Profit analytics for e-commerce brands

Find the money hiding in your store dataThe E[X] Group — ad spend, margin and pricing analysis for Shopify and beyond

I take your store, ad and cost data and come back with a short deck of figures and comparison tables: where spend is wasted, which products actually make money, and what to change before the next peak. No jargon required.

+20%

holiday revenue

from moving ad spend two weeks earlier into the demand ramp

−31%

wasted ad spend

after cutting campaigns that were buying customers who would have bought anyway

18 SKUs

losing money

identified in a top-100 catalogue once returns and shipping were included

Illustrative examples of the kind of result these analyses surface.

Tell me what you want to know

One question is enough to start — "is Meta actually profitable for us?", "which products lose money?", "should we discount in November?". I reply within two business days with what data I would need and what the answer would take.

hello@theexgroup.com
Response within 2 business days
Built for brands doing $1M–$20M in annual revenue

What I analyse

Every one of these ends in the same place: a number, a comparison table, and a recommendation you can act on this quarter.

Ad spend returns by channel

Which of Meta, Google, TikTok, email and affiliates actually pays back — and where the next euro of spend belongs.

A/B and holdout tests

Set up and read tests so a "winner" is a real difference and not noise you paid to chase.

Incrementality and geo tests

How many of those sales would have happened anyway without the ad. The number platforms never show you.

Margin by product and SKU

Which bestsellers quietly lose money once COGS, shipping, discounts and returns are subtracted.

Discount and promo effectiveness

Whether a promotion grew profit or simply pulled forward orders you would have received anyway.

Customer value and CAC payback

What a customer is worth by cohort and acquisition channel, and how many months it takes to earn the acquisition cost back.

Repeat purchase and churn

Who comes back, when they come back, and which products or flows make a second order more likely.

Returns and refunds

Which products, sizes and regions drive returns, what they cost you, and what the fix is worth.

Pricing and elasticity

How volume actually responds to a price change, so a price move is a decision instead of a gamble.

Shipping and fulfilment cost

Free-shipping thresholds, carrier and region cost differences, and the true delivered cost per order.

Inventory and demand forecasting

What to stock ahead of a peak so you avoid both stockouts on winners and cash tied up in dead stock.

Seasonality and holiday planning

When demand actually rises, and how far ahead of it spend and promotions should be moved.

Funnel drop-off

Where revenue leaks between visit, add-to-cart, checkout and paid order — and what closing each gap is worth.

What you get

Two documents at the end of every engagement — one to decide from, one to verify with.

Findings presentation — 5 to 10 pages

Built for the person making the call. Headline numbers, side-by-side comparison tables, before-and-after charts, and a ranked list of actions with the expected impact of each. No statistical vocabulary required.

  • Headline figures and what changed
  • Comparison tables by channel, product and period
  • Charts that make the answer obvious
  • Ranked actions with expected upside

Official technical report

The full argument behind the numbers: every model used, why it was chosen over the alternatives, the assumptions checked, the robustness tests run, and the limits of what the data can support. For your CFO, your board, or any analyst who wants to audit the work.

  • Model specification and justification
  • Assumption and diagnostic checks
  • Robustness and sensitivity analyses
  • Honest statement of limitations

Your data: what I use and where it comes from

A short guide so we start on the same page. Most brands already have everything I need — it is a handful of exports.

Store platforms

  • Shopify & Shopify Plus
  • WooCommerce
  • BigCommerce
  • Magento / Adobe Commerce
  • Wix Stores
  • Squarespace Commerce
  • Salesforce Commerce Cloud
  • Shopware
  • PrestaShop
  • Ecwid

Marketplaces

  • Amazon Seller Central
  • Etsy
  • eBay
  • Walmart Marketplace
  • Faire

Ads, email & analytics

  • Meta Ads Manager
  • Google Ads
  • TikTok Ads
  • Amazon Ads
  • Google Analytics 4
  • Klaviyo

Money & costs

  • Stripe
  • PayPal
  • Shopify Payments
  • Xero
  • QuickBooks
  • Your COGS spreadsheet

How to pull your exports

01

Orders and products from your store

Shopify: Admin → Analytics → Reports → Export, plus Orders → Export as CSV. WooCommerce, BigCommerce and Magento all have an equivalent order export. Easiest of all: add me as a staff user with report access.

