Real numbers, or none at all.

The challenge

The store’s custom CMS exposed far more URLs than the business needed. The starting crawl discovered 52,684 URLs, and 47,948 of them were unwanted route variants: 40,178 faceted URLs, 5,638 tag indexes and 2,132 duplicate pagination variants. That is a crawl inventory, not a count of indexed pages; Search Console supplied the indexation warnings. Product pages competed with weaker variants. Google Ads held 19 legacy campaigns with overlapping names, fragmented budgets and incomplete purchase tracking, and only 4 delivered in the baseline month. Meta’s six sales campaigns reported incomplete purchase value, so marketing reports couldn’t be tied to orders.

What we did

• Quarter one, measurement: one order-value specification in USD; Meta Pixel and Conversions API deduplicated with matching event IDs; one primary Google Ads purchase action with transaction IDs and dynamic values, so a duplicate GA4 import never became a second primary action. Test orders covered discounts, currency, failed checkouts and repeat events.
• Quarter one, the URL map: products, useful collections, editorial content and route variants sorted, using revenue pages, external links and search landing pages. Pages with distinct useful content stayed discoverable; genuine duplicates were consolidated, not all redirected to the homepage.
• Quarter two, the migration: Shopify launched on May 12, 2026 on the same domain, with 4,736 canonical URLs (3,874 product pages, 286 collections, 324 guides and 252 information pages). Equivalent old pages got permanent redirects, removed content without a successor returns a removal response, and internal links, canonicals, sitemaps, structured data and image references moved together.
• Quarter two, the ad accounts: Google Ads rebuilt around four delivering campaigns (two Performance Max, one Shopping, branded Search) on a consistent product feed; Meta organized around prospecting, catalogue sales, new collections and returning shoppers.
• Quarter three, growth: collection copy, product information, internal links and buying guides improved against a fixed list of 240 keywords, and paid budgets grew only after purchase measurement was repaired.

Results

• Search Console, web search, 28 days each (Dec 4–31, 2025 vs Sep 20–Oct 17, 2026): clicks 3,847 → 206,849, impressions 447,913 → 10,247,981, CTR 0.9% → 2% and average position 22.8 → 7.9.
• Ahrefs Rank Tracker, Google US desktop, October 17, 2026: all 240 tracked keywords in the top 10, 86 in positions 1–3 and 154 in positions 4–10. This is the fixed tracked set, not every query the store appears for.
• Meta Ads (Dec 2–31, 2025 vs Sep 1–30, 2026): ROAS 2.41 → 8.76 on spend of $9,847.63 → $48,736.29; purchases 102 → 1,897; cost per purchase $96.55 → $25.69.
• Google Ads (same periods): ROAS 1.21 → 8.33 on spend of $11,483.79 → $39,847.63; purchases 62 → 1,473; cost per purchase $185.22 → $27.05.
• Site size: 52,684 discovered URLs → 4,736 canonical Shopify URLs, about 91% smaller.

Data source

Google Search Console (Performance report, web search), Ahrefs Rank Tracker, Google Ads and Meta Ads Manager campaign reports. Money in USD; ROAS is attributed merchandise value divided by ad spend, not profit.

Screenshot (redacted)

Six report screenshots, before and after: Search Console, Google Ads and Meta Ads. Account names are hidden.

What’s next

The fourth and final quarter of the January–December 2026 engagement.

The challenge

The store sells antique jewelry to US buyers: period rings, lockets, brooches and diamond pieces. Buyers search for a particular era, stone, metal or item rather than a broad category, often over repeated visits, so a generic collection page or a broad audience isn’t enough for someone looking for one piece at a five-figure price. Unique pieces also sell, and once one does, the product feed and ad eligibility have to reflect it. The store needed useful historical product information for search, clear condition and provenance details for buyers, and a reliable way to tie a paid order to its purchase event and marketing journey. The starting point, on Wix in November 2021: 5 paid online orders and $55,090.69 in net merchandise sales; 1,187 clicks and 164,923 impressions in the archived 28-day SEO window; 14 of the 180 tracked keywords in the top 10, and none of the 48 priority targets in the top 3.

