Facebook Ads for Businesses That Deliver

Most Facebook Ads advice is written for a business that can sell to anyone. Yours can’t. You can sell to the people inside the area you’re able to deliver to at a price that still leaves a margin, and everybody else is an impression you paid for and can’t use.

That one constraint changes almost every setting in the account. It changes which objective you pick, how wide the radius goes, what the creative has to show, how much you can afford per order, and what counts as a result. It’s also the thing the standard guides skip, because a national brand has no delivery zone to respect.

This guide is written for the other case: a bakery, florist, caterer or wholesaler running its own vans and taking orders online. It covers the whole account, start to finish. Before you spend anything, it helps to know what Facebook ads cost in 2026, because the budget question decides whether the rest of this is worth doing at all.

The Bottom Line

• Your ad radius should be drawn from your delivery zone, not from your ambition. Every mile you add that a van can’t profitably reach is budget spent on people who will never become an order.
• Facebook clicks are cheap in food categories. WordStream’s 2026 benchmarks put the median cost per click on a traffic campaign at $0.60 across all industries and $0.45 for restaurants and food, off 1,377 campaigns run between April 2025 and June 2026.
• Pick your objective by what you can measure. If orders happen on your site, run a sales campaign with a purchase event. If they happen by phone or email, run leads and accept a messier number.
• A warm audience beats a clever ad. People who reached your order page and left are the cheapest conversions in the account, and most local businesses never build that audience.
• Give it a month and a fixed budget before you judge it. Meta’s system needs volume to learn, and three days of data tells you nothing except how impatient you are.

What changes about Facebook Ads when you deliver the order yourself

Three things, and they compound.

Geography is a hard wall here. A shop with a physical counter benefits from being seen slightly too far away, because somebody might drive in. You don’t get that. If the address is outside your zone, the order either gets refused or gets delivered at a loss. The radius is a business rule you are copying into an ad platform.

Your margin sets the ceiling on what a customer can cost. Delivery eats into the order before you’ve paid for the ad. Fuel, the driver’s time, the trip back, the packaging that keeps the product intact. Whatever an order is worth, subtract the cost of getting it there, and what’s left is what you have to buy the customer with. Businesses that skip this arithmetic discover it four weeks in.

Repeat ordering is where the money is. A single $60 order rarely justifies a $27 cost per lead. The same customer ordering monthly for a year does, easily. That’s why your campaign structure should be built to earn the second order, not just the first, and it’s why retargeting matters more here than in most industries.

Everything below follows from those three.

Setting up your first Facebook Ads campaign

Meta’s structure has three levels, and people lose money by confusing them. The campaign holds the objective. The ad set holds the audience, the placement, the budget and the schedule. The ad holds the creative. Most of the decisions that matter live at the ad set level, but the one that matters most is at the top.

Choose the campaign objective that matches what you can measure

Meta offers six objectives. For a business selling deliverable goods, only three are realistic starting points.

Choose Sales if orders complete on your website and you have the Meta pixel firing a purchase event. This is the right answer for most online storefronts, and it’s the only objective where Meta optimizes toward the outcome you want.

Choose Leads if orders start with a conversation: a catering enquiry, a wholesale account application, a quote for a wedding. You’ll be buying form fills or messages, not revenue, so you need a conversion rate from enquiry to order before the numbers mean anything. WordStream’s 2026 data puts the median cost per lead at $27.39 and the median cost per click on a leads campaign at $1.80, roughly three times the traffic-campaign figure.

Choose Traffic only as a deliberate first step, when you don’t yet have enough conversion volume for Meta to optimize on and you want cheap clicks to test which products and which audiences respond. It’s a diagnostic, not a strategy. Traffic campaigns buy visits, and visits are not orders.

Awareness, engagement and app promotion are not useful to you. Awareness sells reach to advertisers who measure in brand lift. You measure in deliveries.

Set the conversion event and the budget at the ad set

Two settings here decide whether the campaign can work at all.

The conversion event should be the last step before money changes hands. Purchase if you can track it. “Initiate checkout” or “add to cart” only if purchase volume is too low for Meta to learn from, which in practice means fewer than around 50 purchases a week. If you optimize for add-to-cart, expect Meta to find you excellent cart-fillers who never pay.

The budget should sit at the campaign level (Meta calls this Advantage campaign budget) if you’re running several ad sets and want the system to move money toward whichever one works. Keep it at the ad set level if you need to guarantee spend against a specific audience, which is usually the case when one ad set is a warm retargeting audience you don’t want starved.

Set a daily budget you can sustain for four weeks without flinching. Meta’s delivery system spends the first stretch of any new ad set learning, and turning it off on day five throws that away and starts it again.

Matching your ad radius to your delivery zone

This is the setting that separates a local Facebook Ads account that works from one that quietly burns money, and it takes about ten minutes.

Start from the operational side. Draw the area where a delivery still makes money after fuel, driver time and the trip back. Not the furthest you’d be willing to drive once for a big order, and not your whole metro. For most urban food and floral businesses, that zone is smaller than they’d guess.

