The baseline is not zero, and that changes the whole calculation.
A dealership evaluating an SEO proposal is almost never choosing between spending and not spending. It is choosing between two ways of spending.
NADA Data 2025 puts total advertising at $586,246 per franchised dealership for the year, which works out to $739 per new vehicle sold. Inside that figure, three lines account for 60.6% of everything: search engine marketing at $123,698, third-party listing sites at $117,249, and SEO and website optimisation at $114,318. Dealership advertising is now mostly a search budget with some broadcast attached.
That third line is the one this page is about, and it is the one stores are most surprised by. At $114,318 a year it runs to roughly $9,527 a month. The reason it surprises people is structural: the SEO line is usually bundled inside a website or platform contract rather than invoiced on its own, so no single document at the store ever shows it as a standalone number. Nobody signs off on $9,527 a month. They sign off on a platform.
So when a general manager asks whether SEO is worth it, the question being answered is usually the wrong one. The comparison that matters is not a proposed retainer against zero. It is a proposed retainer against a line already running at eight times its size, whose output nobody has ever been asked to defend. Where the rest of the budget goes →
- Search engine marketing
- $123,698 a year, 21.1% of the budget. Stops producing the day it stops being funded.
- Third-party listing sites
- $117,249, 20.0%. Rents an audience on a domain the store does not own.
- SEO and website optimisation
- $114,318, 19.5%. Usually invisible, because it sits inside the platform contract.
- Everything else
- $230,981, 39.4%. Broadcast, print, direct mail, events and the rest.
Three search lines, 60.6% of the budget. Only one of the three compounds, and it is the one nobody sees an invoice for.
The formula, and the two inputs that make most dealer numbers wrong.
The arithmetic itself is not complicated. Incremental gross attributable to organic, minus the cost of the line, divided by the cost of the line. Almost every disagreement about automotive SEO ROI is a disagreement about what belongs in the numerator.
- Branded organic does not belong in it. A buyer who types the store's name has already chosen the store. That traffic converts at rates no non-branded query approaches, and it would have converted without the program. Including it is the most common way a vendor report overstates a return, and it is usually not deliberate: the default report in most analytics tools aggregates all organic into one line. Branded volume is still worth watching, because it rises when the rest of the marketing works, but it is a demand signal rather than a return.
- Last-click attribution does not measure this channel. A buyer researches a model in week one, compares trims in week three, checks a local store's reviews in week five, then clicks a paid ad and submits a lead. Last-click hands the entire gross to the ad. The organic work that put the store in the consideration set gets nothing. This is why SEO reliably looks worse than paid in a dealership's own reporting, and why a store that cuts the line often sees paid performance degrade a quarter later without connecting the two.
- The denominator is bigger than the retainer. The full cost of the line includes what is already bundled into the platform contract, any content quota being paid for, and the staff time spent managing vendors. A comparison that puts a new retainer in the denominator while leaving the existing spend out is not measuring the same thing on both sides.
- Repair orders belong in the numerator. Service demand is the most local and most recurring demand a store has, and fixed operations is where a search program frequently pays first, because the queries are specific and the competition is independent shops rather than franchise stores. Leaving service out understates the return on the work that produced it. Service and parts SEO →
None of these four corrections requires a new tool. All of them require the report to separate branded from non-branded and sales from service, which most standard dealership reporting does not do by default. That separation is the practical starting point for any honest ROI conversation, and it is the first thing worth asking a current vendor to produce. The metrics that predict sales →
Four numbers a store already has, and the break-even they produce.
A dealership can run this calculation before any vendor is involved, using figures already sitting in the DMS and the CRM. No projection is required, and no vendor model needs to be trusted.
Front and back together, averaged over the last twelve months. This is the multiplier on every incremental unit, and it is the number stores most often quote too low by leaving finance and insurance out.
Submitted leads that became delivered units. Most stores know this figure and dislike it. It converts visibility into vehicles and is the input a vendor cannot supply.
Leads and calls from organic sessions that did not begin with the store's own name. Search Console splits this cleanly, and the split is usually more unbalanced than expected.
The retainer, plus whatever the platform contract already allocates to SEO and content. Ask for that allocation in writing. A vendor unwilling to state it has answered the question.
The top row is the point. At a published $1,165 per rooftop per month, break-even sits under a single vehicle a month at every realistic gross level, which means the retainer decision at this price is rarely a return question at all. It is a trust question about whether the work will happen. The bottom row is the uncomfortable one: the spend already running needs between two and five incremental units a month to justify itself, and most stores have never been shown whether it produces them.