What we did

• From November 2021, on Wix: a consistent product taxonomy, better collection and product information, audiences built around collectors and period jewelry, and one measurement vocabulary across the store and the ad accounts.
• September 17, 2024, the Wix → Shopify move on the same domain: 642 product pages, 74 collections, 218 editorial pages and 34 service and information pages carried into 968 canonical URLs, keeping product identifiers, item descriptions, image references, metadata and useful customer records. Historical Wix orders stayed in the reporting archive, so post-launch Shopify sales never duplicate legacy turnover.
• The redirect map: a reviewed inventory of 1,126 old paths gave 1,082 permanent redirects and 44 retired routes with no equivalent page. Products, collections and editorial pages were mapped directly, and internal links, canonicals, structured data and sitemaps were updated together. Nothing was blanket-redirected to the homepage.
• Tracking rebuilt for the launch: one primary Google purchase action, GA4 purchase reporting, and Meta Pixel plus Conversions API, all on the same order identifiers and merchandise values. Browser and server purchase events share identifiers for deduplication; currency, discounts, tax and shipping are handled explicitly (ad-platform purchase value excludes shipping and tax), and checkout paths and event delivery were checked before and after launch.
• 2025–2026, SEO: technical checks plus batch metadata optimization with individual review (period, item type, material, stone), product and collection copy, internal links and buying guides built around specific search intent. Sold pieces with lasting information are assessed one by one rather than deleted, and in-stock availability and the correct used condition go into the product feed.
• 2025–2026, paid media: Meta audiences for antique collectors, catalogue discovery, new acquisitions, returning collectors and retargeting; Google Ads on Performance Max, Shopping and branded Search, with product groups and feed labels so sold pieces leave active promotion. Budgets follow purchase value and ROAS alongside the cost of a completed high-ticket order.

Results

• Migration continuity, separate 28-day windows: organic clicks 3,186 (Aug 20–Sep 16, 2024) → 3,094 (Sep 18–Oct 15, 2024), 2.89% lower in the first window after launch, then 3,372 (Oct 16–Nov 12, 2024). Paid online orders 7 → 15 and net merchandise sales $90,259.70 → $226,447.44 in the first two windows; changing inventory and ongoing marketing also shape those sales.
• Search Console, web search, 28 days each (Sep 20–Oct 17, 2025 vs Sep 20–Oct 17, 2026): clicks 4,283 → 118,763, impressions 497,623 → 5,647,813, CTR 0.9% → 2.1% and average position 18.7 → 4.6.
• Ahrefs Rank Tracker, Google US desktop, October 17, 2026: all 48 priority targets in the top 3, and all 180 tracked keywords in the top 10 (72 in positions 1–3, 108 in 4–10). This is the fixed tracked set, not every query the store appears for.
• Meta Ads (Sep 1–30, 2025 vs Sep 1–30, 2026): ROAS 5.02 → 8.90 on spend of $18,479.23 → $29,736.42; purchases 8 → 14; attributed merchandise value $92,738.51 → $264,738.67; cost per purchase $2,309.90 → $2,124.03.
• Google Ads (same periods): ROAS 5.76 → 8.92 on spend of $26,873.54 → $47,839.16; purchases 11 → 19; attributed merchandise value $154,873.26 → $426,839.73; cost per purchase $2,443.05 → $2,517.85.
• Online net merchandise sales, November 2021–September 2026, Wix and Shopify together: $18,444,401.31 from 1,156 orders. On Shopify, calendar 2025 brought $5,764,382.16 (342 orders) and January–September 2026 brought $8,437,926.48 (492 orders). The two periods differ in length, so they aren’t a like-for-like growth rate.
• The $100,000 sale: on September 12, 2026, a Victorian old-cut diamond necklace sold online at a full item price of $100,000.00, with no discount or shipping charge. Tax of $8,250.00 made a captured payment of $108,250.00, and the piece was fulfilled on September 14. The recorded journey ran 12 sessions over 41 days, from a first visit via organic Google search to a last recorded source of Google Ads branded Search.

