The Ecommerce Analytics Glossary Every DTC Founder Needs (60+ Terms, Plain English)
by Om Rathod
|
9 min read
Aug 24, 2026
Ecommerce analytics tools love to argue with each other. Pull up Triple Whale, Northbeam, and Polar side by side and you'll get three different numbers for the exact same week of ad spend. Nobody tells you that upfront. So founders end up in board meetings citing a "ROAS of 3" without knowing whether that's blended, platform-reported, or attributed on a 7-day click window versus a 1-day view.
This is why an ecommerce analytics glossary for DTC founders needs to exist as its own thing, not buried inside a tool's help docs where the definitions conveniently favor that tool's numbers.
The stakes aren't academic. A founder who mixes up blended CAC with paid CAC can look at a spreadsheet, see a "profitable" number, and approve a 30% budget increase that actually torches margin the second organic traffic softens. That's not a hypothetical. It happens every quarter to brands scaling ad spend on metrics they only half understand.
This glossary is grouped by category, attribution, unit economics, ad platform mechanics, operations, so you can skim it or hit ctrl-F for the one term your agency just threw at you on a call. It's meant to be a working reference for founders and CEOs running the business day to day, not a syllabus. For the actual formulas behind each metric, with edge cases and calculation variants, the data dictionary goes deeper than we can here.
Revenue and Profitability Metrics
MER (Marketing Efficiency Ratio)
What it measures: Total revenue divided by total ad spend across every channel
Formula: Total Revenue / Total Ad Spend
Why it matters: You can't game it by shifting budget from Meta to TikTok to make a channel report look better. It forces you to look at the whole business, which is exactly why more founders lean on MER over channel-level ROAS once they pass seven figures.
POAS (Profit on Ad Spend)
What it measures: Gross profit generated per dollar of ad spend, not just revenue
Formula: Gross Profit / Ad Spend
Why it matters: A ROAS of 4 means nothing if your margin is 20%. POAS is the metric that actually tells you if the ad spend made money. Thin-margin brands (supplements, apparel with high return rates) should treat this as the primary number, full stop.
Contribution Margin
What it measures: Revenue minus COGS minus variable costs (shipping, payment processing, fulfillment, returns)
Why it matters: Gross margin ignores the variable costs that scale with every order. Contribution margin doesn't. It's the number that tells you whether growth is actually adding value or just adding volume.
AOV (Average Order Value)
What it measures: Total revenue divided by number of orders
The common mistake: Tracking one blended AOV number. New customers and returning customers almost always have different basket sizes, and lumping them together hides whether your upsell flows are working at all.
Customer Acquisition and Lifetime Value Terms
Blended CAC
Formula: Total spend across all channels (paid, organic, referral overhead) / Total new customers
What it hides: Nothing, actually. That's the point.
Paid CAC
Formula: Paid channel spend only / Customers attributed to paid channels
The gap: If a brand gets 40% of new customers through organic and referral, blended CAC can run 40% higher than paid CAC. Founders who only look at paid CAC think acquisition is cheaper than it really is, because they're ignoring the true cost of the whole growth engine.
LTV (Lifetime Value)
Historical LTV: Actual revenue or profit from a customer cohort, calculated after the fact from real purchase data
Predictive LTV: A modeled estimate based on early purchase behavior (first 30-90 days), used to make decisions before you have years of cohort data
The catch: Predictive LTV is only as good as the model's assumptions. Treat it as a planning input, not gospel.
Payback Period
What it measures: How many days or months it takes a customer's gross profit to cover what it cost to acquire them
Why it matters: A brand with a 90-day payback period needs very different cash flow planning than one with a 9-month payback period, even at the same CAC.
Cohort Analysis
What it does: Groups customers by acquisition month and tracks their behavior over time, instead of averaging everyone together
Why it matters: Blended retention numbers can look fine while a specific cohort (say, everyone acquired during a discount promo) is quietly churning at twice the normal rate. You only catch that by looking cohort by cohort.
Attribution and Ad Platform Metrics
Last-Click Attribution
What it does: Gives 100% credit to the last touchpoint before purchase
The flaw: It systematically over-credits bottom-funnel channels like branded search and email, because those are usually the last thing a customer clicks, not the thing that actually created demand.
First-Click Attribution
What it does: Gives 100% credit to the first touchpoint
The flaw: Opposite problem. Over-credits top-of-funnel awareness plays and undervalues the channels that closed the sale.
Multi-Touch Attribution (MTA)
What it does: Distributes credit across several touchpoints in the customer journey
Best for: Brands with enough order volume and tracking infrastructure to make the model statistically meaningful, usually mid-market and up.
Marketing Mix Modeling (MMM)
What it does: Uses statistical modeling on aggregate spend and revenue data instead of individual user tracking
Best for: Brands where iOS privacy changes have made pixel-based tracking unreliable, or lower-volume brands where per-user attribution models don't have enough data to be stable.
