Arc'teryx Cormac LS
Nightscape Heather · $140
View productCommerce journey audit · 28 August 2026
Blue shirt purchase journey
Three shirts. One cart. Checkout to payment. The journey completed cleanly, but high-intent shoppers are being sent in the wrong direction before they buy.
The basket
Three genuinely blue shirts
Cotopaxi Pulso Tech
Steel Blue · $179.99
View productOutdoor Research Echo
Naval Blue · $69.99
View productStart here
Three changes with the clearest upside
Ranked by expected business impact against implementation effort.
01 Confirmed
Teach the site what blue means
“Blue shirt” puts a sage green Patagonia top first because the manufacturer colour contains the word blue. The same data rule puts green products inside the Blue category filter.
- Evidence
- 70 of 210 men’s tops are tagged Blue; the first search result is green.
- Business outcome
- Better conversion on high-intent colour searches and fewer colour-related returns.
- Effort
- Medium · merchandising data with engineering support
02 Confirmed
Give filtered views a real address
Colour selections live after a hash in the address. Search engines cannot index that state and shared links expose an unreadable technical string.
- Evidence
- The Men’s Tops → Blue state is a client-side fragment, not a crawlable page.
- Business outcome
- Organic visibility for high-intent terms such as “men’s blue shirts NZ”.
- Effort
- Medium–high · engineering
03 Confirmed
Fix shared-link titles everywhere
Double-escaped social titles publish raw web code such as & and ' when products are shared. One theme-tag correction improves every organic share and paid social advert.
- Evidence
- The home page and Cotopaxi shirt both publish double-escaped social metadata.
- Business outcome
- Cleaner, more trustworthy previews across Facebook, Instagram, WhatsApp and Slack.
- Effort
- Low · theme sharing-tag correction
Complete opportunity map
All 18 findings, ranked
Confidence distinguishes repeated observed facts from a single timed observation.
| Rank | Area | Observation | Impact | Effort | Confidence |
|---|---|---|---|---|---|
| 01 | Search & filter | Green garments are returned as blue because vendor colour names drive shopper-facing colour. | High | Medium | Confirmed |
| 02 | Findability | Filtered collection states use hash fragments, so they are neither crawlable nor cleanly shareable. | High | Med–high | Confirmed |
| 03 | Social sharing | Double-escaped social titles expose raw web code in shared links and paid social previews. | High | Low | Confirmed |
| 04 | Cart | No recommendations appear in the cart drawer or cart page. | High | Low | Confirmed |
| 05 | Tracking | Meta Pixel reports conflicting versions on a live product page. | High | Low | Confirmed |
| 06 | Reliability | One product page produced 8 JavaScript errors and 7 warnings, including failed wishlist and monitoring calls. | High | Low–med | Confirmed |
| 07 | Catalogue | Different colours of the same shirt are separate, unlinked products. | Med–high | Medium | Confirmed |
| 08 | Speed | Men’s Tops made 433 requests to 37 outside hosts and completed at 8.6 seconds on a warm cache. | Med–high | Med–high | Confirmed |
| 09 | Trust | Only 2 of 5 sampled products had any reviews. | Medium | Low | Confirmed |
| 10 | Search relevance | “Blue shirt” returned 21 results, including 4 women’s products; the first result was women’s. | Medium | Medium | Confirmed |
| 11 | Checkout | Address suggestions still triggered an incorrect “add a house number” warning. | Medium | Low | Confirmed |
| 12 | Checkout | The signature choice has no default and sits beneath roughly 90 words of liability copy. | Medium | Low | Confirmed |
| 13 | Filters | Colour is the sixth filter group, inside a separately scrolling panel. | Medium | Low | Confirmed |
| 14 | Page structure | Pages without a main heading include Home, Brands and Contact. Screen readers and search engines receive no single page topic; 84 KB of tracking script before the title also hides it from short-read tools. | Medium | Low | Confirmed |
| 15 | Mobile | The first product begins 592px down an 874px screen: 68% of the first view is interface chrome. | Medium | Low–med | Confirmed |
| 16 | Product page | Brand colour names lack a plain shopper-facing colour label. | Low–med | Low | Confirmed |
| 17 | Product page | The breadcrumb displays a literal ellipsis instead of its middle step. | Low | Low | Confirmed |
| 18 | Search | Search-dropdown images appeared blank, then loaded after roughly 4 seconds in one run. | Low–med | Medium | Likely |
Measured evidence
The page is carrying too much kit
433requests on Men’s Tops
37outside hosts contacted
8.6sfully loaded, warm cache
111scripts on one product page
These are observed browser measurements, not synthetic lab scores. The audit did not estimate conversion or revenue impact.
Protect the good parts
What already works
- Guest checkout is clear. Signing in is optional rather than a wall.
- Delivery messaging is consistent. The $150 threshold appears throughout the journey.
- Every add is confirmed. The drawer shows colour, size and price immediately.
- Address suggestions save effort. Suburb, town, region and postcode filled from one choice.
- The payment promise is kept. Afterpay is advertised and offered at payment.
- Store stock is useful. Click and collect shows availability across 9 stores.
Method & boundaries
A real journey, stopped safely
An anonymous desktop shopper used both navigation and search, selected three separate products in size Small, and proceeded through guest checkout to the payment step. No card details were entered and no order was placed. Mobile was spot-checked at 402 × 874. Signed-in accounts, click and collect, returns and post-payment screens were out of scope.