TRAFFIC & SEASONALITY — INTERNAL
Growth · Traffic Sources + Multi-Year Context

Where traffic comes from —
and is this June normal?

Channel mix over 17 months, the organic-search erosion, the rise of AI traffic, and a two-year seasonality view of spend & CAC that answers whether the current squeeze is the calendar or something real.

The Answers
Organic
−17% YoY

Organic search is eroding — ~56k/mo (May '25) to ~46k (May '26). Steady bleed, not a cliff. Likely tied to the AI shift below.

AI traffic
~5× — but tiny

AI referrals (ChatGPT-led) grew ~412→2,100/mo YoY. Real trend, but still ~0.5% of traffic. It's likely cannibalizing organic search clicks, not adding net demand yet.

Is June normal?
Mostly seasonal

CAC peaks every May–Jun and troughs at BFCM. 2025 hit $95–98 in summer; 2026 is ~$101 — the same shape, running only modestly hotter.

01 — Traffic Mix

Sessions by channel, 17 months

Monthly sessions by GA4 channel. Paid social is the engine (and the BFCM/late-2025 cross-network surge stands out). June 2026 excluded (partial month). Source: GA4.

Channel mix — monthly sessions (k)

Paid social = purple · Direct = teal · Cross-network = amber · Paid search = pink · Email = green · Organic search = red.

Organic Search — the erosion

Down ~17% YoY (56k → 46k). A slow bleed across the period.

AI-referred traffic — rising fast, still small

ChatGPT + Gemini + Perplexity + Claude sessions/mo. ~5× YoY, but only ~2k/mo.

⚠ Organic search down + AI up = the same story

Organic search clicks are bleeding ~17% YoY while AI-assistant referrals are up ~5×. That's the industry pattern: people increasingly ask ChatGPT/Gemini instead of Googling, so traditional organic clicks shrink and a thin AI-referral stream replaces a sliver of them. Net: it's a channel shift, not necessarily lost demand — but it does mean SEO content should be optimized to be cited by AI answers, and the AI channel is worth tracking as it compounds.

02 — Could a competitor be hurting us?

What the owned data can and can't say

Signals we can see (mixed)

Against a competitor story: Direct traffic (a brand-demand proxy) is roughly flat YoY, and CTR on Meta actually improved — both argue brand pull is intact. Possibly consistent with one: Meta CPM rose ~55% into summer (more auction competition can do that), and organic search slipped. But CPM is also seasonal, and organic slipping fits the AI shift above.

What owned data can't answer: whether a specific competitor is outbidding you in the Meta auction or winning brand search. That needs external signals.

→ To actually answer the competitor question

  1. Google Trends (we have the connection): track "mud wtr" brand search interest vs competitor terms YoY — falling brand interest = a demand problem; flat = it's auction/efficiency.
  2. Meta Ad Library (scraper on the roadmap): pull competitor active-ad counts & spend proxies — a competitor flooding the auction is the most likely explanation for a CPM spike that isn't purely seasonal.
  3. Review/share-of-voice: Trustpilot/Amazon review velocity vs competitors for demand-shift signals.

Say the word and I'll pull Google Trends brand-vs-competitor and stand up the Meta Ad Library scan.

03 — Multi-Year Seasonality

Is the current CAC squeeze just the calendar?

Monthly blended CAC, spend, and new customers across ~19 months (Nov 2024 – May 2026, June partial excluded). Two full spring/summer cycles. Source: Daily Stand (Northbeam-fed, Clicks-Only).

Blended CAC — the seasonal shape

CAC troughs at BFCM (Nov '24 $69, Nov '25 $78) and peaks in late spring/summer (May–Jun). 2025 summer hit $95–98; 2026 May is $101 — same shape, ~$3–6 hotter. Target <$90 (dashed).

New customers / month

Peaks at BFCM; lowest in summer. The denominator of CAC.

Total ad spend / month ($k)

Fairly steady ~$1.4–1.9M/mo; Meta ~$0.8–1.0M of it.

✓ Answer: the June CAC squeeze is mostly seasonal — with a modest real overlay

CAC follows a clear annual rhythm: cheapest in Q4 (BFCM volume floods cheap new customers), most expensive May–August. Last year CAC ran $95–98 in May–June; 2026 is tracking ~$101 in May — the same seasonal peak, only ~3–6% hotter year-over-year. So most of "CAC is bad right now" is the calendar; the genuine YoY degradation is small. The bigger near-term swing is the deliberate Meta spend pullback in June, not a structural CAC blowout.

04 — Method

Sources: GA4 (property 310751693) monthly sessions by channel + AI-referral sources (chatgpt.com, gemini, perplexity, claude.ai, copilot; excludes bing.com) Jan 2025 – May 2026. Daily Stand (Northbeam Clicks-Only spend; Shopify new customers; blended CAC) Nov 2024 – May 2026. Partial current month (June 2026) excluded from all charts. Multi-year extends to ~Jun 2024 in monthly format; 2023 would require the older weekly-format tabs.

Related: D2C Complete Overview · 6-Month Pulse · YoY Benchmark.