Decomposing AOV × Orders by Paid Channel to Spot the Drag
A practical walkthrough for splitting AOV × Orders across Meta, Google, TikTok, email and organic — so you can pinpoint exactly which channel is dragging blended revenue down.
Quick answer
Split revenue into AOV × Orders for each paid channel separately (Meta, Google, TikTok, email, organic). Compare each channel's AOV and order count week-over-week against its own baseline — the channel whose numbers moved most in the wrong direction is the drag. Blended AOV almost always hides this because high-AOV channels mask low-AOV ones.
Decomposing AOV × Orders by Paid Channel
Running AOV × Orders revenue decomposition separately per acquisition channel to surface which channel is dragging blended performance.
Channel-level decomposition takes the standard Revenue = AOV × Orders identity and runs it once per acquisition source rather than across the whole store. You end up with a small matrix: rows for Meta, Google, TikTok, email, organic, direct; columns for orders, AOV, and revenue, compared period-over-period.
The diagnostic value is in the deltas. A flat blended AOV can hide a TikTok channel where AOV dropped 22% while Google's AOV climbed enough to mask it. Channel decomposition is the bridge between revenue reporting and ROAS optimization — it tells you which channel deserves the budget conversation this week.
Blended numbers are useful for board decks and useless for decisions. If your storewide AOV is flat at €68 this month, that single line tells you nothing about whether to cut Meta spend or scale TikTok creative.
Channels behave differently because their audiences behave differently. Google Shopping pulls high-intent buyers searching for a specific SKU; Meta prospecting catches cold browsers on impulse; email reactivates existing customers who already trust your bundles. Three different AOVs, three different order frequencies, one blended average that smooths it all out.
Why blended AOV hides channel drag
Mix shift is the silent culprit. If TikTok grew from 8% of orders to 22% of orders this quarter and TikTok's AOV runs €15 below your store average, blended AOV will fall — even if every individual channel's AOV held steady.
The reverse trap is just as common. A Meta retargeting campaign with a €120 AOV can prop up the blended number while Google Shopping quietly slips from €85 to €71 because a competitor is undercutting your hero SKU. The blended view says "all fine"; the channel view says "Google needs attention now".
The Simpson's paradox trap
Every channel's AOV can rise week-over-week while blended AOV falls — if low-AOV channels grew their share of orders. Always inspect channel-level deltas AND channel mix before concluding anything from a blended number.
How to run the decomposition
Start with a clean attribution window. Pick one model and stick with it for the whole exercise — usually GA4's data-driven model or your platform's last-non-direct click. Mixing models across channels in the same view is how teams chase phantom drags.
Pull orders and revenue per channel for the current period and a matched comparison period (last 28 days vs the prior 28). Compute AOV as revenue ÷ orders for each channel, then calculate three deltas: ΔOrders, ΔAOV, and Δshare-of-orders. The last one is what catches mix shift.
Rank channels by absolute contribution to the revenue gap, not by percentage change. A 30% AOV drop on a channel doing 4% of orders matters less than a 6% AOV drop on a channel doing 40% of orders. The drag is whichever channel moved the blended needle most in euros.
Typical channel AOV and order patterns
Indicative AOV and order-share patterns by paid channel for a mid-market Shopify apparel store
| Channel | Typical AOV (€) | Share of orders | AOV volatility | Common drag signal |
|---|---|---|---|---|
| Google Shopping | 75-95 | 25-35% | Low | AOV slips when competitors discount hero SKU |
| Google Search (brand) | 85-110 | 10-15% | Very low | Order volume dip = brand demand softening |
| Meta prospecting | 55-70 | 20-30% | Medium | AOV drops as creative fatigues, orders flatten |
| Meta retargeting | 95-130 | 8-12% | Low | Audience exhaustion → orders fall, AOV stable |
| TikTok | 40-60 | 5-15% | High | Low-AOV impulse buys dilute blended AOV |
| Email/SMS | 85-120 | 15-25% | Low | Order count drops when send cadence slips |
| Organic/direct | 70-90 | 10-20% | Low | Steady; rarely the drag, often the cushion |
Treat the table as a shape, not a target. Your numbers will differ by vertical — beauty AOVs cluster tighter, electronics swing wider — but the relative ordering (TikTok low, retargeting high, Shopping in the middle) holds for most online retailers in the €1M-€15M range.