02

Ad spend from each platform

In Meta Ads Manager, Google Ads or TikTok Ads: Reports → export CSV at campaign and day level, including spend, impressions, clicks, conversions and attributed revenue.

03

Traffic and funnel from analytics

GA4: Explore → export your sessions, landing pages and checkout funnel — or simply grant me viewer access.

04

Email and SMS performance

Klaviyo (or your equivalent): export campaign and flow performance with revenue attributed per send.

05

Your costs

A simple spreadsheet is enough: cost of goods per SKU, average shipping and packaging cost, payment fees, and fixed monthly costs. This is what turns revenue analysis into profit analysis.

06

Send it over

Raw CSVs are perfectly fine — messy, inconsistent or partial exports included. Cleaning and reconciling the data is part of the job, not something you need to do first.

What makes the analysis stronger

  • 12+ months of daily data lets me separate seasonality from a real change.
  • Campaign-level ad spend beats account totals — that is where the decisions live.
  • Per-SKU costs unlock margin work; without them I can only speak to revenue.
  • All data and findings are confidential, and I sign an NDA whenever you need one.

Pricing

Sized for brands between $1M and $20M in annual revenue. Fixed scope, fixed fee — agreed before any work starts.

Focused audit

from €2,500

One question, answered end to end.

  • A single decision — e.g. ad spend efficiency or SKU margin
  • Findings deck plus technical report
  • One review call to walk through the results
  • Typically 1–2 weeks

Most popular

Full profit review

from €6,000

The whole picture, in one engagement.

  • Ad returns, margin by SKU, LTV/CAC, returns and promo effectiveness
  • Incrementality read on your largest spend channels
  • Prioritised roadmap of what to change first
  • Findings deck, technical report and working session

Ongoing partner

from €2,500 / month

A quantitative team member on retainer.

  • Monthly reporting on the metrics that move profit
  • Test design and readouts before you commit budget
  • Forecasts ahead of peak season
  • Direct line for ad hoc questions

Final pricing depends on scope, number of data sources, and the state of the data. You get a fixed quote after a short scoping call — no hourly billing surprises.

How it works

Four steps from your first message to a decision you can defend.

01

Tell me the decision

We start from the choice you are trying to make — scale a channel, reprice a range, plan Q4 — not from a dataset.

02

Get the data in

You export what you have using the guide above. I clean, reconcile and sanity-check it against your own reporting.

03

Analyse and verify

Models are fitted and then stress-tested. If a result does not survive the robustness checks, you hear that instead of a confident-sounding number.

04

Deliver and act

You get the short findings deck, the full report, and a call where we agree what changes next week.

Under the hood

You never need to read this part. But if you or your analyst want to know what is actually behind the figures, here it is — properly specified, properly verified, in R and Python.

Regression (OLS, regularised, hierarchical)A/B & experimental designDifference-in-differencesSynthetic controlRegression discontinuityInstrumental variablesTime series & forecastingCohort & survival analysisGLMs / GLMMsCross-validation & out-of-sample testingBootstrapping & robust standard errorsAssumption & residual diagnostics

Frequently asked questions

Practical answers about working together.

What e-commerce platforms do you work with?
Shopify and Shopify Plus most often, plus WooCommerce, BigCommerce, Magento / Adobe Commerce, Wix, Squarespace Commerce, Salesforce Commerce Cloud, Shopware, PrestaShop and Ecwid. Marketplace sellers on Amazon Seller Central, Etsy or eBay are equally welcome. Anything that exports orders and products as CSV works.
How do I get my data to you?
Export orders and product reports from your store admin, campaign-level CSVs from each ad platform, and a spreadsheet of your per-SKU costs. The quick guide on this page walks through each source. Raw and messy exports are fine — cleaning is part of the engagement.
Do I need to understand statistics to work with you?
No. The findings presentation is written in plain business language with figures and comparison tables. The technical report exists for anyone who wants to check the reasoning, but you never have to read it to act on the results.
What exactly do I receive?
Two documents: a 5–10 page findings presentation with headline numbers, comparison tables, charts and a ranked list of actions; and an official technical report presenting every model used, the argument for choosing it, the assumptions verified, and the limitations.
How long does a project take?
A focused audit typically takes one to two weeks from receiving the data. A full profit review usually runs three to five weeks depending on data quality and how many questions are in scope.
Is my data kept confidential?
Yes. All data, results and the fact of the engagement itself are treated as confidential, and I sign an NDA whenever you require one.