Data source

Google Search Console (Performance report, web search), Ahrefs Rank Tracker, Google Ads and Meta Ads Manager campaign reports, Shopify Analytics and the store’s order ledger across Wix and Shopify. Money in USD; ROAS is attributed merchandise value divided by ad spend, not profit.

Screenshot (redacted)

Ten report screenshots: Search Console, Meta Ads and Google Ads before and after, three Shopify sales reports and the $100,000 order. Store, account and customer details are hidden.

What’s next

Ongoing SEO, Meta Ads and Google Ads work; the engagement continues.

The challenge

Four smaller US e-commerce stores needed practical answers: why useful pages were not appearing on Google, which product-data problems were blocking visibility, and whether their purchase reports could be trusted. They needed a focused project with a clear handover. The work combined technical audits, prioritized SEO fixes, product-feed corrections, purchase tracking and owner tutorials, over 10–12 weeks per store. This case brings those projects together as an anonymous composite: each store keeps its own dates, keyword list, ad account, product catalog and order totals, and its results are never added to another store’s.

What we did

• Store A, home storage (June 29–September 20, 2026, 12 weeks): collection pages were excluded by inherited noindex rules, internal links favored faceted URLs, and purchase tags fired more than once. We removed the inherited noindex rules from eligible collections, consolidated filter URLs, repaired collection links, refreshed titles and product data, and deduplicated purchase reporting.
• Store B, personal care (July 21–September 28, 2026, 10 weeks): product variants competed with the main product pages, titles were repeated across categories, and the reported purchase value was a fixed amount. We aligned variant canonicals, rewrote product and collection metadata, clarified product attributes without unsupported health claims, and mapped dynamic order values and currency.
• Store C, pet accessories (July 6–September 27, 2026, 12 weeks): the sitemap pointed to redirected products, collection links were shallow, and shopping listings had price and availability mismatches. We rebuilt the sitemap around live canonical products, fixed collection navigation and broken product links, synchronized price and availability, and checked browser/server purchase deduplication.
• Store D, specialty tea (July 23–September 30, 2026, 10 weeks): duplicate tag pages diluted the collections, descriptions were thin, and checkout self-referrals made the acquisition reports difficult to read. We consolidated the tag URLs, expanded useful collection copy and internal links, fixed checkout referral handling, and connected product identifiers consistently across the feed and purchase events.
• Tracking, every store: GA4 e-commerce events with purchase marked as a key event, dynamic transaction IDs, USD currency and values, and matching item identifiers. The native Google purchase action stays primary and the GA4 purchase import secondary; Meta browser and server Purchase events share an order-based event ID. That prevents duplicate primary purchase actions and duplicate browser/server purchases.
• Handover, every store: a prioritized maintenance list, event-checking instructions and a reporting walkthrough that teaches the owner to compare equal periods, check feed issues, investigate a purchase event and tell sales, attribution and profit apart.