Put your own gross, close rate and lead count into the same arithmetic.
Open the SEO ROI calculatorThe payback curve is not a straight line, and the first quarter measures the lag.
Paid media and earned search have opposite shapes, and judging one on the other's timeline is how good programs get cancelled.
A media buy produces on day one and stops producing the day the card is declined. The return is immediate, legible, and entirely rented. Earned search produces nothing measurable for weeks, then produces at a rate that keeps rising after the spend flattens, because indexed pages, resolved entities and accumulated corroboration do not switch off at the end of a billing cycle.
This has a practical consequence for how a store reviews the line. In months one to three, the honest metrics are indexation and impression share on non-branded commercial queries. Rankings on the classes that produce leads move in months four to six. Units and repair orders attributable to the work land after that. A ninety-day review that looks for revenue is measuring the lag and calling it a failure.
The same asymmetry explains a pattern stores see and rarely diagnose. A dealership cuts the search line to fund something urgent, sees no immediate drop because the pages keep working, and concludes the line was never doing anything. The decline arrives two quarters later, by which point it gets attributed to the market. What the work actually costs →
- Months 1 to 3
- Indexation of inventory, model and service pages. Non-branded impression share. Technical faults closed.
- Months 4 to 6
- Position on the query classes that produce leads. Citation rate in AI answers. Map pack presence.
- Months 7 to 12
- Leads, calls and repair orders attributable to non-branded organic. First honest ROI figure.
- Month 13 onward
- Compounding, or the absence of it. By now a flat line is a real finding rather than a lag.
A vendor promoting revenue in quarter one is describing paid media and pricing it as SEO.
Local return and national return are different investments.
Local
Demand is bounded by a drive radius, the competitive set is a handful of named stores, and payback is the fastest available in automotive because the queries carry immediate intent. The limit is arithmetic: a market only contains so many buyers, and once a store holds the map pack and the commercial classes, additional spend buys less each month.
For a single rooftop selling vehicles and service inside its radius, this is almost always the better return, and it is where a first engagement should start. The Columbus market read is a worked example of what that looks like against a real DMA. The Columbus read →
National
Slower to earn, larger when it lands, and only justified when the business genuinely sells beyond the radius. Three cases qualify: a dealer group consolidating group-level terms across rooftops, a parts or accessories operation shipping nationwide, and specialist inventory such as classic, exotic or commercial vehicles that buyers will travel for.
The failure mode is a single rooftop buying national work because the volume figures look bigger. Those buyers cannot take delivery, so the traffic converts at a fraction of local traffic and the return calculation worsens without anyone connecting the two. When group-level work pays →
The sequencing that holds for most operators is local first until the map pack and the commercial query classes are secure, then group or national work funded out of what local produced. Running both from a standing start splits a budget that was not large enough to win either.
Measuring the return on the agency relationship, separately from the program.
Program return and vendor return are two different questions, and a store can be getting one without the other. A rising traffic chart is compatible with a vendor doing almost nothing, because some of the rise is seasonal and some of it is work done two vendors ago.
Ask what shipped this month and to be shown it on the live site. A vendor describing activity rather than artifacts is billing for meetings.
Reporting that separates branded from non-branded and sales from service. Aggregate sessions is the metric that hides everything worth knowing.
A vendor that never reports something that did not work is either not testing or not telling. Both are worth knowing before renewal.
Whether the work would still exist if the vendor left. Pages on the store's own domain survive. Content inside a rented widget leaves with the contract.
Those four are deliberately checkable by someone who does not do SEO for a living, which is the point of them. The longer version, including the red flags that show up in a sales conversation before any money changes hands, is set out in the evaluation guide, and the vendor field itself is mapped separately. The evaluation guide → The vendor roster →
When the ROI case does not hold.
Four conditions where the arithmetic on this page produces a number a store should not act on. Each is worth saying before an engagement rather than after one.
- The store needs units this quarter. Earned search compounds rather than switching on, and no amount of urgency changes the curve. That requirement is a paid media job, and any vendor who will not say so is optimising for the retainer.
- Nobody works the lead queue. Additional demand into a queue that sits produces additional unworked leads and a worse conversion rate. Response time is not an SEO deliverable, and fixing it costs nothing.
- The market is genuinely thin. Some rural single rooftops earn more from their service drive and their reputation than from an organic program. Bounded demand cuts both ways, and a market with few buyers has a low limit no matter who does the work.
