Signal Attribution & Tracking agent
Signal monitors the integrity of every tracking signal that feeds your attribution — UTMs, click IDs, ecommerce events, GTM tag health and consent state.
Monitoring scope
UTM and click ID coverage across paid and organic Source / medium classification and channel grouping Ecommerce event reliability (view_item, add_to_cart, purchase) GTM tag health and firing consistency Consent state visibility and cookie compliance Broken or duplicated tracking journeys Example finding
Missing UTMs on paid social
18% of paid social sessions arrived without UTM parameters in the last 7 days.
→ Becomes an Implementation Pack via Forge.
Ledger Revenue Reconciliation agent
Ledger reconciles reported revenue across your store, analytics and ad platforms — surfacing variance, overclaim and silent under-reporting.
Monitoring scope
Shopify, GA4, Google Ads and Meta revenue comparison Transaction ID deduplication and coverage Refund, discount and gift card treatment Currency and timezone reconciliation Daily revenue variance vs tolerance Platform-level overclaim of conversions Example finding
Revenue mismatch · Shopify vs GA4
Shopify revenue is 24% higher than GA4 purchase revenue over the last 7 days.
→ Becomes an Implementation Pack via Forge.
Pulse Funnel & Checkout agent
Pulse tracks the journey from landing to purchase, surfacing where customers silently drop off before revenue is captured.
Monitoring scope
Landing → product → cart → checkout → purchase Mobile vs desktop drop-off by step Add-to-cart, shipping and payment friction Checkout abandonment and recovery signal Page weight and Core Web Vitals on critical steps Error and dead-click hotspots Example finding
Mobile checkout drop-off
Mobile checkout drop-off is 18% higher than desktop on the shipping step.
→ Becomes an Implementation Pack via Forge.
Vector Paid Media Efficiency agent
Vector evaluates paid media spend against true store revenue and the quality of the conversion signal feeding optimisation.
Monitoring scope
Spend vs store revenue by channel ROAS confidence and platform overclaim Conversion action quality and signal strength Smart Bidding and Advantage+ setup hygiene Audience and creative waste signals Wasted spend on out-of-stock or low-margin SKUs Example finding
Bidding optimising to weak signal
Google Ads Smart Bidding is optimising toward a low-confidence conversion action.
→ Becomes an Implementation Pack via Forge.
Scout CRO Opportunity agent
Scout finds high-impact conversion opportunities across product, landing and category pages.
Monitoring scope
High-traffic, low-conversion page outliers Weak product detail pages and CTAs Trust signals and social proof gaps Landing page friction and intent mismatch Element-level interaction patterns Prioritised experiment ideas with expected lift Example finding
High-traffic PDP underperforms
Top-traffic PDP converts 41% below site average despite strong intent.
→ Becomes an Implementation Pack via Forge.
Merchant Product & Catalogue agent
Merchant analyses product and collection performance, inventory exposure and discount dependency.
Monitoring scope
Top revenue-driving products and collections Underperforming SKUs and dead inventory Bestseller out-of-stock with active paid traffic Discount dependency by product Margin and contribution signal Catalogue-level conversion patterns Example finding
Bestseller out of stock with active spend
Active Google Ads spend is driving traffic to an out-of-stock bestseller.
→ Becomes an Implementation Pack via Forge.
Loop Retention & Growth agent
Loop analyses repeat purchase behaviour, lifetime value and post-purchase flows.
Monitoring scope
Repeat purchase rate by cohort Time to second purchase Lifetime value trajectory Email and SMS revenue share Post-purchase flow performance Churn and reactivation signals Example finding
Second-purchase window stretching
Time to second purchase has increased 12 days over the last 90-day window.
→ Becomes an Implementation Pack via Forge.
Forge Implementation Packs agent
Forge turns every high-priority finding into a structured Implementation Pack — owner, steps, QA checklist, success metric and validation plan.
Monitoring scope
Owner role assignment per finding Step-by-step recommended fix QA checklist and acceptance criteria Success metric and validation window Linked agent finding and evidence Status lifecycle from Generated to Validated Example finding
Pack: GA4 purchase under-reporting
Generated owner, 5 fix steps, QA checklist, success metric and 14-day validation window.
→ Becomes an Implementation Pack via Forge.
