How to make Facebook ads for your Shopify store without a designer

Ask a solo Shopify founder what is stopping them from running Facebook ads and you will hear about targeting, the pixel, or "the algorithm." It is almost never the real answer. Ads Manager is a form you fill out once. The hard part of how to make Facebook ads for Shopify is creative: you need several genuinely different ads to test, Meta's delivery system keeps rewarding fresh creative, and you — the person also doing support, shipping, and inventory — have zero design hours in your week.

This playbook takes you from nothing to a live campaign: account setup, picking the photo, choosing the angles, producing the creatives (three honest options, including ours), launching in Ads Manager, and reading the results. It is written for a store run by one person, on a one-person budget.

Step 0 — Prerequisites: get the boring parts done once

Three things need to exist before any ad runs. None of them require a designer, and all of them are one-time setup.

1. A Meta business portfolio and ad account. Create both in Meta Business Suite (business.facebook.com). You will also need a Facebook Page for your store — ads run "from" a Page — and a payment method on the ad account. Do this days before you plan to launch; brand-new ad accounts sometimes hit verification checks, and you do not want that on launch day.
2. The Meta pixel and Conversions API, via Shopify. Install the Facebook & Instagram app from the Shopify App Store and connect it to your Meta business portfolio. It sets up the pixel and the server-side Conversions API for you, so Meta can see page views, add-to-carts, and purchases without you touching code. This is standard Shopify functionality, and it is the single most important prerequisite: without purchase data flowing back, Meta cannot optimize and you cannot measure anything.
3. A product with margin to support paid traffic. Run the math before you spend a dollar. Take your price, subtract product cost, shipping, payment fees, and packaging. What is left is the most you can pay to acquire a customer and still break even on the first order. If that number is $8, Facebook ads will be a grind. If it is $30 or more, you have room to test. Write your breakeven number down — you will use it in Step 5.

Step 1 — Pick the one product photo that can carry an ad

Every static ad you run will be built around a product image, so the photo decision comes before any design decision. You do not need a studio shoot. You need one photo that meets four tests:

• Clean background or an honest in-context shot. A product on a plain, uncluttered surface works. So does the product genuinely in use — a candle lit on a nightstand, a serum on a bathroom shelf. What fails is a busy photo where the eye has to hunt for the product.
• Sharp detail at small sizes. Most people will see your ad on a phone, in a feed, for under a second. Zoom your photo out to thumbnail size. If you cannot tell what the product is, neither can a scroller.
• Breathing room for text. Ads need a headline and usually an offer. A photo where the product fills every pixel leaves nowhere to put words. Favor shots with some negative space above, beside, or below the product.
• Accurate color and texture. The photo is a promise. If the linen looks cream in the ad and arrives beige, you have bought a refund, not a customer.

One good photo is enough to start. The whole point of the next two steps is squeezing multiple distinct ads out of a single image — and if you want to see what one photo can become, that is exactly what a product photo to ad generator is for.

Step 2 — Decide the angles before the design

Here is the mistake that quietly kills most first campaigns: founders make three ads that are really the same ad — same message, different background color. When all three flop, they conclude "Facebook ads don't work for my product." Nothing was actually tested.

An angle is the buying reason an ad leads with. Before opening any design tool, write down 3–5 angles that contradict each other — different enough that a stranger would describe them as different pitches:

• Launch. "New" is a message by itself. "Just dropped: the 40-hour soy candle." Works best for genuinely new products — see the product launch angle.
• Proof. Lead with a review, a rating, or a number. "4.9 stars from 212 burners." If you have real social proof, this angle often beats everything else — the social proof angle is built around it.
• Offer. Lead with the deal: free shipping, a bundle, a launch discount. Easy to test, easy to overuse.
• Premium. No discount, no urgency. Lead with materials, craft, design. Whitespace and confidence. This angle and the offer angle cannot both be your brand — which is exactly why you test them.
• Problem/solution. Name the pain first, product second. "Candles that die at hour ten" → yours doesn't.

For each angle, write one headline and one supporting line. That is your creative brief — maybe ten minutes of work, and it is the highest-leverage ten minutes in this whole playbook. If staring at a blank page stalls you, an AI product ad generator will propose angles from the photo itself, and you keep or kill them.

Step 3 — Make the creatives: three honest options

You need each angle turned into finished ad images, in multiple sizes. There are three real ways to do this, and each wins in a different situation.

DIY in Canva is the default for a reason: it is free and you control everything. The honest downsides are time and the blank-canvas problem — you still have to invent five different layouts, and most founders end up making five near-identical ads (see Step 2). If your store's whole aesthetic is hand-built and you like the work, Canva is a fine answer. Here is our full Canva comparison.

Hiring a designer produces the best single ads, full stop. For a brand system, a rebrand, or a hero launch campaign, a good designer is worth every dollar. For weekly creative testing it is the wrong tool: turnaround is measured in days, revisions in invoices, and the testing loop in Step 5 wants new variants on demand. More on when each makes sense in generator vs. hiring a designer.