- A replatform is already scheduled. Building on a site that will be replaced in two quarters spends the budget twice. Migrate first, then start the program on the site that will still exist. What each platform constrains →
Questions dealers ask about SEO return.
How do you calculate SEO ROI for a car dealership?
Units and repair orders attributable to non-branded organic, multiplied by the store's own average combined gross, minus the full cost of the search line, divided by that cost. Two disciplines make it honest. Exclude branded organic, because a buyer typing the store's name would have arrived anyway. And count the full cost of the line rather than only a new retainer, because the average franchised dealership is already funding SEO inside a platform contract.
How much does the average dealership already spend on SEO?
$114,318 a year, per NADA Data 2025, which is 19.5% of the average franchised store's advertising budget and the third-largest line in it, behind search engine marketing at $123,698 and third-party listing sites at $117,249. That is roughly $9,527 a month. Most stores have never seen the figure on its own, because it is bundled into a website or platform contract rather than invoiced separately.
What is a good ROI for automotive SEO?
There is no benchmark worth quoting, because the denominator changes with what a store already pays and the numerator changes with its gross per unit. The useful test is break-even expressed in units. At $1,165 per rooftop per month and $3,000 combined gross, the program covers itself at roughly 0.39 incremental units a month, about one additional vehicle a quarter. A store can check that against its own DMS before signing anything.
Should branded search count toward dealership SEO ROI?
No. A buyer searching the store's name has already chosen the store, and that traffic converts at rates no non-branded query approaches. Counting it inflates every figure downstream and is the most common way a vendor report overstates a return. Branded volume is worth monitoring as a demand signal, because it rises when the rest of the marketing works, but it belongs outside the calculation for the search line.
How long before dealership SEO pays for itself?
The first ninety days measure the lag rather than the return. Indexation and non-branded impression share move in months one to three. Position on the query classes that produce leads follows in months four to six. Units and repair orders land after that. A store judging the program on quarter one is measuring the wrong thing, and a vendor promoting quarter-one revenue is describing paid media.
What is the break-even on a $1,165 per month SEO retainer?
0.39 units a month at $3,000 combined gross, 0.29 at $4,000, and 0.58 at $2,000. Every one of those is under a single vehicle a month, which is why the retainer decision at this price is usually a trust question rather than a return question. The harder arithmetic is against the line already funded: at $9,527 a month and $3,000 gross, that spend needs 3.18 incremental units a month to break even.
How do you measure ROI when an AI answer resolves the question without a click?
By measuring citation rather than sessions. When an assistant names the store and the buyer then types the name into a browser, analytics credits a direct or branded visit and gives the search line nothing. The reportable proxies are the share of category questions in the market where an engine names the store, movement in branded volume with no matching media spend, and call and direction actions on the business profile.
Is local SEO or national SEO the better return for a dealership?
For a single rooftop selling inside a drive radius, local is almost always better: bounded demand, faster payback, and a competitive set of a few named stores. National work earns its place when the business sells beyond the radius, which usually means a dealer group consolidating group terms, a parts operation shipping nationwide, or specialist inventory buyers will travel for.
How do you measure the return on an SEO agency relationship?
Separately from program return, on four things a store can verify without trusting a dashboard. Whether named deliverables shipped and can be pointed at on the live site. Whether reporting separates non-branded from branded. Whether the vendor states plainly what did not work. And whether the work would survive the vendor leaving. A relationship failing all four can still show a rising traffic chart. The evaluation guide →
When does the ROI case for dealership SEO not hold?
Four situations. The store needs units this quarter, which is a paid media job. Nobody works the lead queue, so more demand produces more unworked leads. The market is genuinely thin, where a rural rooftop may earn more from its service drive. And a replatform is already scheduled, where the sensible order is to migrate first and start on the site that will still exist.
Does VulcanAX report ROI?
Reporting covers non-branded impression share, position by query class, citation rate in AI answers, and the leads and calls a store can match back. The final multiplication into gross stays with the store, because only the store knows its real close rate and real gross per unit, and a vendor supplying both halves of an ROI calculation is grading its own work. Pricing is published, so the denominator is never in question. Full pricing →
Find out what the line you already pay for is producing.
The baseline audit reports non-branded visibility, citation rate in AI answers, and which of your pages are indexed at all. Most stores discover the expensive part of their search budget is the part nobody has ever been asked to defend.