Then copy it into the ad set. Drop a pin on your address and set the radius, or enter ZIP codes if your zone doesn’t follow a clean circle, which is common when a river, a highway or a bridge toll makes one direction expensive. Meta also supports store-set location targeting for businesses with several pickup or dispatch points, documented in Meta’s Business Help Center.

Typical starting radii, as a sanity check rather than a rule: 5 to 10 miles in a dense city, 10 to 15 miles in suburbs, 20 to 30 miles where addresses are spread out. If your van can’t clear a 25-mile round trip inside a delivery window, the radius is wrong no matter what the reach estimate says.

Then check the setting most people miss. Meta gives you four location options, and the default is the wrong one for you:

• People living in this location targets residents. This is almost always what you want, because residents are the ones who reorder.
• People in or recently in this location includes commuters, tourists and anyone whose phone passed through. It inflates your reach and your costs and delivers to people with no address in your zone.
• People travelling in this location is narrower still and irrelevant to a delivery business.
• Everyone in this location is the default and the widest. Change it.

One more leak: exclusions. If you have a wholesale side and a retail side, or a zone you serve only on certain days, build that into the ad set with exclusions rather than handling it with an apologetic phone call later.

Which Facebook ad formats suit goods that get delivered

Format matters more than it looks like it should, because the formats differ in what they can show and in how much work they take to produce.

The short version for a delivering business: single image ads are the cheapest to make and the right default for one hero product or one offer. Carousels earn their place when you have a range, because each card carries its own image and its own link, up to ten of them, per Meta’s carousel documentation. Video works when the product needs demonstrating or when the appeal is the making of it. Collection and catalogue-based formats matter once you have a product feed worth pulling from. Messenger and lead-form ads suit the businesses whose orders start as conversations, which is most caterers.

Which one to run, what each costs you in production time when nobody on the team is a designer, and which formats hold up for perishable goods are all covered in the companion piece on the types of Facebook ads and when each one earns its budget.

One rule that applies regardless of format: the creative has to answer “can you get this to me, and when.” A photograph of a beautiful cake that doesn’t say you deliver to this city by tomorrow is a brand ad. You can’t afford brand ads.

What Facebook Ads cost in 2026

Benchmarks are for sanity-checking your own numbers, not for predicting them. Here’s the current picture, from WordStream’s 2026 Facebook ads benchmark study, drawn from 1,377 campaigns between April 2025 and June 2026, and from Triple Whale’s August 2026 analysis of more than 40,000 ecommerce brands.

Two numbers matter more than the rest.

Food and restaurant clicks are the cheapest on the board at $0.45, with the highest click-through rate at 2.68%. That’s unusual and it’s in your favour. A $300 monthly budget buys a useful volume of clicks in these categories rather than a rounding error.

And the cost per lead spread is enormous. LocaliQ’s benchmark data, last updated in June 2026, runs from $3.16 in restaurants and food to $76.71 for dental practices. Which end you land on depends far more on your audience size and your offer than on how cleverly the account is built.

The full breakdown, including what a realistic monthly minimum looks like and how to work out the most you can afford to pay for an order, is in the guide to what Facebook ads cost and what to budget.

Reaching people who nearly ordered and didn’t

Every business running ads has an audience it isn’t using: people who visited the site, browsed a product, started a checkout and left. They’re already interested, they already know what you sell, and there are very few of them, which is exactly why reaching them again is cheap.

Meta builds these as Custom Audiences from website activity, catalogue activity, video views, form opens or an uploaded customer list. Standard website custom audiences hold up to 180 days of activity, and an audience needs at least 100 people in it before it can be used at all, which is worth knowing before you build a 7-day cart abandoner list and wonder why it won’t run.

For a business that delivers, this works in two directions. Backwards, to recover the cart somebody abandoned when they hit the delivery fee. Forwards, to reach the customer who ordered once in March and hasn’t since, which is a far easier sell than a stranger. The audience building, the retention windows and the offer sequencing are covered in depth in the piece on Facebook retargeting ads and turning browsers into repeat orders.

If you only do one thing from this guide beyond fixing the radius, build the retargeting audience. It costs nothing to create and it starts accumulating people the day you set it up.

Writing ads for a product that has to arrive intact

Creative for a delivering business has a job that creative for a shop doesn’t: it has to remove the doubt about arrival.

Show the product as the customer receives it, not as it looks on your bench. A flower arrangement in the box, a cake in the carrier, a catering tray with its lid on. This is the single most useful photograph a delivering business can take and almost nobody takes it.

Put the zone in the copy. “Delivered across [city] by 10am” does more work than any adjective. It answers the question that stops the order.

Name the cutoff. Ordering deadlines create urgency honestly, without the fake-scarcity countdown clocks that make people distrust a small business.

Price the delivery visibly if you can. Surprise delivery fees at checkout are the most common reason a warm visitor abandons, and a fee disclosed in the ad costs you a click you were going to lose anyway.

