What Does a Study of Dealer Ad Spend Actually Measure?
A Harvard student and a former GM ad art director are leading a research project on dealership advertising spend, as reported by Auto Remarketing on September 2, 2026. The study is the kind of academic exercise the industry needs more of: structured, cross-dealer, designed to surface what the money is actually doing versus what it is claimed to be doing. The findings, if published, will be cited at conferences. Some dealer groups will use them to rebalance their agency relationships.
The research is already surfacing uncomfortable figures.
“62% of ad spend ‘carried no defensible link to a sale or a dollar of profit,’” - Auto RemarketingThat number should land hard at any dealer group still writing agency checks based on impressions delivered and clicks purchased. It is a structural indictment of how most dealers have been trained to think about advertising: reach first, accountability somewhere in the footnotes.
But attribution studies have a blind spot baked into their methodology, and it is not one the researchers can fix by improving the data collection. The blind spot is organic content. Specifically: the topical authority a dealer built, or did not build, before any of the paid impressions in the study were served.
Why Do Channel Attribution Studies Miss the Content Layer Entirely?
Paid search studies measure what happens inside the ad auction. They capture spend, impressions, clicks, and platform-attributed conversions. Channel attribution research by design captures only the ad-platform layer: spend inputs and conversion outputs as the platforms themselves report them. The question those studies cannot answer is why two dealers running identical budgets on identical keywords in the same DMA see materially different cost-per-lead figures.
The variable that explains much of that gap is content authority. A dealer who has published 40 articles ranking for the model queries in their market, who is cited by Perplexity when a buyer asks which local Nissan dealer has the best CPO selection, who has structured content answering the exact questions buyers type into AI answer engines: that dealer is not entering the paid auction cold. Their brand is already warm in the buyer's mind before the ad fires. The paid click is confirmation, not introduction.
The dealer across town, running the same Google budget against a blank content slate, is buying the entire relationship through the auction. Every impression carries the full cost of awareness, consideration, and intent. The CPL figures look like they belong to the same study. The underlying economics are completely different.
How Does Topical Authority Change What Paid Search Actually Costs?
Search auctions price intent, not brand equity. The auction does not know, and does not care, whether the buyer already trusts the dealer who wins the click. Quality Score captures some of the landing-page signal, but it cannot measure whether a buyer has read six articles from that dealer this month and already shortlisted the store before opening Google.
The result is that the content layer operates entirely outside the auction but shapes who enters it ready to convert. A buyer who found the dealer through a blog post ranking for "best certified pre-owned Accord in [city]" may click a paid search ad three days later. The study will record a paid search conversion. The organic article that pre-sold the relationship will not appear in any attribution report. It cannot: there is no ad account to query.
This is not a measurement gap the researchers can close by adding more data sources. It is a structural feature of how channel attribution works. The organic content layer sits upstream of the paid layer, and no current attribution model maps the full journey from first organic impression to paid conversion cleanly. The best a study can do is measure what happened inside the ad accounts. What happened before the ad accounts fired is, by definition, outside the study.
AUTONOMi syncs organic search queries, clicks, and impressions nightly from Google Search Console, creating the organic half of the organic-versus-paid picture, but it captures only the Google half of a buyer's pre-click journey. The earned-authority layer, the citations in AI answer engines, the brand recall from a well-ranked article: those remain invisible to any attribution tool that starts counting at the ad impression.
Why Is the Dealer's Content Slate Invisible to Ad-Spend Researchers?
Attribution studies draw from ad platform APIs. They see Google Ads data, Meta data, and whatever cross-channel modeling the researchers layer on top. They do not draw from Google Search Console. They do not draw from Perplexity's citation index. They do not draw from whatever AI answer engine cited the dealer's inventory page this week when a buyer asked for recommendations.
There is no industry-standard data source that captures organic content authority at the dealer level and links it to downstream paid performance. No standard dealer marketing data feed links organic search authority to paid auction performance at the rooftop level. That means a dealer can have 60 published articles driving meaningful search presence, and that content investment will appear exactly nowhere in a study that measures ad spend efficiency. The study's denominator is ad spend. The numerator is attributed sales or leads. The content is neither: it lives upstream of both.
This produces a systematic distortion in what the research can conclude. A dealer with strong organic presence and a tightly managed paid budget will look like a highly efficient advertiser. A dealer with a weak content slate and a bloated paid spend will look like an inefficient one. The study will correctly identify the output difference. It will not have the inputs to explain the cause.
As the Auto Remarketing piece notes, researchers argue that every dollar recovered from untraced spend is working capital: redeployable toward inventory, floorplan interest, or the channels that demonstrably convert. That framing is correct as far as it goes. But the channels that demonstrably convert are demonstrably converting in part because of the content authority the study cannot see. Recovering spend without building the organic layer underneath is recycling the same structural dependency on paid acquisition that generated the waste in the first place.
