Net revenue by channel
Pacing to target
Channel mix
Top opportunities ranked by estimated quarterly contribution
Brand performance
LTV & CAC
Unit economics on trailing-12-month cohorts. LTV is 12-month gross-margin LTV per customer; nCAC is paid spend divided by paid-acquired new customers. Healthy is LTV:CAC ≥ 3 with payback inside 6 months.
LTV vs CAC by acquisition source
12-month, per customerRatio health
LTV ÷ CACAmazon LTV is modelled. The marketplace does not release customer identity, so Amazon cohorts are estimated from repeat-order patterns rather than matched customers.
Cumulative LTV curve
months since first orderCohort repeat rate
% of cohort with 2+ ordersRead across a row to see how a cohort matures; read down a column to compare cohorts at the same age. The November 2025 gifting cohort is the strongest to date.
Acquisition & ROAS
Platform-reported ROAS overstates performance because each platform claims the same sale. MER (total revenue ÷ total paid spend) is the arbiter; GA4-attributed ROAS and POAS (gross profit ÷ spend) sit between the two.
Paid platforms
nCAC by platform
last 14 daysOnline conversion funnel
Google Analytics 4Shopify and WooCommerce storefronts only. Amazon conversion is reported separately as unit session percentage.
CRM & email
Brevo campaigns and automations, matched to orders so email is credited with revenue rather than opens. Open rates are inflated by mail-privacy proxies; revenue per recipient is the measure to manage.
Recent campaigns
BrevoAutomation revenue
Channels
Shopify, WooCommerce, Amazon and in-store compared on contribution, not just revenue — marketplace fees, returns and paid media are netted off per channel.
Channel economics
Contribution margin by channel
after fees, returns & paid mediaAmazon health
Seller Central · Amazon AdsOperations & service
Fulfilment and allocation from the YALLA OMS, stock cover from Shopify, WooCommerce, Amazon FBA and POS, and service quality from HubSpot.
Stock cover risk
days of cover vs 30-day forecastNPS trend
12 weeksAI insights
The engine scans every connected source for shifts in efficiency, margin and demand, then proposes actions with the evidence, sources, estimated impact and confidence behind each. Nothing is executed without approval.
Ask the data
natural-language queryHow the engine works
1 · Unify
Orders from the OMS, spend from ad platforms, sessions from GA4, email from Brevo — resolved to one customer and one currency.
2 · Detect
Daily scans for anomalies, trend breaks and efficiency gaps against each brand's own baseline.
3 · Size
Each finding is costed in contribution margin, not revenue, with a confidence score from data depth and stability.
4 · Recommend
Ranked actions, each with evidence and sources. Accepted actions are tracked so the engine learns what worked.
Integrations
The YALLA OMS is the system of record for orders across every storefront, so revenue is counted once. Platform connectors supply what the OMS does not hold — spend, sessions, email engagement and marketplace fees.