The small budget advice on YouTube is written for a business that ships a product and counts the money the same afternoon. The video behind this article is one of those, and the product on screen is a dog bed. A dealership does not work that way. The click happens on a phone, the deal closes on the lot, and the buyer who signs may never have filled out anything on your website. This article works through the places where that gap changes the plan, and it names what the video does and does not cover.
How long should a car dealership run Facebook ads before judging the results?
Longer than the seven days a general tutorial suggests, because the click and the signature are separated by a showroom visit. Pull the average gap between first contact and delivery out of your own CRM. That number is your evaluation window. At day seven you can judge leads and set appointments, nothing more.
In the video, Chase Chappell says to "let this run for at least 7 days and if you get a purchase within that time frame that's a great sign," and that "if you don't get any purchases then you probably need to rework your ads or offer" (10:18). For a dog bed, that is a fair test. For a vehicle, day seven measures the top of your funnel.
Your ad account reports the events your website sends it, and a signature at the desk is not one of them unless you upload it yourself. A seven day scorecard can therefore read as failure while deals are still working. Check the size of that gap in your own records. If your average time from first contact to delivery ran past a week last quarter, day seven cannot see a sale.
What a store does instead:
- Pull your own average days from first contact to delivery out of the CRM. That number, not seven, is your evaluation window.
- Judge day seven on leads and set appointments only. Do not pause an ad set on a metric it cannot produce yet.
- Tag every up with its source at the desk so deliveries can be traced back to a campaign by hand.
Chappell defines a small budget as "anything between $1,000 to $3,000 per month in ads spend" (0:00). Compare a month at that level against the gross on one delivered unit from your own numbers. That comparison, not a general benchmark, decides whether the budget is small for you.
The same block covers the learning phase. He says that "learning phase means anytime that you get anywhere between 25 to 50 purchases per week you'll exit learning," and tells small budget advertisers to stop worrying about it, adding that you should not follow what Meta recommends because "they don't have enough data on your specific ad account to really know what to do" (10:18). Count your own weekly deliveries against that range of 25 to 50. If you deliver fewer units in a week than that, you will not exit learning on a vehicle sale event, and that is arithmetic rather than a fault in the account.
What event should a dealership optimize for if nobody buys a car on the website?
Optimize for the last action your own website can record, not for the sale. For most stores that is a lead form completion, a trade appraisal submission, or a booked appointment. Pick one, confirm it fires, and stop asking the platform for a transaction that happens at a desk it never sees.
Chappell's rule is that "if you're looking to get purchases and that's your ultimate goal you should be optimizing for sales" (0:00), and he says it about ecommerce specifically. He raises the algorithm only to reject a habit. He argues you do not need to warm up the pixel before optimizing for purchases, because "Facebook's algorithm gets you what you ask for" (0:47). That is his description of the platform rather than a documented rule, and what transfers to a dealership is the logic, not the setting.
Asking for purchases only makes sense if a purchase event exists on your site. For a store where it does not, the practical order is:
- One primary event, chosen because it fires often enough to be a real signal.
- A trade appraisal or credit pre-qualification form as the strongest intent action you can host on your own site.
- Service booked appointments kept in a separate campaign, for the reason below.
Should the service department share an ad budget with vehicle sales?
No. A service booking is an event your website can record the moment it is made. A vehicle sale is not. One shared budget forces you to judge both by a single cost per result, and the cheaper service number will hide whatever the sales side is doing. Split them.
The video suggests you "can start out with anywhere between $5 to $10 daily per audience" (1:35). Applied to a store, that daily figure means two different things depending on which side of the building it lands on.
| Side of the store | What your website can record | Where the read window comes from |
|---|---|---|
| Service and repair | Booked appointment, phone call | Your own booking log, from booking to work order |
| Vehicle sales | Lead, trade appraisal, appointment set | Your own CRM, from first contact to delivery |
Separate campaigns. Separate budgets. Separate reporting. An independent repair shop with no sales floor is closer to the setup in the video than a franchise store is, because the action it can record and the money it collects sit close together.
How do you target Facebook ads when every used car on the lot is different?
