Six months, every metric,
week by week

Funnel (GA4), acquisition spend & CAC (Northbeam / Daily Spend Tracker), and the traffic mix — each on its own weekly chart. The through-line: the recent decline is an acquisition story, not a funnel one.

The Story in Three

Why traffic dropped

Meta pullback

Meta spend held ~$245k/wk Feb–May, then fell to ~$177k the first week of June. Paid-social sessions dropped in lockstep (~92k → ~67k). The traffic decline is mostly fewer Meta dollars.

And it got pricier

CAC $92 → $117

Blended CAC drifted up all spring and jumped to ~$117 by early June (target <$90). Fewer dollars buying costlier customers — new customers fell ~4,300/wk → ~2,600.

Funnel is steady

CVR ~2%, seasonal

Conversion and add-to-cart have held their range all six months; the June dip is the normal summer slide. The funnel isn't the problem — acquisition volume and cost are.

01 — Acquisition

Spend, cost, and customers

Spend & CAC series start the week of Feb 9 (when the Daily Spend Tracker's per-channel history begins). Northbeam attribution: Clicks-Only.

Total ad spend

all channels, $/week

Meta spend

Facebook/Instagram, $/week

Blended CAC

D2C spend ÷ Shopify new customers · target <$90

New customers (D2C)

Shopify, per week

02 — Traffic

Volume and mix

Sessions overall, the Meta-driven paid-social line, and the full channel mix. Paid social is the swing factor.

Total sessions

all traffic, per week

Paid-social sessions

Meta-driven, per week

Sessions by channel

paid social · paid search · direct · organic · email

03 — Funnel

Conversion, cart, order value

The on-site funnel (GA4). Steady-to-seasonal — confirming the decline lives upstream in acquisition, not here.

Conversion rate

orders ÷ sessions

Add-to-cart rate

% of sessions

Orders

per week (GA4 transactions)

AOV

revenue ÷ orders

04 — Reading It

→ The acquisition flywheel slowed; the site held

Lay the charts side by side and the causal chain is clear: Meta spend pulled back → paid-social sessions fell → total sessions fell → with CAC already climbing ($92→$117), new customers dropped from ~4,300/wk to ~2,600. Meanwhile CVR, add-to-cart, and AOV stayed in their normal ranges. So the six-month decline is an acquisition-efficiency and spend-level story, not a conversion problem.

The lone on-site exception worth a glance: the AOV dip the week of 5/25 (~$36 vs ~$48 norm) — a discount/promo week, not a trend.

→ Where to push

CAC is the lever. It broke the <$90 target in March and hit ~$117 by June. Until CAC comes down, adding Meta spend just buys expensive customers. Diagnose the CAC climb (creative fatigue, audience saturation, auction costs) before scaling spend back.
The June Meta pullback explains the traffic break we flagged on May 29 — confirm it was intentional (budget decision) vs. a delivery/tracking issue.
Protect the funnel. It's healthy; the separate ~Jun 4 checkout-completion dip (Shopify) is the one on-site item still worth fixing.

05 — Method

Sources: Funnel + traffic = GA4 (property 310751693), weekly Dec 15 – Jun 1. Spend = Daily Spend Tracker spend_long (Northbeam-fed, Clicks-Only), weekly from Feb 9. New customers = Shopify D2C (Daily Spend Tracker customers_daily). Blended CAC = total D2C spend ÷ Shopify new customers (per standard — not Northbeam's attributed counts). 6-month platform composition: Meta 68% of spend ($5.9M of $8.6M). Current partial week (Jun 2+) excluded so no chart ends on a false cliff.

Related: 2026 vs 2025 Benchmark · Shopify checkout break · Funnel Steps W/W.