Beyond Impressions and Clicks: The Three Classes of Automotive Attribution

Beyond Impressions and Clicks: The Three Classes of Automotive Attribution 

Most agencies, advertising platforms, and media vendors say they provide attribution. The problem is that the word “attribution” has become so broad that two reports can use the same label while answering completely different questions.

One report may show impressions, clicks, and video completions. Another may connect an ad exposure to a later website visit. A third may compare campaign exposure with actual vehicle sales. All three are useful, but they do not provide the same level of insight.

After 21 years in automotive digital marketing, I have found it helpful to place attribution reporting into three classes. The framework is especially relevant for upper funnel programmatic media like; OTT, CTV, online video, streaming audio, and display. These channels often influence shoppers without generating an immediate click or, sometimes, even a single click, so evaluating them with traditional click based metrics alone creates a very incomplete picture.

The purpose of this framework is not to dismiss basic campaign reporting. Dealers need delivery metrics, engagement signals, and outcome measurement. The point is to understand what each class can prove, what it cannot prove, and how it should influence a dealership’s marketing decisions.

Why Attribution Gets Confusing

Attribution is supposed to help answer a basic question: What role did our marketing play in producing a business result? That question becomes complicated very quickly. A vehicle shopper may see a streaming CTV ad, hear a streaming audio spot, search for the dealership later, visit several VDPs, return through a direct visit, submit a form, and ultimately purchase at the store. Depending on the reporting system, paid search might receive the credit because it generated the last click. The CTV campaign might receive credit because the shopper saw an ad before visiting. Or both might receive credit for the same sale (hint; both should receive the credit).

That does not automatically mean either report is wrong. It means each report is viewing the customer journey through a different measurement window and a different set of rules.

Dealers should therefore begin every attribution conversation by asking what is actually being connected. Is the report measuring media delivery? Is it connecting exposure to digital behavior? Or is it matching ad exposure to a completed sales outcome?

Those questions separate the three classes.

Class 3: Delivery Metrics

Class 3 is the foundation of campaign reporting. It usually includes metrics such as:

  • Impressions
  • Clicks and CTR
  • Video and Audio completions
  • Reach and Frequency
  • CPM/eCPM
  • GA4 site activity

These metrics tell a dealer whether the campaign delivered as planned. They help answer operational questions. Did the campaign spend its budget? Did it reach enough people? Were viewers completing the video? Was the frequency too high? Did one creative generate more clicks than another?

That information is necessary, it’s basic campaign delivery reporting. The limitation is that delivery is not the same as attribution.

An impression confirms that an ad was served under the platform’s measurement rules. A completed video confirms that the ad played to completion. A click confirms an immediate interaction. None of those actions, by itself, confirms that the person visited the dealership’s site later, entered the showroom, or purchased a vehicle.

This limitation matters most where a click is not the natural response. A CTV viewer or streaming audio listener may remember the message but act hours, days, or weeks later on another device. 

Class 3 reporting should be viewed as campaign health reporting. It can show whether the media was delivered efficiently and whether people engaged with the format. It cannot fully explain the business effect of that media.

Class 2: Influenced Traffic Attribution

Class 2 begins connecting advertising exposure to later consumer behavior. In addition to delivery metrics, it may include:

  • Influenced or attributed site traffic
  • Influenced or attributed site activity 
  • Specific pages viewed after exposure (VDPs, SRPs, Service, etc.)
  • Site conversions after exposure (chats, calls, from submissions, etc.)
  • The time between exposure and the website visit

This is where attribution becomes much more useful for upper funnel media.

Imagine that a shopper sees a CTV ad on Tuesday evening. On Thursday morning, that shopper searches for the dealership, enters through an organic listing, and views three VDPs. Standard web analytics may credit the visit to organic search because that was the immediate source of the session. Class 2 reporting can add important context by showing that the website activity followed an advertising exposure.

That context helps dealers understand how channels work together. CTV may create awareness, search may capture the resulting demand, the website may move the shopper into consideration. The final conversion may occur through a phone call, chat, form, or showroom visit.

The strongest Class 2 reporting also shows the traffic source that ultimately brought the shopper to the website and the amount of time that passed between ad exposure and action or website visit. These details make the report more than a total count. They begin to reveal the sequence of the shopping journey.

Class 2 still has limitations. A website visit is not a vehicle sale. A shopper can view several vehicles and never contact the store. Another shopper can purchase after conducting very little measurable website activity. Shared devices, privacy settings, consent choices, browser restrictions, and identity matching limitations can also affect what can be observed.

For those reasons, influenced traffic should not be presented as proof that an ad caused a sale. It is evidence that advertising exposure and meaningful shopping behavior occurred within a specific measurement window. That is useful, but it remains one step removed from the final business outcome… the SALE.

Class 1: Sales Matchback Attribution

Class 1 connects media exposure to completed vehicle sales. Depending on the available data, permissions, and measurement methodology, reporting may include:

  • The purchasing customer record
  • The VIN purchased
  • The ads or creative viewed before purchase
  • The sequence of media exposures
  • The publishers or platforms involved
  • The time from first exposure to sale
  • The time from most recent exposure to sale
  • The number of sold customers who were exposed to specific campaigns

This is the most outcome focused class because it connects marketing activity with the vehicle sale, down to the individual customer who purchased a specific VIN. 

Instead of ending the month with only impressions and clicks, a dealer might learn that 87 of 127 sold customers had been exposed to one or more measured campaigns before purchase. The dealership is then able to examine which campaigns reached those buyers, which creative they saw, what vehicles they purchased, and how long the measured path to sale took.

This lets the dealer move beyond asking whether a campaign delivered or generated traffic to asking which audiences, messages, formats, and exposure patterns were most common among buyers.