And run more than one ad. Three or four variations in an ad set lets Meta find the one that works instead of you guessing, and creative fatigue on a small local audience arrives fast, usually within two to three weeks.

Measuring Facebook Ads when the order finishes offline

If everything completes on your site, install the Meta pixel, define the purchase event and read the numbers. That’s the easy case.

The harder and more common case is a business where the ad starts the process and a person finishes it. A caterer gets an enquiry and quotes by email. A wholesaler gets a call. In that world, Meta’s reported cost per lead is a partial truth, and treating it as revenue is how businesses conclude Facebook Ads don’t work when what doesn’t work is their attribution.

Three things fix most of it.

Ask. A one-line “how did you hear about us” field on the enquiry form is crude and it works better than most tracking.

Track the step you can see. If the enquiry form submits on your site, that’s a pixel event you can count, even if the order closes weeks later. Record what share of enquiries turn into orders, and you can convert Meta’s cost per lead into a cost per customer.

Use a separate landing page or phone number per campaign, if the volume justifies the setup. A campaign-specific number makes the phone orders countable.

Then judge the account on cost per order and on what a customer is worth over a year, not on click-through rate. Click-through rate tells you whether the creative is interesting. It says nothing about whether you made money.

Your first 30 days on Facebook Ads

A realistic sequence for a business starting from nothing.

In week one, do the plumbing. Install the pixel, confirm the purchase or lead event fires, create the website custom audience so it starts collecting, and set your location targeting to residents inside the delivery zone.

In week two, run one campaign with two ad sets: a cold audience inside the radius, and one broad interest or Advantage+ audience. Same creative in both. Fixed daily budget. Then leave it alone; changing an ad set resets the learning phase.

In week three, read the numbers instead of reacting to them. Look at cost per result by ad set and by creative. Kill the clearly worst ad, not the whole ad set. If nothing has any conversions yet, the budget is probably too small for the conversion event you chose, and dropping to a mid-funnel event for a while is the fix.

In week four, turn on retargeting. By now the custom audience has enough people to run. Point a small, separate budget at it with a specific offer, and expect its cost per order to be the best in the account.

After that, the loop is creative refresh, audience expansion and a monthly look at whether cost per order still clears your margin.

Where local Facebook ad budgets leak

In rough order of how much money they waste:

• Radius too wide, or set to “everyone in this location.” The default setting alone can put a large share of your spend outside your delivery zone.
• Judging a campaign in under two weeks. You’re reading noise and then acting on it, which resets the learning phase and generates more noise.
• No retargeting audience. You are paying full price for strangers while your warmest prospects go untouched.
• Optimizing for the wrong event. Clicks and add-to-carts are cheap to buy and easy to mistake for progress.
• One ad, running for months. Small local audiences see the same creative repeatedly, and performance decays.
• Advertising into a broken ordering flow. If the checkout hides the delivery fee until the last step, better ads only buy you faster abandonment.

Frequently asked questions

How much should a local business spend on Facebook Ads per month?

Enough to generate roughly 50 conversion events a month so Meta’s system has something to learn from, which for most food and floral businesses at a sub-$1 cost per click means $300 to $500 a month. Below about $200 a month you can still buy clicks, but you won’t get enough conversion data for the optimization to do anything, and you’ll be judging the channel on a sample too small to judge.

Do Facebook Ads work for a business with a small delivery radius?

Usually better, not worse. A tight radius means a small audience, and small audiences are cheap to reach repeatedly. The trade-off is that creative fatigue arrives faster because the same people see your ads more often, so plan on refreshing the creative every few weeks rather than every few months.

Should I use Advantage+ or build audiences manually?

Start with one of each and let them compete. Meta reports an average 17% improvement in cost per conversion for its Advantage+ shopping campaigns against manually configured ones, but automated targeting needs conversion volume to work with, and a small local account often doesn’t have it yet. Keep the geographic constraint locked either way.

Is it better to advertise on Facebook or Instagram for a delivery business?

You don’t pick one. Both run through the same Meta Ads Manager, and placements are a setting inside your ad set rather than separate campaigns. Let Meta distribute across both and look at the placement breakdown after a few weeks. Visual, impulse-led products often earn their keep on Instagram; enquiry-led wholesale and catering usually convert better on Facebook.

How long before Facebook Ads produce orders?

Expect two to three weeks before the data means anything and four weeks before you can judge cost per order honestly. Retargeting audiences produce results faster once they’re populated, because you’re reaching people who already wanted the thing.

Where to start

If you take nothing else from this: fix the radius first, install the pixel and create the retargeting audience on day one, and pick a conversion event you can measure. Those three decisions determine more of your result than any amount of creative work.

Then give it a month at a budget you can sustain, and judge it on delivered orders and repeat rate rather than on clicks. Facebook Ads are a reliable way to fill a delivery route in a defined area. They’re a fast way to lose money in an undefined one.