Ecommerce Analytics: The Reporting Stack That Reconciles to Shopify

Ecommerce analytics is the discipline of building one set of numbers that everyone in the business trusts — and for a Shopify brand, that means anchoring every report back to what Shopify actually recorded as revenue, not to what the ad platforms claim they drove. The gap between those two things is where most reporting goes wrong: Meta says it drove the revenue, Google says it did, TikTok says it did, and the sum lands well above the line item in your Shopify admin. The job of an ecommerce analytics stack is to be the referee, reconcile the self-reported numbers against the commerce source of truth, and surface the handful of metrics that actually decide whether the business is healthy.

This is the operator-grade version of that stack — what to measure, why platform-summed numbers overstate, how cross-channel deduplication works, and where the dashboard layer (including Level dashboards) fits. Written for a founder or operator who wants to interrogate their own reporting, not for a data engineer.

Why platform-summed numbers lie

The single most common reporting failure in ecommerce is adding up what the ad platforms self-report and treating that sum as truth. It is not. Every ad platform has a structural incentive to overstate its own contribution, and the mechanics guarantee overlap:

• Shopify attributes by last click within a 30-day cookie window — it credits the final referrer.
• Meta uses a 7-day click / 1-day view window.
• Google Ads uses data-driven attribution with up to a 90-day lookback.
• TikTok uses a 7-day click / 1-day view window.

A single customer who watches a TikTok, clicks an Instagram ad the next day, then converts through a branded Google search gets counted as a conversion by all three platforms. Your dashboards show three sales; your Shopify admin shows one (via Improvado). This is not fraud — it is self-reported attribution, and self-reported attribution is inherently biased toward the reporter.

The overstatement compounds with smaller data-quality issues. A currency-conversion error plus a duplicate-transaction rate from misfiring pixels can add up to a high-single-digit revenue overstatement on its own; on a brand reporting $2M monthly that is six figures of phantom revenue steering budget decisions (via Definite, Upstack Data). The fix is not a smarter attribution model layered on top of the platform numbers. It is reconciliation against the commerce ledger. We walk through the tooling that does and doesn't do this in our cross-channel marketing dashboard comparison, and the pattern of why the summed numbers mislead in why most agency reports lie.

Shopify is the source of truth — use it that way

The discipline that fixes most ecommerce analytics problems is one rule: Shopify is the financial source of truth. Revenue, orders, AOV, COGS-adjusted contribution — those numbers come from the order ledger, full stop. The ad platforms are inputs you analyze against that ledger, never substitutes for it.

This splits your stack into two clearly different jobs:

• Shopify (the ledger) answers what actually happened — total revenue, order count, AOV, new-vs-returning split, refunds. Use it for P&L, financial reporting, and any number that has to tie out to the bank (via Improvado).
• GA4 and the ad platforms (the lens) answer how it happened — pre-purchase behavior, traffic source, the user journey, on-site funnel drop-off. Use these for behavioral analysis, never as the revenue number.

The two will disagree, and that is expected. UTM parameters track the source of a session, not the source of an order; Shopify reads the tags and stores them on the order only if the buyer converts in the same session, while GA4 reads the same tags through a different attribution window. For a high-ticket brand with longer consideration windows, the gap between Shopify and GA4 can run 15-30% (via WeltPixel). Disciplined UTM hygiene narrows that to a 10-15% band — the threshold where budget conversations stop being arguments between two dashboards (via Improvado). The point is not to make them match exactly. It is to know which one to believe for which question.

What to actually measure

Platform ROAS is the metric founders ask for first and the one that matters least, because it is the number most distorted by the overlap problem above. The metrics that decide whether the business is healthy are these.

Contribution margin. What remains after every direct cost of the sale — COGS, shipping, payment fees, refunds, and ad spend — is removed. A $100 AOV at 40% gross margin is $40 of gross profit; a $50 CAC against it means you are down $10 on the first order and need the repeat purchase to break even (via Saras Analytics). Contribution margin is the number that tells you whether growth is profitable or just expensive.

MER and new-customer MER. Marketing Efficiency Ratio is total revenue divided by total ad spend across all channels — a blended, cross-channel number that cannot be inflated by per-platform attribution overlap because it never touches the platform-reported conversions (via Common Thread Collective). The refinement that matters most: split it into new-customer and returning-customer revenue, because new-customer acquisition economics differ dramatically from returning-customer economics, and a blended MER hides which one is carrying the business (via adlibrary.com).

CAC payback period. How many months it takes for a new customer's cumulative contribution to recover their acquisition cost. This dictates how fast marketing investment turns cash-flow-positive and is the single most important number for a brand financing growth out of revenue rather than venture capital (via Saras Analytics).

Repeat rate by cohort. Returning customers convert more easily, spend more per order, and carry higher margin. Measured by cohort — grouped by acquisition month — repeat rate is a leading indicator of retention quality that blended numbers obscure. The mechanics of building that cohort table correctly — and the three mistakes most dashboards make — are covered in our LTV by cohort piece.

The honest framing is stage-gated: at launch, watch contribution margin per order, blended CAC, and AOV; in growth, prioritize CAC payback and channel-level contribution; at scale, move to LTV:CAC by cohort and contribution margin by SKU-by-channel matrix (via Luca). Most founders skip ahead to the metric that sounds most advanced. The right one is whatever the brand has enough clean data to read. For the fuller list of dashboard metrics beyond these four core numbers, our marketing KPI dashboard guide rounds out the set.

