Skip to content

Dealership Margins Fell in 2026. Here Is Why Data Is the Margin You Have Left

By DealerApex TeamSeptember 12, 202614 min read
Dealership Margins Fell in 2026. Here Is Why Data Is the Margin You Have Left

New-car gross is down about a third. Cars now take 60 days to turn. Financing is expensive and demand is soft. The dealers pulling ahead are not the ones with the best inventory. They are the ones with the best data.

By DealerApex Team   September 12, 2026  9 min read

A joint report from DealerApex and DemandNow.ai.

Key Takeaways

  • Front-end gross on a new car has fallen roughly 30 percent, from a peak near $3,250 to $4,500 down to about $2,247 per unit.
  • New vehicle supply jumped 48 percent to 2.8 million units, and the average car now takes 60 days to turn instead of 30.
  • In a 14-month analysis of one anonymized two-rooftop group, the dealer’s own website converted leads to sales at 11 to 13 percent while several marquee third-party listing products converted at close to zero.
  • Most stores still cannot join CRM leads to DMS deals, so they cannot calculate a true cost per sale on their single biggest controllable expense.
  • Two cheap wins are sitting untouched: replying to internet leads inside 60 seconds, and filling a service drive that now carries roughly 64 percent of store overhead.

For three years, cars sold themselves. Short supply, record grosses, buyers paying over sticker. That market is over. Plants are back to full output, lots are full again, and the buyer is back in control. If you run a store or a small group, you have already felt it on the monthly statement.

This report is a partnership between DealerApex and DemandNow.ai. DealerApex builds dealer websites and the data platform behind them. DemandNow.ai runs the marketing and the analytics. We combined our industry research with anonymized data from real dealer clients and wrote this for one reason: in 2026, the margin the market used to hand you is gone, but the margin sitting inside your own data is still there for the taking.

The margin the market handed you is gone

The numbers are not subtle. Front-end gross on a new car has fallen roughly 30 percent, from a peak near $3,250 to $4,500 down to about $2,247 per unit. Used front-end gross is down close to 16 percent, to around $1,399. Average pre-tax profit per franchised store dropped from $1.80 million to $1.04 million, a fall of 24 to 43 percent depending on the study. Return on sales slid from 3.4 percent back to 2.1 percent, right where it sat before the pandemic.

Franchised dealership profitability, peak market versus 2026. Figures in US dollars per unit or per store, compiled by DealerApex from published industry studies, September 2026. Dashes mark metrics where the source reports a change rather than a prior level.
Metric Peak market 2026 Change
New front-end gross per unit $3,250 to $4,500 $2,247 Down about 30%
Used front-end gross per unit $1,399 Down about 16%
Average pre-tax profit per franchised store $1.80M $1.04M Down 24% to 43%
Return on sales 3.4% 2.1% Back to pre-pandemic
New vehicle supply 2.8M units Up 48%
Days to turn a vehicle 30 days (2023) 60 days Doubled

Pricing power went with it. About half as many dealers can charge over MSRP today as could two years ago. New transaction prices have leveled off near $47,278. Used prices have slipped to about $25,981 and are still drifting down. The lever you used to pull, price, barely moves anymore.

Money got expensive and demand went soft

Interest rates are the anchor. Prime auto loans run 6.5 to 9.5 percent. Non-prime runs 14 to 18 percent. More than half of dealers say high rates are hurting showroom traffic, lead conversion, and finance approval. The buyer who qualified last year now does not, or walks the moment they see the monthly payment.

At the same time, inventory came flooding back. New vehicle supply jumped 48 percent to 2.8 million units. The average car now takes 60 days to turn, double the 30 days it took in 2023. Every extra week a unit sits is floor-plan interest eating your gross before the car is even sold. More than a third of dealers now name floor-plan cost as a real drag on earnings.

The squeeze

Lower grosses, softer demand, and higher carrying cost, all at once. You cannot price your way out of it. You have to operate your way out of it. That is where data stops being a nice-to-have and becomes the job.

Data is the new margin in a thin-gross market

When margin is thin, precision pays. There are four places where data now decides the outcome.

Buy right, because stocking the wrong unit is no longer a small mistake

With supply loose again, you can afford to be selective. Real data on what turns in your market, in your price band, keeps you from buying metal that will age on the lot and bleed carrying cost.

Price to your turn data, not to the top of a search result

Listing sites rank by price, so it is tempting to list below market to win the top spot. That spot brings bargain hunters who grind your gross and fill your CRM with junk leads. Pricing to your actual turn data protects front-end and back-end gross together.

Move fast, because velocity is now a line on your P&L

Live days-supply and aging data tell you which unit to promote, mark down, or wholesale this week, before the carrying cost passes the gross you were trying to hold.

Measure marketing ROI live, not in a quarterly recap

This is the one most stores still cannot do. Your CRM knows the lead. Your DMS knows the deal. Nothing connects them, so you cannot see which ad actually sold a car. You end up flying blind on your single biggest controllable expense.

Stop renting all of your traffic from third-party listing sites

Dealers now spend $500 to $700 in marketing for every car they retail. Paid-search leads cost $45 to $65 each and keep climbing. For years, 15 to 20 percent of that budget went straight to third-party listing sites like CarGurus and Cars.com.

Those platforms have a place. But leaning on them alone is a risk, and our own client data shows why. DemandNow analyzed 14 months of CRM data for an anonymized two-rooftop independent used-car group in the Northeast, covering more than 25,000 leads and over 880 car sales at a single rooftop. The pattern was hard to miss: the channels the group owned crushed the channels it rented.