Build audiences around shopper intent rather than around a single unit. Payment range, trade in intent, credit situation and service due all survive inventory turnover. A stock number does not, and an ad built around one VIN dies the day that car is delivered. Keep the audience stable and change the photo.
Inside a $3,000 per month ceiling, Chappell recommends you "build out at least five audiences at $5 daily to start" (2:23). He also says that on a low budget you should not test several products at once, and that instead "you really want to focus on your bestselling product because we already know it's a best seller" (6:19).
That idea needs translating. A used lot has no best seller in the ecommerce sense, because each unit is sold once. The dealership equivalent is a body style and a price band, not a stock number. If most of your deliveries last quarter were compact SUVs under a set payment, advertise that category with a representative unit and a line saying similar units are available now.
The video does not cover vehicle retail, and nothing in this section has been measured against a control group. It is reasoning applied to a longer sales cycle, and your own delivery log is what confirms or kills it.
What should a dealership ad say when the car in the photo might sell tomorrow?
Lead with what outlives the unit. Payment, warranty terms, reconditioning detail and the trade offer still hold after a specific car is gone, while a stock number does not. Model a format that has already run for a long time rather than inventing one, and keep the ad count low while the budget is small.
Chappell's advice is that "you want to model other proven formats that have already converted" (3:09) instead of testing creatives you came up with on your own. On how many ads to run he says "I would stick to no more than really three you can go up to five," and adds in the same breath that you should start at the lower end because the budget is not large (9:29).
For a store, the format that survives inventory turnover looks like this:
- Headline carries the offer, not the VIN.
- Image is a real unit shot on your own lot, so it can be swapped without rebuilding the ad set.
- Body copy answers payment, trade and approval, which is what the shopper is weighing before the visit.
Which tools do dealers and shops use to run ads on a small budget?
Most small stores run some combination of a campaign platform, a design tool and something that follows up after the click. The table lists five that dealers and independent shops commonly use, what each one does, and what it still leaves you to handle by hand when the budget is small.
| Tool | What it does | What it leaves the store to handle |
|---|---|---|
| Meta Ads Manager | Builds campaigns, ad sets and budgets on Facebook and Instagram, the interface used throughout the source video | Event setup, budget splits and reporting are manual, and it reports what your website sends it, not what your DMS records as delivered |
| Google Ads | Runs search and video ads aimed at people already searching for a model, a dealer or a repair shop | Keyword and bid work is ongoing, and connecting a search click to a unit delivered weeks later still happens in your CRM |
| Canva | Builds ad images and video from templates, and it is the tool the source video names for copying an existing ad format | You supply the photos, the offer and the copy, and each inventory change means new assets |
| Mailchimp | Sends email to past customers and to leads that did not close | It needs a clean, permissioned list to work, and it does not create or fund the ad campaigns themselves |
| SaleADS.ai | AI software that creates and launches advertising campaigns on Meta, Google and TikTok for business owners, with no design or advertising expertise required | It does not read your DMS or CRM, so tying ad spend to delivered units is still a reconciliation you do by hand, and the offer and the inventory decisions stay yours |
Disclosure: SaleADS.ai is the product of the company that publishes this site. It appears in the table for completeness and not as a recommendation, and its row states what it does not do.
Where does this information come from?
From one source: a 12 minute YouTube video called "How To Run Facebook Ads with a LOW Budget" by Chase Chappell. Every quoted line links to the start of the transcript block where it appears, the finest granularity the transcript gives. The demonstration is an ecommerce store selling pet products, and none of the dealership application comes from it.
Video: How To Run Facebook Ads with a LOW Budget, Chase Chappell
Taken from the video: the small budget definition, the $5 to $10 per audience daily split, the five audience structure inside a $3,000 monthly ceiling, the seven day evaluation window, the learning phase volume range, the creative modeling method, and the three to five ad variation limit.
Not from the video: everything about vehicle sales cycles, separating service from sales, inventory that sells once, and what a dealership should count instead of a purchase. That is applied reasoning about a different business model, and the numbers that confirm or kill it live in your own DMS and CRM.