← catspring-tea

Last 28 Days — All Traffic

GA4 property 321793786, all channels. Last 28 days (Mar 31 - Apr 27) vs previous 28 days (Mar 3 - Mar 30).

Sessions down 11% and conversions down 25%, so conversion rate slipped from 1.49% to 1.26%. The biggest channel mover is Paid Search, which lost 44% of its sessions (2,670 → 1,504). Direct stayed steady around 2,200. Email actually grew 59% off a low base. Two outliers in the channel data are worth a flag: Cross-network jumped from 30 to 730 sessions but with only 49% engagement and 21 conversions, and Paid Search lost half its traffic without a corresponding spend cut on the Google Ads side.

The shape mostly tracks the previous period one band lower, but Apr 17-20 is a noticeable trough (103, 103, 216, 143) compared to ~250/day in March. The Apr 22 spike to 408 is the standout — worth checking what content went out or what referrer drove it.

Other key metrics

Engagement rate is essentially flat. Time on site is slightly up. The conversion drop isn't a quality-of-traffic story — visitors are engaging similarly, fewer of them are converting. That points at either a checkout/storefront issue or a channel-mix shift toward less commercially-intent traffic.

Channel breakdown

Three things stand out:

1. Paid Search lost 1,166 sessions (-44%). That's the biggest absolute drop. Cross-check with the Google Ads report — clicks were only down 4% but GA4 says sessions are down 44%. Likely a UTM/auto-tagging/attribution mismatch worth diagnosing.
2. Cross-network jumped from 30 to 730 sessions with 21 conversions. Performance Max attribution probably reshuffled and some Paid Search is now bucketing as Cross-network. That would mean the headline "Paid Search is down 44%" is partly a labelling shift, not a real drop.
3. Direct is the largest channel and held flat. That's encouraging — brand and returning customer traffic is steady.
4. Organic Social fell off. 95 → 29 sessions. Worth checking if Instagram/TikTok activity changed.

Top landing pages

A few patterns to call out:

1. The lymphatic drainage blog post is the second-biggest landing page (565 sessions) but it converts almost nothing — 1 conversion this period, 1 last period. That post is huge for awareness but converts 0.18%. Either the on-page CTAs aren't working, or the audience isn't ready to buy. Worth A/B testing a sticky product offer or starter-pack callout.
2. Product PDPs are mostly up: Marfa Dark Roast (74 → 108, +46%), Pedernales Green (61 → 99, +62%), Variety Pack (101 → 123, +22%), Best Sellers (79 → 103, +30%). The merchandising is working — more people are landing on commerce pages.
3. Home page is essentially flat (1,222 → 1,190). Conversions on home dropped from 41 to 35 — the conversion rate from home dipped slightly.

Monthly trend

Jan 2026 (64,833 sessions) is a massive outlier — roughly 8× a normal month. Almost certainly a press hit, viral social moment, or referral surge. Worth identifying and trying to reproduce. Outside of January, traffic has been settling in the 5K-17K range. April 2026 is partial through Apr 27 and tracking similar to March.

What to do next

1. Diagnose Paid Search vs Cross-network attribution shift. Google Ads says clicks are flat but GA4 says Paid Search sessions are down 44% while Cross-network jumped from 30 to 730. Almost certainly a Performance Max re-labelling. Confirm in Google Ads → check campaign network mix.
2. Audit the lymphatic drainage post for conversion paths. 565 sessions / 1 conversion is a huge missed opportunity. Add a sticky product card, or run a free shipping bar specifically for that landing page.
3. Find what drove January 2026's 64K sessions and try to replicate it. That single month is doing more than the rest of the year combined.
4. Investigate the Apr 22 spike (408 sessions). Probably an email send or a press mention. If we can identify it, we can plan more of them.
5. Email is small but growing (179 → 285). Worth doubling down — same conversion rate as Paid Search at near-zero cost.