Dealership Inventory Source Attribution

Most stores can tell you what a car cost. Far fewer can tell you which sourcing channel keeps producing the units that actually turn.

Ask a used car manager where their inventory comes from and you will get percentages. Ask which of those channels produced the units that turned fastest at the best gross over the last two quarters, and the room goes quiet. Not because nobody cares, but because the answer lives in three systems and one person's memory.

That gap costs real money in a market where sourcing is the constraint. When auction prices move, a store that knows its consumer acquisition channel produced a better return per unit will shift effort there deliberately. A store without attribution shifts on feel, usually late, and usually toward whatever channel the last good week came from.

The frustrating part is that this is not an analytics problem. It is a data entry problem wearing an analytics costume. The information exists at the moment of purchase and gets lost within a week because nobody wrote it in a field that anything reports on.

A taxonomy that is too coarse tells you nothing and one that is too fine never gets filled in. Most stores land somewhere around six to eight categories. A workable starting list:

• Trade in on a retail sale. The unit came in against a delivery.
• Consumer acquisition. A local owner who asked for an offer on their vehicle and sold it to you without buying one.
• Service drive. Acquired from a customer in for service, which behaves differently enough from a sales trade to be worth its own line.
• Physical auction and online auction, kept separate because the fee structure and the risk are not the same.
• Dealer trade or wholesale purchase.
• Fleet, lease return or rental buy.
• Street purchase or walk in.

Two rules make the taxonomy survive contact with a busy Saturday. Every unit gets exactly one source, decided by where the car actually came from rather than who gets credit. And the list is fixed, so nobody invents a ninth category in the notes field. Loose labels are the reason most attribution reports are unreadable a year later, exactly as they are with marketing sources.

Knowing that thirty percent of your units came from auction tells you nothing you can act on. The channel only becomes a decision when these travel with it.

All in acquisition cost. Purchase price plus fees plus transport. Auction units carry costs a trade does not, and comparing sticker to sticker flatters auction badly.

Reconditioning spend. The number that most often reverses a ranking. A channel that looks cheap at purchase and averages an extra fifteen hundred in recon is not cheap.

Days to front line. Time from acquisition to being retail ready. A unit sitting in recon for eleven days is not inventory, it is a bill.

Days to sale and turn rate once it hits the line.

Front and back gross at delivery.

Together those give you a return per unit by channel, and that is the number worth building a sourcing plan around. It is also the number that changes when the market changes, which is why it has to be a standing report rather than an annual exercise. There is a calculator angle to this at acquisition ROI.

Stores that have tried this and given up usually hit one of these.

The tag is entered by someone who was not there. A clerk stocking in a unit three days later guesses the source. Guessed data is worse than missing data because it looks complete. Whoever bought the car should record where it came from, at the time, on their phone.

The field lives somewhere nothing reports on. A note in a comment box is not attribution. It has to be a structured field that a report can group by, which is the whole difference between having the information and being able to use it.

The buckets shift. Someone renames a category, another store in the group uses different words, and now two quarters cannot be compared. Freeze the list and make renaming a deliberate decision with a date attached.

The fix in every case is the same: capture at the moment of acquisition, into a structured field, using a fixed list. That is a discipline problem with a small software component, not the other way around.

Being direct about the boundary, because inventory attribution touches systems we deliberately do not sell.

We do not provide inventory accounting, floorplan or vehicle costing, general ledger entries, deal posting to accounting, or title and registration work. Recon spend and stock in accounting live in the dealer management system you run and in your recon tool. If you want a single system that books the cost, that is not what we are.

What we do provide is the acquisition side and the reporting around it. Inventory Link ingests your live inventory feed with VIN, trim, mileage, price and photos, so the vehicles being advertised are your actual vehicles rather than a stale list. Every lead carries its source, through lead feeds and custom source feeds, so the demand side of a unit is traceable. Campaign reporting shows which acquisition campaigns produced offer requests. And the customer profile carries multiple vehicle slots, including trade in rows alongside the vehicles a customer is shopping, so an appraisal conversation and a purchase conversation stay attached to one person.

Neither a DMS integration nor an inventory feed is required to run the platform, which is why brokers and single point stores use it. But if you have a feed, the attribution picture gets considerably better. More on that at inventory feed management.

Buying directly from local owners is the channel with the most upside and the least measurement, because it is the newest habit in most stores and because the pipeline is longer than a trade.