What Happens to a Dealer's Paid Search CPL When AI Answer Engines Enter the Funnel?
The research timing matters. This study is being conducted as AI answer engines are inserting themselves between the buyer's intent and the search results page. AI answer engines including Perplexity, Google AI Overviews, and ChatGPT Search synthesize answers from indexed web content and cite sources the buyer can click through to. For dealers with structured, well-sourced content, that creates a new pre-click touchpoint the paid-search channel never had to compete with. For dealers without content, it creates a new way to be absent.
AI answer engines are already crawling dealer websites and blog content. The question is not whether they will influence dealer selection; it is whether any given dealer's content is structured to be cited or whether the dealer gets filtered out at the answer-engine layer before a buyer ever reaches a search results page. A dealer who loses the AI answer engine citation does not show up in an ad-spend attribution study as having lost that opportunity. The impression was never served. The study has no record of the absence.
This is the next generation of the same structural blind spot. Attribution studies will continue to measure what the ad accounts report. The content authority layer will continue to shape who enters those accounts ready to convert, and who is already gone. The gap between the two is widening, not closing, as more of the buyer's research process moves into AI-mediated surfaces before Google even renders a results page.
Dealers who are asking their agencies for accountability on paid spend are asking the right question. They are rarely asking the same question about the organic content layer that either supports or undermines everything the paid budget is trying to do.
How AUTONOMi's ECHO Engine Builds the Content Layer the Research Cannot Measure
ECHO, AUTONOMi's organic content engine, produces per-dealer blog content at machine cadence: 20 articles per month, each built from keyword research using Google Trends and Google Autocomplete signals, planned against the dealer's own market and franchise, and optimized for the specific queries buyers in that DMA are running. The output is not generic automotive content. It is topical authority targeted at the exact searches that precede the paid auction the dealer is running.
Every ECHO article ships with a machine-extractable claim graph: every factual assertion structured as a Claim row with a subject, predicate, object, verbatim supporting quote from an external source, and a lastVerified timestamp. That structure is what AI answer engines prefer to cite. A dealer whose content carries verified claim graphs with sourced timestamps is more likely to surface in an AI Overview or a Perplexity answer than a dealer whose blog is a wall of unstructured marketing prose. ECHO tracks AI answer engine citations as a KPI, with nightly probes recording every time a dealer article is cited by an AI answer engine, split by engine.
AUTONOMi's Google Search Console integration syncs organic search queries, clicks, and impressions nightly, building the organic half of the picture alongside the paid campaign data AEGIS manages. This does not close the attribution gap that the Harvard research will face. Nothing can, because that gap is structural. What it does is give the dealer visibility into the organic layer that a study like this will never capture, and the ability to build that layer deliberately rather than by accident.
The ECHO module is available as a selectable add-on on every plan, including the entry tier. ECHO is priced separately at $1,799 per month per rooftop and is not bundled into any plan tier. It covers 20 articles per month with Claim Graph verification, cover and inline imagery, and social crossposting to the dealer's owned channels. It buys no media. It builds the layer that paid media alone cannot build, and that ad-spend research structurally cannot measure.
What the Next Generation of Dealer Ad Research Will Have to Reckon With
The Harvard and GM study will produce findings worth reading. Any rigorous attempt to trace dealer ad spend to actual business outcomes is useful, and the industry is overdue for that kind of accountability. The conversation about AI in dealer marketing has been dominated by tool vendors and conference panels for long enough; empirical research from outside the vendor ecosystem is genuinely valuable.
But the findings will have a ceiling. They will explain variation in paid-channel efficiency without being able to explain why that variation exists. The dealers who look most efficient will, in many cases, be the dealers who built organic topical authority before the study's observation window opened. The dealers who look least efficient will, in many cases, be the dealers running their entire brand relationship through the paid auction because there is no content layer to do the pre-sell work.
The next generation of dealer ad research that wants to actually explain CPL variation will need to include organic search authority as a covariate: published article count, keyword ranking coverage in the dealer's DMA, AI answer engine citation rate, Search Console click data. Until those variables are in the model, the research will correctly identify what the money is doing and consistently miss why some of it works. The dealers who understand that gap now, and build accordingly, will be producing the outlier results the next study finds most difficult to explain.
If you want to see what building that content layer looks like in practice, the right starting point is not a new agency contract. It is a content engine running at the cadence the organic layer actually requires. Sign up to see how ECHO maps a content strategy against your market and your franchise, and what that layer does to the economics of the paid budget running beside it.
Source: Auto Remarketing