Reading the matrix: where the drag usually hides
Meta prospecting is the most common AOV drag. Creative fatigue pulls in lower-quality clicks, the audience skews toward single-item buyers, and AOV erodes by €5-€10 over 3-4 weeks before anyone notices. The order count often looks fine, which is why it goes unflagged in weekly stand-ups.
Google Shopping is the most common order-volume drag. A competitor entering the same product category, a Merchant Center disapproval, or a feed price change can knock 15-20% of orders out within a week. Because Shopping's AOV is steady, the symptom is a clean drop in orders only — easy to spot once you're looking per-channel, invisible in blended views.
What to do once you've found the drag
If the drag is AOV-driven, the fix lives on-site: bundle offers, threshold-based free shipping, cross-sell modules on the PDP and cart drawer. Channel-level AOV problems rarely get solved inside the ad platform — they get solved by raising the basket once the click lands. Tie any change to a clean A/B test so you can confirm the lift is real, not a mix-shift artefact.
If the drag is order-volume, the fix is upstream: budget reallocation, creative refresh, audience expansion, or feed/landing-page diagnostics for Shopping. This is also the point where channel ROAS becomes more useful than blended ROAS — once you know which channel moved, you need its isolated efficiency number to decide spend. See the companion piece on blended vs channel ROAS for the trade-off.
Frequently asked questions
Weekly during stable periods, daily during sales events or after a major creative or budget change. The diagnostic only works if you spot the drag within 7-10 days — beyond that, mix shift has already absorbed into your baseline and the signal fades.
Channel ROAS tells you the efficiency of spend on each channel. AOV × Orders decomposition tells you the mechanism behind a revenue change — whether basket size or order count moved. You use the decomposition to diagnose, then channel ROAS to decide where to reallocate budget.
Google captures buyers mid-search with explicit intent, often for higher-ticket items they've already researched. Meta catches browsers in-feed where the purchase is more impulsive and the typical basket is one item. The €15-€25 AOV gap between the two is structural, not a problem to fix.
Pick one and use it consistently across all channels in the same analysis. Data-driven is generally better for prospecting-heavy mixes; last-non-direct click is fine for retargeting-heavy mixes. Switching models mid-analysis is the single fastest way to draw the wrong conclusion.
No. A low-AOV channel only drags the blended number when its share of orders grows. If TikTok holds steady at 10% of orders, its low AOV is priced into your baseline. The drag signal is the delta in share-of-orders, not the absolute AOV level.
It can absolutely be the drag — usually on the order-count side when send cadence drops, deliverability degrades, or your list grows stale. Email's high AOV means a small order-count dip moves real revenue, so it's worth watching weekly even though it rarely surprises you.
Aggregate to a longer window (28 or 56 days) for low-volume channels and a shorter window (7-14 days) for high-volume ones. Mixing windows across channels in the same report is acceptable here — the goal is statistical stability per channel, not visual symmetry.
Yes, but with a caveat: a large share of direct traffic is mis-attributed paid or organic. Treat direct's AOV as a sanity check on the rest of your mix rather than as an independent acquisition source you can optimise.
GA4's Acquisition reports give you orders and revenue per source/medium — that's enough to compute AOV manually in a spreadsheet. If you want it without the spreadsheet, a CRO/analytics layer that joins GA4 with your store data can surface the matrix automatically each week.
Storewide AOV × Orders decomposition tells you whether revenue moved because baskets shrank or orders dropped. Channel-level decomposition tells you which channel caused it. Run the storewide version first to confirm there's a real change worth diagnosing, then drill into channels.
Track CAC, channels, and funnel conversion in one place
Metricuno connects ad spend, funnel events, and revenue so you can see CAC by channel, cohort, and campaign — without stitching together five tools.