computing, and the number is the same for every salesperson and every customer. That is what unbiased means at the desk. It is also why a store can run with half the desk managers it has now: the tower stops being where deals wait for someone’s opinion.

Appraisals in minutes, from the store’s point of view

An appraisal in VINTheorem starts with the VIN and ends with two numbers. It decodes the build and options, pulls live comps within 300 miles, checks title and accident history, and reads the customer’s photos for condition, with a person confirming what the model saw before it counts. Then it applies the store’s own rules. If the car fills a hole on the buy list, the pay-up rule kicks in and the retail number is what the store should offer to keep it. If it does not, the wholesale number is what the store can safely pay to flip it. The manager sees both, with the evidence, in minutes.

• The desk. The trade on a deal is appraised while the deal is being structured, not after the customer has been waiting forty minutes.
• Jerry. A customer texts photos and mileage and gets a real range in the conversation. Podium already has the photos; VINTheorem turns them into a number.
• The service lane. Every repair order is an appraisal waiting to happen, with the technician’s eyes on the car.
• The public. The same engine behind a Podium-branded instant offer on the dealer’s site, which is the KBB Instant Cash Offer replacement.

Every appraisal is remembered

Appraisals are tracked against what happened next. Did the store get the trade or lose it, and to what number? If it got it, what did it sell for and how fast? Lost trades become a report instead of a rumor, and the pay-up rules tighten over time from the store’s own results. A manager who consistently loses trades by a few hundred dollars finds out. So does a manager who consistently overpays.

The desk should be the best-informed seat in the store. Today it is the least. That gap is the margin.

Our own instant offer, and our own Accu-Trade

Kelley Blue Book Instant Cash Offer is the consumer-facing side of Cox’s acquisition machine: the customer asks for a number on KBB.com or the dealer’s site, Cox generates it, and the dealer pays for the lead and honors the offer after an inspection. Accu-Trade is the dealer-facing side: a condition-adjusted appraisal with a guaranteed floor the dealer can lay the car off to if they do not want to keep it. Together they mean Cox owns the customer, Cox owns the number, and the dealer rents both. VINTheorem already produces the number. Podium already owns the conversation. That is both halves.

What we can and cannot do

The constraint is trademark and terms of use, not copyright. Kelley Blue Book, Instant Cash Offer, Accu-Trade, and MMR are Cox brands, so the product cannot carry those names or be described as a version of them, and their values cannot be pulled, cached, or displayed as our own. None of that stops anyone from building a competing instant offer from their own data, which is what every other valuation company has done. The offer is generated by VINTheorem from live market comps, real transactions, and the store’s own sales history, under a name Podium chooses. Where the document above shows MMR beside our number, that is the dealer’s own MMR access displayed beside a separately computed value, which is comparison, not reuse. Counsel should confirm the exact presentation, and the plan should assume Cox eventually turns that access off.

How ours works

The customer starts on the dealer’s website, in a text thread, or in the service lane. They give a plate or VIN, mileage, and a few photos. VINTheorem reads condition from the photos, appraises the car against comps within 300 miles and the store’s own history, and applies the store’s pay-up rules. If the car fills a hole on the buy list, the offer goes up. If it does not, the offer is the wholesale number. Jerry delivers it in the conversation, answers the questions a customer actually asks, and books the appointment. The store confirms condition when the car arrives, and the offer stands or adjusts against what was disclosed.

The floor only Podium can offer

Accu-Trade’s real product is the guarantee: a dealer can make an aggressive offer knowing Cox will take the car at the floor. Podium can build a better floor without owning an auction. When a store passes on a car, the same appraisal goes to nearby Podium dealers whose buy lists have that hole, and the first one to take it at the floor gets it delivered. The customer still gets one offer from one store. The store still makes the deal. The car goes to the rooftop that wants it most, inside the network, with no auction fee. Every Podium dealership makes every other Podium dealership’s offer safer, and that is a network effect Cox cannot copy.

Cox built the instant offer to own the customer. Podium already owns the customer. Building our own is not catching up; it is taking the lead away.

Self-checkout: Jerry desks the deal and presents the menu

This is a product on its own, not a feature of the inventory layer. Plenty of customers would rather check themselves out of a dealership the way they check themselves out of everywhere else, and no store can offer that today because the desk and the F&I office are staffed by people whose job is to be in the middle. A dealer will pay for the program that lets the customer do it, because every deal that runs through it is a deal that did not need a desk manager or an F&I manager to build.

Every piece of the desk is already in the building once the layers above exist. The car and its true status come from the truth layer. The price and the walk-down come from the pricing model the dealer described. The trade number comes from the appraisal engine. The customer, their conversation, and their preferences are already in Podium. What is missing is the person who sits in the tower and puts those together, and that person is the bottleneck in every store on a Saturday.

