#The three tiers, and why they collide
Automotive advertising splits into three tiers: the manufacturer nationally, the regional dealer association, and your store. All three can be running paid search in your market at the same time.
This means your performance is partly determined by what is running above you, which you do not control and often cannot see. A quarter where tier one is heavy on a model changes your auction dynamics for that model regardless of what you do.
It also means some of what looks like your success is inherited. If national advertising drove someone to search the model, and you captured the click, the campaign report credits you for demand you did not create.
FIGURE
All three tiers can run paid search in your market at once, so part of your result is set above you.
#Branded search is the number that lies
Bidding on your own dealership name converts extremely well. It has to: the person searching already decided to find you. It will make any account look healthy.
The real question is incrementality: how many of those people would have reached you anyway through the organic result directly below the ad? Some would. Possibly most. That portion of the spend bought nothing.
There are defensible reasons to bid on your own name: competitors bidding on it, controlling the message, occupying the page. Just know which reason applies to you, and separate branded from non-branded in every report so growth in one is never mistaken for the other.
#What tier three is good at
Specific inventory intent. Someone searching a model, trim and year with local intent is close to a decision, and a store with that vehicle on the ground has something national advertising cannot offer.
Service and parts. Frequently the most under-invested and least competitive paid opportunity a dealership has, aimed at customers with immediate need and a local bottleneck.
Conquest in a defined radius, where you can articulate why someone should drive past a closer store. If you cannot answer that in one sentence, the campaign will not answer it either.
What it is bad at: broad model-awareness terms already saturated by tier one, and anything where you are the fourth store in the same auction with the same inventory and no differentiator.
#The landing page is usually the problem
Most underperforming dealership paid search is not a bidding problem. It is that the ad promises a specific vehicle and the click lands on a filtered inventory page, or worse, the homepage.
Every step between the promise and the thing costs conversions. If the ad says a specific trim in a specific color, the landing page should be that vehicle, and it should load quickly on a phone on a mediocre connection.
This is where paid and organic work stop being separate disciplines. The page speed, structure and clarity that make a page rank are the same properties that make paid traffic convert, which is why treating them as separate budgets with separate vendors usually produces two mediocre outcomes.
FIGURE
Most of it is not a bidding problem. The ad promises a specific vehicle and the click lands somewhere else.
#Reading the reports plainly
If a vendor resists, that is itself informative.
Check whether conversions are deduplicated against your other channels. The same shopper commonly appears in the paid report, the website provider's report and the CRM, and summing vendor reports can produce more leads than the store received.
Look at cost per acquired customer instead of cost per lead. A channel producing cheap leads that never close is more expensive than one producing fewer, better ones, and lead-level reporting hides that completely.
#Account structure that survives contact with inventory
Most dealership accounts are structured around campaign types instead of around the business, which is why they need constant rebuilding as stock changes. A structure organized by intent survives longer: branded, model-specific non-brand, service and parts, and conquest: each with its own budget and its own success measure.
Keeping service separate matters more than it looks. It has different seasonality, a different customer, a shorter decision window and a much lower cost per acquisition. Blended into a single account, it gets compared with sales campaigns and quietly defunded.
Whatever the structure, it must tolerate inventory turning over. Campaigns pinned to specific stock need a process for what happens when the vehicle sells. Otherwise you are paying for clicks to pages that no longer have the car, which is both wasted spend and a poor first impression.
FIGURE
Organized by intent, each part with its own budget and its own success measure, so it survives stock turning over.
TABLE
Where dealership paid search earns its budget, and where it buys demand you had
Split branded from non-branded before reading any report. Most disagreements about paid search are disagreements about this table.
| Campaign type | What it buys | Honest verdict | Watch for |
|---|---|---|---|
| Branded: your store name | People already coming to you | Cheap, converts, mostly incremental only against competitors bidding on you | Being credited for demand SEO created |
| Non-branded model intent | Shoppers deciding between vehicles | The strongest tier-three case | Landing on a filtered SRP with nothing to do |
| Service and repair intent | High-intent local ownership demand | Consistently under-funded | Capacity: do not book past the shop |
| Conquest on competitor names | Attention at their moment | Works in a tight radius, expensive outside one | Brand rules and OEM co-op terms |
| Broad inventory feeds | Volume | Depends entirely on feed hygiene | Sold units still advertising |
| Always-on display and remarketing | Impressions | Rarely the bottleneck | Attribution taking credit for everything |
Spend levels depend on market and inventory and are set per account rather than benchmarked here.
A blended cost-per-lead figure hides which of the five campaign types is actually earning it, a confusion worked through in the paid media and marketing insights.