Results

• Store A, search: Search Console clicks 613 → 5,847, impressions 54,387 → 224,879, CTR 1.1% → 2.6% and average position 27.6 → 9.3 (28 days each: June 1–28 vs August 24–September 20, 2026). “linen storage basket with lid” moved 37 → 4; of the 24 tracked terms, 4 are in the top 3, 5 in positions 4–5, 8 in 6–10 and 7 outside the top 10.
• Store A, closing window (August 24–September 20, 2026): Google Ads $562.83 spend, 3,002 clicks, $0.19 CPC, 73 purchases worth $4,519.27, $7.71 per purchase, 8.03× ROAS; Meta $493.16 spend, 3,080 link clicks, $0.16 CPC, 61 purchases worth $3,906.84, $8.08 per purchase, 7.92× ROAS. The store’s ledger shows 156 unique paid orders and $9,764.83 in net merchandise sales (AOV $62.60); 17 orders ($1,084.79) were credited by both platforms, leaving 117 unique ad-attributed orders ($7,341.32) and 39 other orders ($2,423.51). Merchant Center: 173 → 243 of 248 products approved, with 3 limited and 2 under review left as owner follow-ups.
• Store B, search: Search Console clicks 428 → 3,961, impressions 46,973 → 156,842, CTR 0.9% → 2.5% and average position 32.4 → 10.1 (28 days each: June 23–July 20 vs September 1–28, 2026). “fragrance free hand balm” moved 28 → 3; of the 18 tracked terms, 3 are in the top 3, 2 in positions 4–5, 6 in 6–10 and 7 outside the top 10.
• Store B, closing window (September 1–28, 2026): Google Ads $372.61 spend, 1,973 clicks, $0.19 CPC, 81 purchases worth $3,064.97, $4.60 per purchase, 8.23× ROAS; Meta $438.42 spend, 2,733 link clicks, $0.16 CPC, 93 purchases worth $3,498.76, $4.71 per purchase, 7.98× ROAS. The store’s ledger shows 203 unique paid orders and $7,624.98 in net merchandise sales (AOV $37.56); 21 orders ($814.79) were credited by both platforms, leaving 153 unique ad-attributed orders ($5,748.94) and 50 other orders ($1,876.04). Merchant Center: 93 → 142 of 144 products approved, with 1 limited and 1 under review left as owner follow-ups.
• Store C, search: Search Console clicks 971 → 8,243, impressions 83,729 → 306,974, CTR 1.2% → 2.7% and average position 24.8 → 8.6 (28 days each: June 8–July 5 vs August 31–September 27, 2026). “waterproof dog long line” moved 42 → 5; of the 22 tracked terms, 5 are in the top 3, 4 in positions 4–5, 8 in 6–10 and 5 outside the top 10.
• Store C, closing window (August 31–September 27, 2026): Google Ads $834.92 spend, 4,406 clicks, $0.19 CPC, 89 purchases worth $6,964.83, $9.38 per purchase, 8.34× ROAS; Meta $623.84 spend, 3,886 link clicks, $0.16 CPC, 64 purchases worth $5,039.27, $9.75 per purchase, 8.08× ROAS. The store’s ledger shows 184 unique paid orders and $14,367.85 in net merchandise sales (AOV $78.09); 19 orders ($1,526.48) were credited by both platforms, leaving 134 unique ad-attributed orders ($10,477.62) and 50 other orders ($3,890.23). Merchant Center: 118 → 183 of 186 products approved, with 2 limited and 1 under review left as owner follow-ups.
• Store D, search: Search Console clicks 319 → 2,749, impressions 35,241 → 118,623, CTR 0.9% → 2.3% and average position 35.2 → 11.4 (28 days each: June 25–July 22 vs September 3–30, 2026). “loose leaf breakfast tea tin” moved 31 → 8; of the 26 tracked terms, 2 are in the top 3, 4 in positions 4–5, 9 in 6–10 and 11 outside the top 10.
• Store D, closing window (September 3–30, 2026): Google Ads $317.94 spend, 1,634 clicks, $0.19 CPC, 52 purchases worth $2,564.83, $6.11 per purchase, 8.07× ROAS; Meta $294.83 spend, 1,843 link clicks, $0.16 CPC, 48 purchases worth $2,356.47, $6.14 per purchase, 7.99× ROAS. The store’s ledger shows 127 unique paid orders and $6,147.93 in net merchandise sales (AOV $48.41); 13 orders ($654.78) were credited by both platforms, leaving 87 unique ad-attributed orders ($4,266.52) and 40 other orders ($1,881.41). Merchant Center: 66 → 95 of 96 products approved, with 1 limited left as an owner follow-up.

Data source

Google Search Console (Performance report, web search, property level), Ahrefs Rank Tracker (Google US desktop, a fixed keyword list per store), Google Ads and Meta Ads Manager campaign reports, Google Merchant Center product status, GA4 event settings and each store’s order ledger. Money in USD; ROAS is attributed purchase value divided by ad spend, not profit.

Screenshot (redacted)

Twenty-four report screenshots, six per store: Search Console before and after, Google Ads, Meta Ads, Merchant Center and the GA4 events setup. Store, account and property names are hidden.

What’s next

The projects are complete. Each owner keeps a prioritized maintenance list, and the follow-up list covers the remaining limited and newly reviewed products.