NTN ECOMM cheat sheet

Model taxonomy: B2C, B2B, C2C, D2C, marketplace, subscription and dropshipping differ in who holds inventory, who sets price, and who owns the customer relationship.

GMV = total selling value of goods transacted before returns, discounts and fees; it measures scale, not company revenue.

Take rate = platform revenue / GMV; managed marketplaces typically retain 10-20%, pure listing models single digits.

Contribution margin per order = price - COGS - payment fees - shipping - fulfillment labor - packaging.

CAC = total sales and marketing spend / new customers acquired in the same period.

LTV = ARPU x gross margin % / churn rate (equivalently ARPU x margin x average customer lifespan in periods).

Screening thresholds commonly applied to paid acquisition: LTV:CAC of at least 3:1 and CAC payback of 12 months or less.

Break-even ROAS = 1 / gross margin %; at a 40% gross margin the break-even ROAS is 2.5.

AOV = revenue / orders; storefront conversion rate = orders / sessions.

Platform tiers: hosted SaaS (Shopify, BigCommerce), licensed self-hosted (Adobe Commerce, WooCommerce), and headless/composable with an API backend and independent front end.

Core Web Vitals good thresholds at the 75th percentile: LCP at or under 2.5 s, INP at or under 200 ms, CLS at or under 0.1.

Instrument the full funnel: landing/PLP > PDP > add-to-cart > cart > checkout start > purchase, each step with its own event.

Staging and pre-production environments must be blocked from indexing via robots.txt or noindex; duplicate URLs are consolidated with rel=canonical.

Product data model: parent product > variants (size, color) > SKU, with attributes powering faceted navigation and filters.

Guest checkout and reduced-step checkout lower abandonment; every additional required form field adds friction.

PCI DSS SAQ A applies only when checkout is fully outsourced to a validated provider; SAQ A-EP applies when your site scripts touch the payment page.

Monitor TTFB, error rate and uptime against an SLA rather than availability alone; latency directly depresses conversion.

Product, Offer and Review structured data enable rich results with price and availability in search listings.

CPC = spend / clicks; CPM = (spend / impressions) x 1,000; CTR = clicks / impressions.

ROAS = attributed revenue / ad spend; ROI = (revenue - spend) / spend; break-even ROAS = 1 / gross margin %.

Channel CAC = channel spend / new customers attributed to that channel; compare it against channel-level LTV, not blended LTV.

Google Analytics 4 uses data-driven attribution by default; last-click over-credits lower-funnel channels and under-credits upper-funnel.

Google Ads Quality Score (1-10) is built from expected CTR, ad relevance and landing page experience; higher scores generally lower CPC.

SEO rests on three pillars: technical crawlability/indexability, on-page relevance, and off-page authority from backlinks.

Email reporting: list growth rate, click-to-open rate and revenue per email sent; open rate is unreliable since Apple Mail Privacy Protection.

Retargeting audiences decay after conversion; frequency caps and exclusion windows prevent spend on already-converted users.

Tag every campaign with UTM source, medium and campaign; untagged traffic collapses into direct/none and breaks channel reporting.

Price elasticity of demand = % change in quantity / % change in price; absolute value above 1 is elastic, below 1 is inelastic.

Initial markup % = (original retail - cost) / original retail; maintained markup uses the actual selling price after markdowns.

Markdown % = (original price - sale price) / original price.

Sell-through rate = units sold / units received in the period; weak early sell-through signals markdown timing.

GMROI = gross margin dollars / average inventory at cost; a value below 1.0 means margin does not cover the inventory investment.

Pricing tactics: skimming (high launch price), penetration (low launch price), bundling, anchor/decoy framing, and rules-based dynamic repricing.

Front-load the primary keyword in the title tag; product titles beyond roughly 60 characters are truncated in search results.

Duplicated manufacturer copy creates thin, duplicate content; unique descriptions plus original imagery improve both indexation and conversion.

Cross-sell displays complementary items while upsell offers a higher-value variant; both are the levers that move AOV in merchandising slots.

Card flow: authorization > capture > settlement; a void cancels before capture, a refund reverses after settlement.

PCI DSS scope: SAQ A when a validated third party handles all card data; tokenization removes card data from your own systems entirely.

PSD2 strong customer authentication in the EEA requires two factors; 3D Secure 2 satisfies SCA and shifts fraud chargeback liability to the issuer.

Visa chargeback monitoring threshold: 0.9% of transactions and 100 chargebacks per month; Mastercard: 1.5% and 100 per month.

Fraud risk signals: billing/shipping mismatch, AVS or CVV failure, BIN country vs IP country mismatch, order velocity per device, high-value first orders.

TLS 1.2 or higher is required for transmitting cardholder data; older protocols fail PCI DSS.

Offer the locally preferred payment methods (wallets, BNPL, bank transfer); a missing familiar method is a leading cause of checkout abandonment.

Retry soft declines on a schedule and never retry hard declines; soft-decline recovery is a material revenue lever.

Order lifecycle: placed > authorized > picked > packed > shipped > delivered > settled/closed, with returns branching off after delivery.

Reorder point = average daily demand x lead time in days + safety stock.

Safety stock = Z x demand standard deviation x square root of lead time; Z = 1.65 for 95% service level, 2.33 for 99%.

Fill rate = units shipped / units ordered; OTIF = orders delivered on time and complete / total orders.

Perfect order rate is the product of its component rates (on time x complete x damage free x correct documentation).

Dimensional weight bills the greater of actual weight and (L x W x H)/carrier divisor, so carton size drives shipping cost.

A free-shipping threshold lifts AOV only when set above the current AOV and remains margin-positive after shipping cost.

Returns tracking: RMA authorization at intake, restocking fee policy by reason code, return rate = returned units / shipped units.

Fulfillment models are in-house, 3PL and dropship; distributed inventory shortens transit time but raises total safety stock.

Net Promoter Score = % promoters (9-10 ratings) - % detractors (0-6 ratings); the scale runs from -100 to +100.

CSAT = satisfied responses / total responses; customer effort score uses a 1-7 scale where a lower score is better.

Contact rate = contacts / orders in the same period; a spike points to post-purchase friction such as unclear tracking or delivery delays.

First contact resolution is the stronger cost driver than average handle time because repeat contacts multiply total cost.

Self-service deflection rate = issues resolved in the help centre / total help-centre visits.

Proactive order, dispatch and delay notifications reduce inbound where-is-my-order contacts.

Churn rate = customers lost in period / customers at start of period; retention rate = 1 - churn rate.

Set SLA tiers by channel and issue severity, with published first-response and resolution targets per tier.

Revenue per session = revenue / sessions, which decomposes into conversion rate x average order value.

GA4 ecommerce events to implement: view_item_list, view_item, add_to_cart, begin_checkout, add_payment_info, purchase.

A/B test design: one pre-declared primary metric, sample size computed in advance, 95% confidence, 80% power, full business cycles to avoid day-of-week bias.

Cohort retention: group customers by first-purchase month and track repeat purchase rate in each subsequent period.

UTM taxonomy must be lowercase and consistent; source/medium pairs drive GA4 channel grouping and unattributed traffic falls to direct/none.

Leading indicators (add-to-cart rate, PDP views, email signups) predict outcomes; lagging indicators (revenue, LTV) confirm them.

Server-side tagging with a documented data layer reduces distortion from ad blockers and third-party cookie loss.

Funnel drop-off from step A to B = 1 - (sessions at B / sessions at A); prioritize the largest absolute loss, not the largest percentage.

GA4 is event-based with session-scoped and user-scoped dimensions; high-cardinality dimensions trigger thresholding in reports.