Our seller leads are inbound and opt in. Local owners fill out a vehicle offer request and ask to be contacted by a dealership, and every submitted request inside your mapped zone comes to you exclusively rather than being resold. Nothing is filtered or scored. Because those arrive as leads with a source attached, the whole funnel is measurable in a way an auction run list never is.

The chain worth tracking is offer requests received, appraisals completed, offers made, vehicles bought, and then the retail outcome of those units. Each step tells you something different. A store buying a low share of the vehicles it appraises has an offer problem. A store appraising a low share of the requests it receives has a follow up problem, and follow up processes with texting and voicemail drop fix that faster than any change to the offer.

Illustrative math, not a promise: if a store completes appraisals on 40 of 100 offer requests and buys 10 of those, the cost of the program divided by 10 is the acquisition cost per unit for the channel, before recon. Run the same arithmetic against your auction fees and transport for an honest comparison. Your numbers will differ, and we cannot guarantee volume or purchase rates. See vehicle acquisition software for how the pipeline runs.

Three cautions, all learned by people who made confident decisions on thin data.

Small samples lie. A channel that produced nine units last quarter cannot be ranked against one that produced ninety. Set a minimum unit count before you let a channel drive a decision, and be suspicious of any ranking that flips month to month.

Mix beats channel more often than people think. If your consumer acquisition units skew toward five year old domestic trucks and your auction units skew toward compact sedans, you are comparing segments, not sources. Compare inside a segment before you conclude anything about the channel.

Seasonality is real. Tax season, model year changeovers and weather move both sides of this equation. Compare against the same period last year rather than against last month whenever you have the history to do it.

Turn and days supply are the metrics that keep the analysis honest over time, and they belong on a standing dashboard rather than in a quarterly slide. There is more at inventory turn dashboard and on stocking strategy.

You do not need a project plan. You need one field and one meeting.

Week one, agree the channel list and write it on a whiteboard in the used car office. Week two, require the buyer to record the source at acquisition, no exceptions, and check compliance daily until it sticks. Week three, backfill the last ninety days from memory and paperwork, accepting that the backfill is rougher than the new data and labeling it that way. Week four, pull the first report, look at cost and days to sale by channel, and pick one channel to push and one to reduce.

Then repeat monthly. The first report will be imperfect and still better than what you have now. The second one will be the one you start trusting.

On our side, the CRM including campaign and source reporting starts at $199 a month on CRM Only. The seller side, which is the piece most stores are missing, runs on Marketplace Acquisitions from $999 a month, and the combined buyer and seller plan is $1,599. Everything is month to month with no long term contract, and territory is mapped around your store. Current figures are on the pricing page, and contact us if you want to walk your channel mix with someone before deciding anything.

Does LeadLocate track reconditioning cost per unit?

No. Recon spend, vehicle costing and floorplan accounting live in the dealer management system you run and in your recon tool. We cover the acquisition and demand side: sourced leads with attribution, campaign reporting, appraisal conversations and inventory advertising through Inventory Link.

How do seller leads get attributed to a specific unit?

Every offer request arrives as a lead with its source attached and sits on a customer profile with vehicle detail. When that vehicle is purchased, the record ties the acquisition back to the campaign and zone that produced it, which is what makes cost per acquired unit calculable.

What channel list should we use?

Six to eight fixed categories works for most stores: retail trade in, consumer acquisition, service drive, physical auction, online auction, wholesale or dealer trade, fleet and lease return, and street purchase. Keep it fixed, because renaming buckets destroys year over year comparison.

Do we need an inventory feed for this to work?

Not to run the platform. An inventory feed is not required. If you do have one, Inventory Link ingests it with VIN, trim, mileage, price and photos, which makes the connection between a sourced unit and its advertising far cleaner.

How long before the report is trustworthy?

Usually two to three months of disciplined capture. Backfilled history is rougher than data captured at the moment of purchase, so label it as such rather than mixing the two silently. Set a minimum unit count per channel before letting the report drive a decision.

Will attribution improve our gross?

We cannot guarantee that, and no vendor can, because gross depends on your market, your recon and your pricing. What attribution reliably does is stop you from expanding a channel that looks cheap at purchase and is expensive by the time it hits the front line.

Book a walkthrough and we will map an acquisition zone around your store, show the seller lead pipeline end to end, and set up the source reporting behind it. Month to month, no long term contract.