Incrementality
What it measures: Whether ad spend actually caused a sale, versus a sale that would have happened anyway
How it's tested: Holdout groups and geo lift tests, where you turn spend off in one region and compare results to a matched region running as normal
Why founders should care: As iOS privacy changes keep degrading pixel accuracy, incrementality testing is becoming the actual gold standard, not platform dashboards.
Platform-Reported ROAS vs True ROAS
Platform-reported: The number Meta or Google shows you inside their own ad manager
True ROAS: Revenue attributed through a unified, cross-platform view that accounts for overlapping claims
The gap: Meta and Google both claim credit for the same sale constantly. Add up their individually reported numbers and you'll often "spend" more revenue than the store actually made.
Ecommerce Operations and Funnel Metrics
Engaged Sessions and Conversion Events (GA4)
Engaged session: A visit lasting 10+ seconds, with a conversion event, or with 2+ pageviews
Conversion event: Any action you've defined as valuable (purchase, add-to-cart, signup)
Why it trips founders up: GA4's default attribution model (data-driven, cross-channel) rarely matches what Meta or Google Ads report for the same period, and that's expected, not a bug to chase down.
Funnel Drop-Off Rate
Product view to add-to-cart: Typically loses 85-90% of viewers [VERIFY industry benchmark]
Add-to-cart to checkout: Usually the steepest drop, often 60-70% [VERIFY industry benchmark]
Checkout to purchase: The smallest drop if checkout is clean, often under 30%
Why it matters: Knowing which stage is bleeding tells you whether the problem is product pages, pricing friction, or checkout UX. Blended conversion rate tells you none of that.
Inventory Turnover and Sell-Through Rate
Inventory turnover: How many times inventory is sold and replaced over a period
Sell-through rate: Units sold divided by units received, over a given window
Why it matters for ad spend: Pushing paid traffic hard at a SKU that's about to stock out wastes money and annoys customers who land on a sold-out page.
Reconciliation
What it means: Matching reported sales (what Shopify or Amazon "shows" you sold) to the actual cash that lands in your bank account, after fees, refunds, chargebacks, and reserves
Why it's harder than it sounds: Amazon in particular holds reserves and settles on its own schedule, so "sales this month" and "deposits this month" are rarely the same number, and founders who assume they are get cash flow surprises.
Forecasting and Automation Terms
Demand Forecasting
What it does: Predicts future demand using seasonality, trend, promotional history, and external factors
How it differs from a trailing average: A trailing average assumes next month looks like the last three. Real demand forecasting accounts for the fact that November isn't a normal month and neither is the week after a price increase.
Anomaly Detection
What it does: Automatically flags unusual shifts, a sudden CPM spike, a conversion rate that dropped overnight, a channel that went quiet
Why it matters: Most founders find these problems by noticing the P&L looks off three weeks later. Anomaly detection catches it the day it happens.
AI Insights Layer / Copilot Tools
What it does: Surfaces the "why" behind a metric change (a specific SKU driving a margin dip, a specific campaign driving a CAC spike) instead of just showing you the number moved
Why it matters: A dashboard tells you ROAS dropped. An insights layer tells you it dropped because one ad set's CPM doubled after an audience overlap issue.
This is the category Trivas's Wingman and forecasting tools sit in, surfacing the driver behind a change across Amazon, Shopify, and ad platform data instead of leaving you to dig through five tabs to find it. Not the only way to solve this problem, but a real example of what "AI insights layer" actually means in practice.
How to Actually Use This Glossary
Don't try to memorize 60 terms. Pick the 3-5 that matter most at your current stage, blended CAC and payback period if you're scaling fast, POAS and contribution margin if margin is getting squeezed, and check whether your current dashboard reports them consistently. Half the time, the answer is no.
Watch for vanity-metric syndrome. It's easy to obsess over ROAS because it's the number every ad platform shoves in your face daily, while payback period and POAS quietly tell the real story about whether the business is healthy. A brand with a strong ROAS and a nine-month payback period is not in as good shape as it looks.
For the exact formulas, calculation windows, and edge cases behind every term here, the data dictionary has the full breakdown.
If you're tired of reconciling five different definitions of the same metric across five different tools, that's the exact problem Trivas's unified reporting is built to solve, pulling Amazon, Shopify, and ad platform data into one place with one consistent set of definitions. Worth a look next time your ROAS and your bank balance disagree.
Revenue growth leader and co-founder driving Trivas's commercial strategy. Om has led the product vision and execution from scratch. With a strong background in SaaS sales and GTM strategy, Om bridges product innovation with real-world customer needs.
Continue Reading
explore more insights
Anomaly Detection for Ecommerce Data: What It Is and How It Actually Works
3 min read
Implementation Strategies for Cohort Analysis Success
.d53b12e5.png&w=3840&q=75 "Trivas.ai")
by Om Rathod