1. 1Jerry lands on a car. The customer has a vehicle of interest, it is on the lot and front-line ready, and the status is available. Layers 1 through 3, already done.
2. 2Jerry works the trade. Year, make, model, mileage, and photos come in over text. VINTheorem returns a real number in minutes with the store’s pay-up rules already applied. Jerry presents it as a range or a firm figure, whichever the dealer’s policy says.
3. 3Jerry structures the deal. Selling price from the pricing model, trade allowance, cash down, and term. Payment options come from the lender programs the store actually uses, with rate tiers by credit band. The customer sees three ways to buy the car, not one.
4. 4Jerry takes the credit application. The customer completes it in the conversation and Jerry submits it through the store’s existing lender portal. Approvals and stipulations come back into the same thread.
5. 5Jerry presents the F&I menu. Service contract, GAP, prepaid maintenance, tire and wheel, appearance protection, priced into the payment, with every product explained and every disclosure shown. The same menu, the same way, to every customer.
6. 6A human signs off. A licensed manager reviews the structure, the approval, and the products, and releases the deal to signing. One manager can review what four used to build, and the review is the only step that needs a person.

Why the menu belongs in the conversation

The F&I office exists because the store needed a person to gate the paperwork and to sell products in the last fifteen minutes of a four-hour visit. Customers hate that room, and product penetration depends on which manager happens to be working. In the conversation, the menu is presented when the customer is deciding how to pay, not after they have already decided they are done. They can read it, ask Jerry what GAP actually covers, and take it or leave it without a person across the desk.

It is also the most defensible way to present products. Every customer is offered every product at the same price on the same menu, and the record of what was shown and what was chosen is in the thread. That is exactly the consistency regulators ask for and the one thing a busy F&I office cannot promise.

Why a dealer pays for this separately

Desk managers and F&I managers are the two most expensive seats on the variable side of the store, and a dealer staffs both for the Saturday peak, not the Tuesday average. A program that structures the deal, takes the application, and presents the menu inside the conversation lets the store staff for review instead of for volume. Half the desk managers and half the F&I managers is the honest sizing. The customers who still want a person get one; the ones who do not are no longer waiting behind them. That is a payroll line the dealer principal can see on the first month’s statement, which is what makes it sellable at a price of its own.

A dealer running this can sell the car faster, price more aggressively than the store across the street because each deal costs less to build, and give the customer the experience they keep saying they want: digital, transparent, and lightly human assisted.

Capture service with the sales data you already imported

Service is the most profitable department in the store and the one that quietly loses the most customers. Most buyers come back once or twice under warranty and then drift to the independent shop down the street, and the dealer never sees the moment it happens. Podium already lives in the service drive. What it has been missing is the trigger.

The trigger is in the sales data VINTheorem is already importing from the DMS. Every delivered deal is a customer, a VIN, a delivery date, mileage at delivery, and the trade that came with it. From the VIN, VINTheorem knows the factory maintenance intervals. From the delivery date and the customer’s driving pattern it can estimate where the odometer is today. That is enough to know, for every car the store has ever sold, what service is due and when, without waiting for the customer to remember.

The lane closes the loop

Once the customer is in the drive, the same data works in the other direction. Every repair order is a vehicle with a known owner, known mileage, and a technician’s eyes on its condition. VINTheorem values it against the store’s current holes and hands Jerry a real offer while the customer is in the waiting room. The cars the store buys that way are the cheapest inventory it will ever own, and they arrive with a service history the store wrote itself.

Recon runs through the same shop. Layer 2 already tracks every used unit through service, so internal work is scheduled, costed, and measured like any customer RO, and the service manager can see what recon is actually costing the used-car department per unit.

A dealer sells a car once. With the sales data driving service, they get to talk to that customer twenty more times before the next one.

Build order

Each layer ships on its own and each one makes Jerry measurably better before the next one starts.

1. 1DMS inventory and sales into VINTheorem, VINTheorem into Jerry and out to every third party. Solves in-stock, replaces the syndication vendor, and starts the sales history every recommendation is built on. The fastest win and the one that removes the most embarrassing failures.
2. 2Recon board with vendor tracking. Solves location, inside and outside the building, moved automatically by DMS status, RO, and PO records where the store keeps them. Also the first thing a used-car manager will actually log into every morning.
3. 3Status inference from conversations. Solves availability. Podium is the only company positioned to do this, because it owns the conversation.
4. 4Appraisal, buying, and the desk. The revenue product. Bulk appraisal is already built; the email and auction-CSV buy lists sit directly on top of it.
5. 5Service triggers from sales data. Every sold vehicle gets a maintenance clock, and Jerry runs the campaigns. Pure retention revenue on data the store already sent us.
6. 6Self-checkout: desk-to-F&I in Jerry, sold as its own program, and the MMR side-by-side. The part that changes what a dealership is.