More importantly, this data can be used to improve future media strategy and audience targeting. By analyzing which audience segments, channels, publishers, creative messages, ad formats, frequency levels, and exposure sequences appeared most often among buyers, dealers can identify the combinations associated with stronger sales outcomes. Those insights can help refine audience segments, adjust media allocation, improve creative messaging, manage frequency, and reduce spending on tactics that are producing fewer measurable results. In this way, attribution becomes more than a reporting tool. It creates a continuous feedback loop between campaign delivery, buyer outcomes, and future campaign optimization. 

However, even Class 1 reporting needs to be interpreted carefully. A match establishes a measurable relationship between campaign exposure and a sale. It does not automatically prove that the advertisement caused the purchase. The shopper may have been influenced by pricing, inventory availability, a prior relationship with the dealership, an OEM offer, a recommendation, or other marketing that was not included in the report.

Sales matchback is strongest when it is used to understand contribution rather than claim exclusive credit. Automotive shopping is rarely driven by one exposure or one channel. A useful Class 1 report helps dealers see more of the path, compare patterns, and make better decisions without pretending that a complex purchase has a single cause.

Attribution Is Not the Same as Causation

This distinction deserves its own section because it is one of the most common sources of confusion whenit comes to reporting automotive marketing results.

Attribution assigns credit according to a defined methodology. Causation asks whether the outcome would have happened without the advertising. Those are not the same question.

Dealers can improve confidence by looking beyond raw match totals. Useful comparisons may include exposed buyers versus nonexposed buyers, sales rates across comparable audiences, geographic patterns, time to purchase, frequency ranges, creative differences, and changes observed when campaigns begin or end. Where scale and methodology allow, controlled testing can provide stronger evidence of incremental impact.

The best attribution reporting acknowledges uncertainty. A credible report should make its matching rules, lookback window, data sources, and limitations understandable.

Questions Every Dealer Should Ask

Dealers do not need to become data scientists to evaluate attribution. They do need to ask direct questions.

What exactly is being attributed?

Is the report connecting an ad to a click, a website visit, a lead, a showroom visit, or a completed sale? The word attribution is not specific enough on its own.

How long is the lookback window?

A seven day window and a 30 day window can produce very different totals. The appropriate period may vary by channel, campaign objective, and typical purchase cycle. What matters is that the rule is disclosed and applied consistently.

How are people and devices matched?

Consumers move among televisions, phones, tablets, computers, and physical locations. Dealers should understand, at a practical level, how exposure on one device is connected to activity or a sale associated with another.

Are sales matched at the household or customer level?

A household match can be valuable, but it is not identical to an individual customer match. Reports should clearly label the level of precision rather than blending different match types together.

What data is missing?

No attribution system sees every customer interaction. Consent requirements, privacy choices, device limitations, incomplete customer records, offline behavior, and platform restrictions can create gaps. A transparent vendor should be able to explain those gaps.

How is customer privacy protected?

Outcome measurement should have appropriate permissions, contracts, security controls, retention policies, and procedures for consumer privacy requests. More detailed reporting is only valuable when the underlying data is handled responsibly.

How to Use the Three Classes Together

The three classes should not be treated as competing scoreboards. Each serves a different purpose.

Class 3 tells you whether the campaign was delivered properly. Class 2 tells you whether exposed shoppers demonstrated measurable interest. Class 1 tells you whether media exposure appeared in the paths of actual buyers.

A healthy review process moves through all three levels. First, confirm campaign health. Review spend, impressions, reach, frequency, completion rates, and any unusual delivery patterns. Second, review influenced behavior. Look at website visits, inventory engagement, conversion actions, traffic sources, and the delay between exposure and activity. Third, examine business outcomes. Review matched sales, VINs, exposure sequences, time to sale, and overlap among channels.

Finally, interpret the findings together. A campaign with strong delivery, but little downstream activity may have an audience, message, inventory, or measurement problem. A campaign with few clicks, but substantial, influenced traffic or sales matches may be doing exactly what an upper funnel channel is supposed to do. A campaign with many matches but excessive frequency may still have room to become more efficient. No single metric should make the entire decision.

Set the Right Minimum Standard

For click oriented media such as paid search, immediate response metrics remain important. For upper-funnel channels such as CTV, online video, streaming audio, and display, delivery metrics alone are no longer enough to provide a useful performance picture.

At a minimum, dealers should expect to understand whether exposed audiences later demonstrated meaningful shopping behavior. When reliable sales matchback is available, it can add another layer by connecting ad exposure to actual sales.

The goal is not to produce the largest possible attribution number. The goal is to produce a clearer, more honest view of how marketing contributed to the customer journey.

When dealers know which class of reporting they are looking at, they can ask better questions, compare partners more equitably, and make more confident budget decisions. Future campaigns will perform better when you understand the channels that contributed to a sale and match those patterns. You can also use the customer journey attribution data to help build and refine future target audiences for future campaigns. That is ultimately what attribution should do. It should turn campaign data into a better understanding of how people move from exposure to consideration to purchase.

Before reviewing your next media report, ask one simple question: Are we looking at delivery, influenced behavior, or actual sales matchback? The answer will tell you far more than the attribution label alone.

  1. Danny Benites
    Vendor

    This is one of the clearest breakdowns of attribution I’ve seen, and it’s overdue. Too many dealers sit through vendor reviews where impressions, influenced traffic, and sold-unit matchbacks all get called “attribution” as if they’re the same thing, and this framework gives them the language to push back and ask better questions. I especially appreciated the section on attribution versus causation, since that honesty is what builds real trust with a dealer. Great work, I’ll be sharing this one.

Table Of Contents