The group’s own website converted leads to sales at 11 to 13 percent. Referrals converted in a 20 to 37 percent range. Meanwhile, several marquee paid products from the big listing sites barely registered. A CarGurus re-engagement product converted at 0.22 percent. A Cars.com alert product converted at 0 percent. One automated “buying signals” lead product delivered nearly 1,200 leads and zero sales. A paid social-marketplace channel ran at a net loss over the period.

Lead-to-sale conversion by source. DemandNow analysis of 14 months of anonymized CRM data from one two-rooftop independent used-car group in the Northeast, more than 25,000 leads and over 880 sales at a single rooftop, to September 2026. Conversion equals confirmed sales divided by leads. Spend-based figures are modeled from historical channel efficiency, not invoiced. Illustrative of one group, not an industry benchmark.
Lead source Type Lead-to-sale conversion
The dealer’s own website Owned 11% to 13%
Referrals Owned 20% to 37%
Lender-driven channel (phone) Paid About 13%
CarGurus, main listings Third-party listing About 1% to 3%
CarGurus re-engagement Third-party listing 0.22%
Cars.com alert product Third-party listing 0%
Automated “buying signals” product Third-party 0 sales on about 1,200 leads

The listing sites were not all bad. CarGurus drove real volume, close to 400 cars over the 14 months. But its efficiency was eroding, and at one rooftop it converted at just 1.4 percent. The single best paid source was not a classified site at all. It was a lender-driven channel that carried the lowest acquisition cost of any source and converted phone leads at around 13 percent.

Then came the deeper problem, and most stores share it. The group’s ad spend lived in platform invoices. Its sales lived in the CRM. Until the two systems were joined, no one could calculate a true cost per sale. When the numbers were finally modeled together, blended return on ad spend came in well under the 2x to 4x a healthy channel mix should produce, and the portfolio bad-lead rate ran near 18 percent. Money left the building every month with no live view of what it bought.

The fix is not to quit the listing sites. It is to stop being a pure renter. Build owned channels: your own website, your own search presence, your own customer database. When you own the channel, you control the traffic, and you can finally measure every dollar against a sold car instead of a lead. That is the same argument we made in the case for treating your dealer website as a storefront rather than a brochure.

Two fast wins most stores leave on the table

Speed to lead inside 60 seconds

More than 30 percent of internet leads get no timely reply, and the industry average response time runs into hours. Dealers who answer and qualify a lead inside 60 seconds book about 27 percent more appointments and close about 26 percent more of their leads. It costs almost nothing, and almost no one does it.

Fixed ops, the most reliable gross in the building

Service and parts now cover roughly 64 percent of the average store’s total overhead, and the fleet on the road is a record 12.6 years old. Every service call that rings out is customer-pay gross walking to the shop down the street. Answering the phone and filling the service drive is the most reliable money in the building right now.

The AI adoption gap is the opportunity

Most dealers know AI is coming. About 43 percent already use some tool, and another 47 percent plan to. But most of that is a website chatbot skimming the top of the funnel. The high-return work, instant lead response, live pricing, and true marketing attribution, is still rare across the industry. That gap is exactly where an early mover pulls ahead, and it will not stay open forever.

Why DealerApex and DemandNow wrote this together

DealerApex gives you the website and the platform that tie your marketing back to the car it actually sold. DemandNow.ai runs the analytics and the owned-channel marketing on top of it. One shows you the truth in your numbers. The other acts on it.

The margin the market handed you is gone. The margin in your own data is still sitting there, waiting. In 2026, the dealers who go get it are the ones who will still be standing in 2030.

A storefront still needs a scoreboard. See everything in the platform, or get a free read on where your own store’s numbers stand — we will join your CRM leads to your DMS deals and send back cost per sale by channel, plus the full data set behind this report.

Frequently asked questions about dealership margins in 2026

Why are car dealership margins falling in 2026?

Dealership margins are falling because supply came back while borrowing costs stayed high. New vehicle inventory rose 48 percent to 2.8 million units, so dealers lost the pricing power they held during the shortage, and prime auto loan rates of 6.5 to 9.5 percent cooled demand at the same time. Front-end gross on a new car fell about 30 percent as a result, to roughly $2,247 per unit.

Should dealers stop advertising on third-party listing sites?

No, but no store should rely on them alone. Track every channel against sold cars rather than leads. In one anonymized two-rooftop group DemandNow analyzed over 14 months, the dealer’s own website converted leads to sales at 11 to 13 percent while some big-name listing products converted at close to zero. Owned channels give you control of the traffic and a clean line from spend to a sold unit.

What is the fastest way for a dealer to protect gross right now?

The two fastest levers are speed to lead and fixed ops. Answering internet leads within 60 seconds lifts appointments by about 27 percent and closings by about 26 percent, and filling the service drive captures the most reliable gross in the store, since service and parts now cover roughly 64 percent of total overhead.

Client figures in this report come from DemandNow analysis of 14 months of anonymized dealer CRM data for a single two-rooftop independent used-car group in the Northeast. They are illustrative of one group’s channel mix, not an industry benchmark. Industry figures are compiled by DealerApex from published studies current to September 2026. Want the full data set, or a straight read on where your own store’s numbers stand? Reach us at dealerapex.ai and demandnow.ai.