Cross-channel deduplication: the part most stacks skip

If platform-summed numbers overstate because the platforms double-count, the obvious question is: how do you actually deduplicate? Not by picking a "better" attribution model and applying it on top of the summed conversions. You deduplicate by anchoring to a shared key.

The clean version reconciles every platform-reported conversion against the Shopify order ledger using order_id or a deterministic click-ID match. One order, one conversion, credited once — regardless of how many platforms claimed it. The platforms' view-through and click windows all touch the same conversion path, and summing them runs the totals well above actual sales for any active multi-channel account (via Cometly).

We break down which tooling tiers actually do it in our agency dashboard tools comparison.

The dashboard layer: three jobs, not one

The ecommerce analytics stack is not one dashboard. It is three distinct jobs, and conflating them is where teams waste money on tools that overlap and still leave gaps.

1. The commerce-reconciled profit view. Built on the Shopify ledger, this is where contribution margin, MER, CAC payback, and cohort LTV live. Profit-focused platforms purpose-built for Shopify — Polar Analytics, for example, ships pre-computed contribution margin, LTV-by-cohort, COGS, and inventory metrics — sit here (via Polar Analytics). This is the layer that answers "are we actually making money."

2. The paid-channel performance view. A live, reconciled read on Meta, Google Ads, and TikTok spend and efficiency, refreshed frequently enough to act on. This is where the day-to-day media decisions get made, and where attribution-window discipline matters most. Broad aggregators like Triple Whale connect 100-plus sources and sit at the executive-reporting end of this layer (via Triple Whale).

3. The behavioral / journey view. GA4 and on-site analytics — funnel drop-off, source/medium of sessions, pre-purchase behavior. Never the revenue number; always the why behind it.

Level — the reporting platform we built at Marketing Bar — is purpose-built for job #2: a single live, reconciled view of paid-channel performance across Meta Ads, Google Ads, and TikTok Ads, with hourly sync and user-token authorization so the connections don't silently drop. Klaviyo email-attribution dedup and full Shopify reconciliation are on our roadmap, not shipped today — so if your pain is concentrated in the paid-channel slice and the constant disagreement between Meta, Google, and TikTok reporting, that is exactly what our reporting platform solves now. If your pain is profit reconciliation against the Shopify ledger, a profit-first tool from job #1 is the better current fit, and we will tell you that rather than oversell. For the tool-by-tool comparison across all three jobs, see our marketing reporting software roundup. Automating the weekly assembly of these views is its own discipline, covered in automated reporting tools for agencies, and the layout that makes them readable is in our agency report template.

When to bring in a marketing analytics agency

Most ecommerce brands hit a point where the analytics stack outgrows the operator who set it up. The CSV-export-and-reconcile-in-a-spreadsheet routine that works at $500K ARR quietly eats ten-plus hours a week by the time the brand is running four paid channels, email, and a subscription program (via Improvado). That is the inflection point where a marketing analytics agency earns its fee — not by owning another dashboard, but by owning the reconciliation discipline the dashboards depend on.

What a marketing analytics agency should actually do for an ecommerce brand:

• Stand up the source-of-truth hierarchy. Define Shopify as the financial ledger, GA4 as the behavioral lens, and the paid platforms as inputs — then enforce it so every report ties back to the same revenue number.
• Fix the tracking foundation first. UTM hygiene, server-side event deduplication, and currency normalization, before any dashboard is built. A pretty dashboard on broken tracking is worse than no dashboard, because it manufactures false confidence.
• Build the metric set that matters. Contribution margin, new-vs-returning MER, CAC payback, and cohort repeat rate — wired to update without a weekly manual export.
• Run the reconciliation cadence. A standing weekly check that platform-reported numbers still tie to the Shopify ledger within tolerance, with drift flagged before it corrupts a budget decision.

The mistake we see most often is hiring a marketing analytics agency to produce reports rather than to own data integrity. A report is an output. Data integrity is the asset. An agency that hands you a beautifully formatted deck built on un-reconciled platform numbers has sold you the deck, not the analytics. Ask any prospective partner the same diagnostic the tooling question uses: do your numbers reconcile to Shopify, and can you show me where they don't? The honest ones can.

If you want a reporting layer scoped to your channel mix rather than another off-the-shelf subscription, contact us for a scoped quote — pricing depends on channels, integration depth, and reconciliation scope, so it is quoted per engagement rather than priced off a tier sheet.

How to build the stack in order

The sequence matters more than the tool choice. Brands that buy the dashboard first and fix the data later spend twice.

Get the order right and the tooling decisions get easier, because you are choosing tools to serve a defined metric set against a known source of truth — not hoping a tool will tell you what to measure.

Where to next

If you want the tier-by-tier breakdown of cross-channel reporting tools, our cross-channel marketing dashboard comparison covers the four working categories and where dedup actually happens. For the deeper read on why summed platform numbers mislead, see why most agency reports lie. If you want agency analytics deployed on your paid-channel data, the Level page walks through the current Meta / Google / TikTok scope — and for a stack scoped to your brand, contact us for a scoped quote.