Most store owners I talk to are running Meta ads the way they learned in 2022. Twelve ad sets, tight interest stacks, one hero video, and a spreadsheet that stopped making sense around March. Then they wonder why the CPM keeps climbing while the revenue doesn’t.
The platform moved. The playbook didn’t. This guide is the version I’d hand a client on day one.
What “Meta ads” covers
Facebook and Instagram sit inside one ad system. When you publish a campaign, the same creative can land in the Facebook feed, Instagram feed, Reels, Stories, Marketplace and the Audience Network, and Meta decides the split based on where it thinks the conversion is cheapest.
So “should I run Facebook ads or Instagram ads” is mostly the wrong question. You run Meta ads. The placement mix sorts itself out, and for most ecommerce brands the majority of impressions end up on mobile, in vertical 9:16 formats.
The shift you need to understand first
Meta rebuilt its ad ranking system. The retrieval layer, which narrows tens of millions of eligible ads down to a few thousand candidates before the auction runs, now evaluates far more variants at once and leans hard on creative signals.
Two practical consequences:
Similar ads get collapsed. If you upload twenty variations of the same concept with a different caption, the system treats them as one entity and makes them compete against each other instead of expanding your reach. Meta now surfaces a Creative Similarity Score in Ads Manager, and once it climbs past roughly 60%, delivery suffers.
Fatigue arrives faster. Creative that used to last six weeks often burns out in two or three. Rising CPM against flat reach is usually the first symptom, not a broken account.
Advantage+ Shopping was renamed Advantage+ Sales in early 2025, and it now covers lead gen and app installs too, which is why half the guides online contradict each other on the name. The product is the same idea: one consolidated campaign, minimal manual controls, Meta handling placement, audience and budget allocation.
Across ecommerce accounts, Advantage+ campaigns are posting roughly 4.5 ROAS versus about 3.7 for comparable manual builds, with CPA somewhere between 17% and 32% lower depending on the vertical. Worth reading with one eye open, though. The algorithm happily takes credit for buyers who were already coming to you through brand search and retargeting pools. Turning Advantage+ on is not a strategy by itself.
A structure that actually works
Keep it boring. One Advantage+ sales campaign carrying most of your prospecting spend, plus one lean retargeting or catalogue campaign for people who viewed products and abandoned carts. That’s it for a store under $100k a month.
Fragmenting your budget across ten ad sets splits your conversion data ten ways, and the campaign needs around 50 purchase events per week before the algorithm optimises reliably. Below that threshold, you’re paying for guesses.
If you’re consolidating an old account, expect a rough fortnight while everything re-learns. Resist the urge to fiddle. Every budget change above 20% in a day risks resetting learning, which is exactly the kind of self-inflicted wound I spend the first month of a Meta ads management engagement undoing.
Creative is the whole job now
Targeting used to be the lever. It isn’t anymore. Meta’s own position is that creative quality accounts for more than half of ad performance, and the account data backs it up.
What that means in practice:
Run 15 to 20 genuinely different creatives at any time. Different hooks, different formats, different people, different use cases. Not twenty crops of one photo.
Prioritise vertical video. Static still works for catalogue and retargeting, but the inventory is vertical and the users are on phones.
Watch cost per thousand unique reach rather than plain CPM as your fatigue signal. When it climbs steadily, you’ve run out of responsive people for that concept.
Build a refresh cadence you can sustain. Six to ten concepts in rotation, refreshed monthly, beats a heroic quarterly launch that dies in week two.
One more thing: since March 2026, Meta requires disclosure on ads containing AI-generated or AI-modified content. Worth checking your creative process against that before it becomes a delivery problem.
Tracking, or none of this counts
Install the Conversions API alongside the pixel. Server side events fill the gaps that iOS and browser restrictions leave, and accounts with clean CAPI setups report meaningfully better returns than pixel-only ones.
Upload your customer list. First party data now beats third party interest targeting for feeding the algorithm.
For attribution, 7-day click plus 1-day view suits most products. If your average order value is above $200, consider 7-day click only so view-through conversions don’t flatter your numbers.
And judge the account on blended MER, not platform ROAS. Meta will always claim more than it deserves.
Benchmarks worth knowing
Ecommerce CPMs rose roughly 20% year over year, which is the single biggest driver of rising costs. CTR improved around 13% and conversion rates about 8%, so CPA stayed close to flat. Median CPM for US DTC sits near $14, and Q4 runs 25% to 60% above the annual average.
Treat these as a fence, not a target. Two brands at the same $45 AOV can post wildly different ROAS in the same week because one has eight winning creatives running and the other has two.
Where to start this week
Pull your last 90 days. Count how many distinct creative concepts actually spent money. If the answer is under five, that’s your bottleneck, and no amount of campaign restructuring will fix it.
Then check whether CAPI is live and deduplicating properly. Then consolidate. In that order.
Meta is usually the biggest line item, but it isn’t the whole picture. If you’re weighing it against TikTok, Pinterest or Snapchat, that’s a broader paid social advertising conversation, and the answer depends more on your margins than on the platform.
Frequently asked questions
How much should an ecommerce brand spend on Meta ads per month?
Start at roughly 30 times your target cost per purchase, so a $25 CPA needs about $750 a month at minimum. The real constraint is data volume: campaigns need around 50 purchases a week to optimise properly. Below that, spend more per campaign rather than spreading a small budget across several.
Are Meta ads still worth it for ecommerce in 2026?
Yes, but the margin for sloppiness has gone. CPMs are up around 20% year over year while CTR and conversion rates improved enough to keep CPA roughly flat. Brands with a real creative pipeline are doing fine. Brands running two ads and hoping are the ones declaring the channel dead.
What is a good ROAS for ecommerce on Meta ads?
Advantage+ campaigns average about 4.5 ROAS across ecommerce accounts, with manual builds nearer 3.7. Your break-even matters more than the benchmark though. A 2.0 ROAS is healthy at 70% gross margin and loss-making at 30%. Work backwards from contribution margin before you judge the number.
Should I run Facebook ads or Instagram ads for my store?
Run both through one campaign and let Meta allocate. They share the same ads manager and the same auction. Instagram Reels and Feed tend to lead on CTR for fashion, beauty and lifestyle, while Facebook often converts better for older demographics and higher ticket items.
How many creatives do I need for an Advantage+ campaign?
Between 15 and 20 active creatives, and they need to be meaningfully different. The ranking system collapses near-identical ads into a single entity, so twenty caption variants of one video perform like one ad. Vary the hook, format, talent and angle, not just the text overlay.
Why did my Meta ads suddenly stop working?
Check creative fatigue first. Cost per thousand unique reach climbing while your reach stays flat is the classic signature, and fatigue windows have shortened to two or three weeks. After that, look for recent budget jumps above 20% in a day, a broken pixel or catalogue feed, and seasonal CPM spikes.