A generator built for this job is the third option. With Product AdKit you upload the one photo from Step 1, point it at your angles and offer up front, and a few minutes later you have a pack of 14 finished static ad concepts — text baked into the image, sized for Facebook feed, 1:1, 4:5, 9:16, and Pinterest. It is a $29 one-time pack, not a subscription, and only ads that pass quality checks count against it; if fewer than 12 pass, the difference auto-refunds. There is a free preview — one watermarked ad, no credit card — so you can judge the output on your actual product before paying. The honest limitation: it is not a canvas editor. You steer the direction and regenerate, or ask for more like a winner; you do not drag text boxes around.

Whichever route you pick, the exit criteria are the same: 3–5 visually and verbally distinct ads, each carrying one angle from Step 2, exported in at least 4:5, 1:1, and 9:16.

Step 4 — Set up the campaign in Ads Manager

With creatives in hand, the Ads Manager part is genuinely quick. Meta keeps renaming things — its Advantage+ shopping campaigns and manual sales campaigns have been folding into a single flow — so treat the button labels loosely; the structure below is what matters.

1. Create a Sales campaign. In Ads Manager, choose the Sales objective. Your conversion event is Purchase (this is why the pixel from Step 0 had to come first — the event must already be firing).
2. One campaign, one ad set. Resist the urge to build a complicated structure. One campaign with one ad set keeps your budget concentrated and your read clean.
3. Go broad on targeting. The 2026 default is to let Meta's Advantage+ audience do the finding. Set your country, maybe an age floor, and stop. Hand-stacked interest targeting is mostly a relic; the creative itself does the targeting now — which is the other reason distinct angles matter.
4. Load all 3–5 ads into the ad set. Each ad is one angle. Meta will shift spend toward whichever resonates.
5. Use placement customization for sizes. Run automatic placements, but supply the right asset per placement group: 4:5 for Facebook and Instagram feeds, 1:1 where square is called for, 9:16 for Stories and Reels. A square ad letterboxed into a Story placement looks like an ad nobody finished making.
6. Set a budget that can actually exit learning. Rule of thumb, not law: the ad set should be able to generate enough purchases for Meta to optimize — think on the order of dozens of conversions in a week. For many small stores that lands somewhere around $20–50/day, but it genuinely depends on your price point: a $25 product can learn on less than a $150 product. The wrong move is $5/day "just to try it" — the ad set never gets enough signal and tells you nothing.

Publish, then leave it alone. Seriously — that is Step 5's first instruction.

Step 5 — Read the results and refresh the creative

For the first several days, do not touch anything. Every significant edit can reset the learning phase and scramble your data. Check in daily, change nothing.

When you do read results, four numbers tell most of the story:

• CTR (link click-through rate). Is the creative stopping anyone? Around 1% is a workable benchmark for cold traffic for many stores — but treat it as a ranking tool between your own ads, not a pass/fail line.
• CPC (cost per link click). CTR's sibling. If one ad's clicks cost triple another's, the market has voted on the angle.
• Cost per purchase. The number that matters, judged against the breakeven figure you wrote down in Step 0 — not against anyone else's screenshot.
• Frequency. How many times the average person has seen the ad. When frequency climbs past roughly 2–3 on cold traffic and results decay, the creative is wearing out — fatigue, not failure.

Kill rules, hedged because every store's economics differ: if an ad has spent roughly 2–3× your breakeven cost per purchase with zero sales, pause it. If one ad's CTR is a fraction of its siblings' after a fair amount of spend, pause it — the angle did not land. Do not pause anything in the first few days on tiny numbers; small samples lie.

Then the loop that actually compounds: mark the winner and make more like it. Whichever angle won, produce new variations of that angle — new layout, new headline phrasing, same buying reason — and retire the losers. This is the rhythm Facebook ads reward: not one perfect ad, but a steady supply of fresh variants of whatever is working. (In Product AdKit this is literally a button — "make more like this" on the winning concept — but the loop is the same whatever tool you use.) The full system, with benchmarks and cadence, is in our ad creative testing guide.

Common mistakes that burn small budgets

• Calling one ad a "test." One ad tests nothing — if it fails you do not know whether the product, the angle, or the layout failed. Minimum viable test: 3 distinct angles.
• A lifestyle photo where the product is invisible. A gorgeous beach scene where your sunscreen tube is 40 pixels wide is a postcard, not an ad. The product must be legible at thumbnail size.
• Claims the product page cannot back. If the ad says "5,000 five-star reviews" and the page shows 12, you pay for the click and lose the sale — and you drift toward policy trouble. The ad may only promise what the landing page proves.
• Editing the campaign daily during learning. Budget nudges, audience tweaks, swapping creatives every morning — each edit can reset learning and burn your spend re-teaching Meta what it already knew. Decide, launch, wait.
• Stretching one layout to all placements. A 1:1 design center-cropped into 9:16 puts your headline half off-screen in Stories. Supply each aspect ratio its own version — this is exactly why Step 3's exit criteria included three sizes.

That is the whole playbook. The setup is a weekend; the photo and angles are an evening; the creatives are minutes to days depending on your route; and the loop in Step 5 is the part you keep. The founders who win at Shopify Facebook ads are rarely better media buyers — they just never run out of distinct creative to test.