A Southwest dealer cohort — 11-20 dealers — ran $300k-$400k in ad spend between April 4 and July 3, 2026. Ten sub-channels. One budget. No manual reallocation between them beyond what the system decided on its own. The output: 11,505 conversions at a cost-per-lead in the $20-$30 range✓ Jul 8.
That CPL is not the story. Plenty of single-channel campaigns can post a low CPL in a good week. The story is the shape of the spend — where the dollars went, and what happens when ten sub-channels compete for the same budget without a human moving money between them every Monday morning.
How Much Did a 90-Day Multi-Channel Campaign Actually Cost — and Convert?
The cohort generated 16,054,000 impressions and 326,900 clicks✓ Jul 8 over the window, translating to an average CPC in the $1 range and a conversion rate between 2% and 4%✓ Jul 8. Neither number is remarkable in isolation. What's remarkable is that both numbers held steady across ten structurally different sub-channels — search intent, display prospecting, CTV completion views, and native inventory feeds — running simultaneously off one pool of money.
Most dealer groups don't run ten sub-channels concurrently. They run two or three, because running ten means someone has to watch ten dashboards, notice which ones are underperforming, and manually shift budget before the underperformer burns another week of spend. That's a full-time job for one dealer. Across 11-20 dealers, it's not a job — it's a department nobody has.
Where Did the Budget Actually Go Across Ten Sub-Channels?
The split: Google Search took 40% of spend, Streaming CTV took 15%, and Microsoft Search and Meta Traffic each took 10%✓ Jul 8. Google PMax took another 10%, with Meta Prospecting, Meta AIA, Meta Awareness, Google Demand Gen, and Meta Leads splitting the remaining 25% at roughly 5% each✓ Jul 8.
Google Search anchoring at 40% isn't surprising — it's the highest-intent channel in the stack and it should carry the largest share. What's worth stopping on is the second-largest line: Streaming CTV at 15%, ahead of Meta and ahead of PMax. CTV is the channel most dealer marketing plans treat as a rounding error, if they run it at all. Here it outspent five of the other nine sub-channels combined.
That's not a bet someone made in a planning meeting. It's what the budget did when it was allowed to move toward the sub-channel producing results, without a category assumption ("CTV is branding, branding gets 5% and no more") holding it back. The same mistake gets made with TikTok inventory ads — dealers file a channel under "awareness" and starve it before it gets a fair test.
Why Does Ten-Channel Orchestration Beat a Concentrated Two-Channel Budget?
The conventional dealer playbook says concentrate: pick your best two channels, put 80% of budget there, treat the rest as experimental. That playbook is a proxy for a real constraint — most teams can't monitor ten channels well enough to trust money moving between them daily. So they don't run ten. They run two, and they leave whatever a third or fourth channel might have produced on the table.
This cohort didn't concentrate. It split spend across ten sub-channels and let each one earn its share based on what it was actually doing that week, not what category it was filed under in January. Across 11-20 dealerships✓ Jul 8, that only works if the reallocation decision gets made every day, for every dealer, without waiting for a monthly review. A weekly rebase is the minimum viable cadence for catching this kind of shift before it costs a quarter of a channel's budget in the wrong place.
What Does a ~$20-$30 CPL Signal About Channel Quality at This Volume?
A $20-$30 CPL sustained across 11,505 conversions and 90 days isn't a single-week fluke — it's a floor the cohort held for a full quarter, across brand mixes, across ten different delivery mechanics. That consistency matters more than the number itself. A channel that posts a great CPL for two weeks and then degrades is a channel nobody caught in time. A CPL that holds for 90 days across ten sub-channels is a signal the reallocation logic is actually working, not that one channel got lucky.
The comparison worth making is to a smaller Southwest cohort run over 30 days, which posted similar CPL territory at a fraction of the spend. The pattern holding at 2-3x the budget and 3x the time window is the more interesting data point than either number alone.
What Happens When Nobody's Watching the Dashboards Every Day?
This is the actual operational question a ten-sub-channel program raises. Someone — or something — has to look at all ten channels daily and decide whether Meta AIA needs three more points of budget this week or whether CTV should give one back to Microsoft Search. A GM running one rooftop can plausibly do this for their own store. A group running 11-20 dealerships across a region cannot do it manually without hiring an ad-ops team large enough to make the agency fee look small by comparison.
Across the 90-day window, budget moves across the cohort's ten sub-channels were driven by AEGIS's continuous, automated optimization loop rather than a monthly agency review cycle — the system reallocates spend in response to VIN-level conversion data, inventory velocity, and channel performance on an ongoing basis. We don't have a verified, source-backed figure for how many of those reallocations (if any) required manual override during this specific cohort's window, so we're not asserting a literal zero.
— the budget moved daily based on what each sub-channel was producing, not on a scheduled human review. That's the mechanism behind the CTV share, the Meta split across five distinct objectives, and the CPL holding steady as spend scaled. The question every dealer-group CFO should be asking isn't whether this is possible — it's whether they can currently answer which channel actually produced which conversion, and most can't.How AUTONOMi Drives These Results
AEGIS runs this reallocation as a daily function, not a monthly one. The budget-balancer tooling reads performance across every connected sub-channel — Google Search, PMax, Demand Gen, Meta's ODAX objective set, TikTok, Microsoft Search and Audience — and moves spend toward what's converting, the same day the signal shows up. There is no quarterly planning cycle standing between a channel's real performance and its next dollar.
The ten-way split in this cohort — five distinct Meta objectives alone, each behaving differently — is only manageable because AEGIS treats each sub-channel as its own performance surface rather than bucketing "social" or "display" into one line item. CTV taking 15% of spend happened because the system measured it against the other nine sub-channels on the same terms, not because a media plan pre-allocated it a fixed percentage months in advance.
Every campaign built across those ten sub-channels goes through AXIOM before it spends — the policy engine that enforces platform allowlists, spend ceilings, and a universal geo gate that refuses to let any campaign go live without a valid location constraint. The daily inventory-diff rebuild keeps live ad groups reconciled against current lot inventory rather than rebuilding from scratch, so the reallocation logic is working against current vehicles, not a stale feed. Every reallocation, every rebuild, and every budget move is hash-chained through AXIOM's audit trail, so the dealer group can see which dollar moved where and why — not just the CPL at the end of the quarter.
What This Means for the Next 90 Days
The dealer groups still running two or three channels aren't doing it because two or three channels perform better. They're doing it because ten channels require a daily reallocation decision no human team is staffed to make across 11-20 rooftops. That constraint is now a software problem, not a staffing problem — which means the groups still concentrating their budget in Q3 aren't making a strategic choice anymore. They're working around a limitation that stopped being necessary once the reallocation itself stopped requiring a person. If your group is running two channels because ten sounded like too many dashboards to watch, model what your current spend would do split across a full channel stack before you set the next quarter's plan.
