CRO Audit · Page Performance

Page-Level Performance
What Works, What Doesn't

Where traffic goes, what it's worth, and the pages we need to fix or investigate next.
US Frame Factory|clips.usframefactory.com|Mar 5 – Jun 2, 2026 (90d)|GA4 · bot-filtered
Section 01

Traffic & Conversion by Category

Every landing page, grouped by what it sells. Across 30,466 sessions and $116.5k in tracked revenue, the store converts at 1.18% overall — but the average hides everything. Two categories carry the store; one large category quietly leaks a quarter of all traffic.
CategoryPagesSessions% of TrafficCVRRevenue% Rev
Pony Wall — PDPs127,861
25.8%
2.51%$48,68341.8%Engine
Truss Spacer — PDPs2699
2.3%
4.15%$8,7737.5%Best CVR
Homepage (/)12,103
6.9%
1.47%$15,19813.1%Soft CVR
TSN Clips & Connectors — PDPs4577,228
23.7%
0.76%$27,00723.2%Leak
Other Products — long tail1,8547,632
25.0%
0.30%$10,8449.3%Leak
Fasteners & Screws — PDPs2112,194
7.2%
0.77%$2,6792.3%Attach
Collections351,469
4.8%
0.34%$2,8622.5%Routing
Blog / Content10545
1.8%
0.00%$00.0%No handoff
All traffic, bot-filtered (China excluded). Cart/checkout re-entries (~$47.6k of returning-buyer revenue) are excluded so this reflects genuine acquisition only. CVR = GA4 session key-event rate; revenue is the hard number.
Section 02

Top Pages by Traffic

The ten pages pulling the most visitors. The healthy signal: the biggest traffic pages are mostly Pony Wall PDPs that also convert. The two that break the pattern (rows 7 & 10) are where traffic arrives and nothing happens.
#PageSessionsCVRRevenue
1Homepage /2,1031.47%$15,198Soft
2LGPW24 Pony Wall Lite 24"1,7392.82%$7,550Converts
3PW36 Pony Wall Heavy 36"1,5992.81%$12,750Converts
4PW24 Pony Wall Heavy 24"1,0652.72%$8,065Converts
5LGPW48 Pony Wall Lite 48"7712.59%$4,600Converts
6CDTBR24 Truss Spacer 24"6854.23%$8,773Converts
7CircleTrak 550CT125 (TSN)5070.00%$0Dead
8Blog: Wood-Stud Pony Wall Build Guide4960.60%$548No handoff
9LGPW36 Pony Wall Lite 36"4773.35%$4,220Converts
10Backyard Garden-House Framing Kit4100.00%$0Dead
Section 03

What's Working

The pages earning their traffic. These convert above the site average — the move here is simple: send them more traffic and budget.
Page / GroupSessionsCVRRevenueWhy it works
Pony Wall family (12 PDPs)7,8612.51%$48,683Converts 2× the site average with strong engagement. 42% of all revenue. The core of the business.
Truss Spacer (CDTBR24)6994.15%$8,773Best conversion rate on the store — and under-trafficked. Most upside if we feed it more visitors.
Pony Wall Supports (collection)3883.61%$3,216A collection page converting like a strong PDP. Proof that good category routing works.
Section 04

What's Not Working — & What to Test

Pages pulling real traffic and converting almost none of it. Low conversion on a product page traces to one of four things — and our job is to find which:
Traffic — wrong or low-intent visitors landing here
Price — not competitive enough to win the buy
Shipping — cost or speed kills low-ticket orders
Page — thin specs, missing info, weak trust
Page / GroupSessionsCVRThe ProblemWhat to Test / Figure Out
TSN Clips & Connectors (457 PDPs)7,2280.76%24% of all traffic across hundreds of thin, near-duplicate pages — a third of the site's conversion rate.Consolidate variants into parent products, then check all four: Traffic — is the feed sending broad/irrelevant clicks? Price — are we competitive on these commodity clips? Shipping — does cost sink these low-ticket carts? Page — enough specs/trust to close?
Long-tail products (1,854 PDPs)7,6320.30%A quarter of all traffic landing on pages that almost never convert and barely engage.Lead with Traffic (pull the source — low-intent feed vs. real intent), then Price and Shipping on the low order values these carry. If intent is real and price/ship are fine, it's the Page.
CircleTrak PDPs (550 / 362)7650.00%Heavy traffic, literally zero sales, low engagement.Traffic — likely a broad Shopping query mismatch. Also check the Page for missing specs/load data, and whether Price/Shipping on a niche bendable track is scaring buyers off.
Backyard Garden-House Kit4100.00%High traffic, zero revenue — looks like wrong-audience visitors (DIY dreamers, not contractors).Traffic — trace the keyword/source (likely a targeting cut). It's also a high-ticket Price sticker-shock candidate: test price framing, financing, or a "request a quote" path.
Homepage (/)2,1031.47%#1 traffic and holds visitors 62s — but converts at half the rate of the product pages it should funnel to.Page — A/B test the headline / value prop and a clearer path to the top Pony Wall products. Attention is there; the pitch isn't landing.
Build-guide blogs (2 posts)~900<0.8%Strong, engaged readership that never gets pointed at a product.Page — add in-content "shop the parts in this build" links/CTAs to the Pony Wall PDPs. Cheapest win on the list.
One caution before pruning: a few very-low-traffic B2B pages book large orders (e.g. VertiClip SL800 — 7 sessions → $4,386) — don't judge those on conversion rate; they're the large-buyer / PO path.
Section 05

Next Step — How We Diagnose & Fix

There are too many low-converting pages to review one by one. Instead we group them with a single rule, diagnose the group, then act. Four steps:
  1. Group every low-converting page into one view. One regex rule in a GA4 Exploration collapses the ~2,300 product pages — everything except the proven winners — into a single segment. Filter on Landing pagematches regex /products/, plus does not match regex pony-wall|truss-spacer|cdtbr. No setup, works on the data we already have.
  2. Review the traffic source. Break that group down by Session source / medium and default channel grouping, with engagement rate alongside. High traffic + near-zero engagement = a traffic-quality problem (broad Shopping/feed clicks). Real engagement but no sale = the problem is price, shipping, or the page.
  3. Run the four-lever check per product family. For each group, confirm which lever is failing — Traffic (qualified?), Price (competitive?), Shipping (cost/speed killing low-ticket carts?), Page (specs + trust to close?).
  4. Decide: fix or cut, then re-measure. Page / price / shipping problem → fix and re-test. Traffic-quality problem → tighten targeting, cut spend, or de-index. Re-pull this report in ~30 days to confirm the leak closed.