By Steve Merrill, Founder of WRKNG Digital | June 15, 2026
Most Shopify brands overpay for Meta ads because of structural incentives, not incompetence. Your agency probably knows about these problems. They just don't get paid to fix them.
1. Agency Fees Are Tied to Ad Spend — So the Budget Never Shrinks
The standard agency model is a percentage of ad spend, typically 10–15%. That structure sounds neutral. It isn't. When your incremental ROAS drops below profitable thresholds — say, your last $5,000 of monthly budget is generating 0.8x return — your agency still collects a fee on that $5,000. There's no financial reward for telling you to cut. I ran Meta ads for my own brand and hit $10M in revenue. I know exactly what a performance conversation looks like when your agency is also watching their own invoice. If your agency has never recommended a budget reduction, ask why. The answer is usually obvious.
2. Billing for Audience Strategy That Meta's Algorithm Already Handles
Meta shifted to Advantage+ and broad targeting years ago. The algorithm now outperforms manual interest stacking in most accounts. But many agencies still invoice for audience strategy — building custom segments, stacking interest layers, refining demographic exclusions — that doesn't move the needle anymore. It's billable work. It looks like effort. It just doesn't perform. Ask your agency to show you a head-to-head test: their manually built audiences against Advantage+ with broad targeting. If they haven't run that test, you've been paying for something that was never validated. Meta's own guidance on Advantage+ Shopping campaigns makes their position clear — the platform wants broad inputs, not tight manual control.
3. Creative Volume Is Not a Testing Strategy
Agencies love to present creative output as a deliverable. "We launched 20 new creatives this month." That's a production metric, not a performance one. Real creative testing requires a hypothesis before you spend: what specific variable are you testing, what does winning look like, and what decision will the result drive? Most agencies batch-produce creatives with vague briefs and call variation a "test." It's not. When you run 20 ads without a structured framework, you get noise data — not signal. You can't make a confident call on what's working, which means you keep spending on everything instead of scaling what actually converts. Industry benchmarks from WordStream show that click-through rates vary dramatically by creative type — but only structured tests tell you why.
4. Retargeting People Who Already Bought
This one is embarrassingly common and almost never gets flagged. Poor exclusion audience management means you're serving retargeting ads to customers who converted last week. They see your ad. They think "I already bought that." You pay for the impression and the click. Nothing happens. Exclusion audiences — purchasers, email subscribers, recent converters — need active management. They go stale. They need refresh cadences. Most agencies set them up once at campaign launch and never revisit. Run a quick audit: pull your retargeting campaign audience and cross-reference against recent purchaser lists. The overlap will surprise you. Meta's exclusion audience documentation is a starting point, but you need someone actively working it — not setting it once.
5. Attribution Window Confusion That Inflates Reported Performance
Same campaign. Same spend. Two completely different ROAS numbers — depending on who's reading the report. Brands often measure on a 7-day click window, which is the Meta default for most Shopify integrations. Agencies frequently report on 28-day view-through attribution, which counts a conversion if a customer saw (not clicked) an ad anytime in the last month before buying. That's not a measurement error. It's a reporting choice that inflates the appearance of performance. If your agency's ROAS number never matches your Shopify Analytics number, attribution model mismatch is almost certainly why. Shopify's guide on Meta Ads integration covers how attribution is tracked natively — match that against what your agency reports. The gap tells you a lot.
How We Chose This List
These five issues came directly from auditing Meta ad accounts for Shopify brands — and from running Meta campaigns inside my own clothing business for years before building WRKNG Digital. They're structural, not accidental. And they're consistent across account sizes.
FAQ
- How do I know if my agency's fee structure is creating conflicts of interest?
- Ask them directly: have they ever recommended a budget cut? If they haven't in 12+ months of management, ask them to show you the ROAS curve as spend increases. If incremental return is declining and they haven't flagged it, you have your answer.
- Should Shopify brands stop using manual audience targeting entirely?
- Not necessarily. Manual audiences still have use cases in prospecting for niche products or entering new markets. But they should be validated against broad targeting, not assumed superior. If your agency hasn't run that test, push for it.
- What's the right attribution window for Meta ads?
- There's no universal answer, but the window needs to match how your customers actually buy. If your average purchase decision takes three days, a 7-day click window is probably fine. If your agency is using view-through attribution, get them to explain — specifically — why view-throughs should count for your product category.
- How often should exclusion audiences be refreshed?
- Purchaser exclusions should be updated at minimum every 30 days. If your account has high repeat purchase rates, exclude recent buyers at a tighter window — 14 days or less. Your agency should be doing this automatically. If you have to ask, they're not.
- What does a structured creative test actually look like?
- One variable per test. Clear hypothesis before launch (e.g., "UGC outperforms studio creative for our top SKU"). Defined sample size and time window. A decision rule before you start: what result changes your budget allocation? Everything else is just running ads and hoping.
If you want to see what AI-ready ecommerce infrastructure actually looks like — and whether your Shopify store is visible to the platforms driving product discovery in 2026 — start here: wrkngdigital.com/agentic-commerce-landing-page.
