Skip to content

MarketingLast updated 5 min readBy The PixelCrayons team

eCommerce revenue leak audit: nine checks, worked on two stores we built

The short answer

A revenue leak is a buyer who wanted to buy and did not, or bought less. This audit gives nine checks with the evidence for each, the checkout fixes to make before recovery emails, and two stores we built where a specific leak was designed out.

Nine numbered checks from product questions to tracking, with options, reorders and checkout marked

A revenue leak is a point where a buyer who wanted to buy did not, or bought less than they would have. Below are nine checks to find leaks, the checkout fixes to make before any recovery email, and two stores we built with specific leaks designed out.

Leak or loss: why growing stores miss them

A loss is an event you notice: the site goes down during a sale and orders stop. A leak is quieter. A product page that leaves one question unanswered, or a delivery charge that appears only at the last step, costs a few orders every day and never shows up as an incident.

Growth hides leaks. When revenue rises because traffic rises, a falling conversion rate or a shrinking repeat rate can go unseen for months. The audit below looks at each stage of the journey separately, so a good month at the top does not mask a problem lower down.

Start with measurement. A missing purchase event looks exactly like abandonment, so reconcile analytics with real orders before calling anything a leak; the checks are set out in our post on what to fix first in conversion optimisation.

The nine checks

Leak What to look for Where the evidence is
1. Unanswered product questions Pages with many views and few add-to-carts; the same pre-sale question asked every week Analytics by product page; support and chat transcripts
2. Hard-to-compare options Buyers choosing the smallest size or leaving at the variant picker Variant mix in orders; recordings of the product page
3. Slow pages Slow loading or delayed response to taps on product and cart pages Real-user field data by device, then a reproduced test
4. Mobile friction A phone conversion rate far below desktop for the same traffic source Analytics split by device and source
5. Checkout steps Drop-off at account creation, address or payment Checkout funnel by step; failed payment logs
6. Surprise costs Exits at the step where delivery and tax first appear Funnel step exits; customer emails about charges
7. Unrecovered carts Carts with an email address and no follow-up sent Email platform flow reports
8. No second order A low share of customers ordering again within your normal reorder window Order history grouped by first-order month
9. Blind spots Decisions made on figures that do not match orders A monthly reconciliation of analytics, orders and refunds

Count the cost of each before fixing anything. For example, add up the sessions that reached the step where delivery cost appears and left there. That number, not a benchmark from another store, tells you which leak to fix first.

A leak audit, in the order to run it
  1. 01ReconcileAnalytics against real orders, before calling anything a leak
  2. 02Nine checksEach stage checked separately, product page to second order
  3. 03Count the costSessions that reached a step and left there
  4. 04Fix the checkoutGuest checkout, the full cost early, payment options
  5. 05Recovery emailsWin back some of the carts still lost
  6. 06Ask whyCheckout recordings and a one-question exit survey
  7. Then redesignOnly once recordings and the survey show where buyers stall

Two stores, two leaks designed out

Real Mad Honey: options buyers could not compare. The store sells a rare wild honey in five jar sizes with an add-on spoon, to customers in many countries. Five sizes and an add-on could easily have become a cluttered choice.

We built the size choice so each jar shows its price per gram and its saving, with the most bought and best-value sizes labelled and the spoon beside them. A country choice shows each visitor the right prices and delivery, which covers check 6 as well. The record is in the Real Mad Honey case study.

Close-up of the size choice: jars from 3.5 oz to 17.6 oz, each with its saving and price per gram, Most bought and Best value labels, and an optional bamboo honey spoon
Real Mad Honey's size choice: each jar with its saving and price per gram, the Most bought and Best value labels, and the optional spoon. Captured 9 October 2026. The Real Mad Honey case study

310 Nutrition: the second order. The brand sells meal shakes, electrolyte drinks, superfoods and teas direct to customers, and its store had grown with the brand. We rebuilt it on Shopify around the path to a subscription.

The range was restructured into collections with filtering, the subscription offer was made more prominent and more reassuring on product pages and in the cart, and related products appear only where they add to an order. Pages got lighter, and the checkout has fewer distractions.

The aim was check 8 as much as the first sale: for a subscription brand, the store’s job includes the next order, so the offer belongs where the buying decision is made. The record is in the 310 Nutrition case study. Neither case record gives a revenue figure, and we do not add one.

Close-up of the buy box: product name, rating, benefit list, flavour and size choices, and quantity bundles
310 Nutrition's buy box: flavour and size choices and quantity bundles where the buying decision is made. Captured 8 October 2026. The 310 Nutrition case study

Fix the checkout before sending recovery emails

Recovery emails win back some carts. A checkout that stops losing them is worth more, and these are the fixes to check first:

  1. Guest checkout. Offer an account after the order, not as a gate before it.
  2. The full cost, early. Show delivery, taxes and any fees before the final step, with an order summary of items, delivery and discounts before payment.
  3. Payment options. Offer the cards and wallets your buyers expect, and check which method failed when a payment fails.
  4. A visible end. A progress indicator tells the buyer how many steps remain.
  5. Returns in plain sight. A clear returns policy, linked from the cart, answers a doubt that otherwise ends the visit.
  6. Delivery thresholds. If you offer free delivery above an order value, say so in the cart while the buyer can still add to the order.

Then find the remaining reasons. Recordings of checkout sessions show where buyers stall, and a one-question exit survey at checkout asks them why. Use both before redesigning anything.

When a leak audit is not the answer

  • A new store with little traffic. There is not enough data to tell a leak from noise. Acquisition is the problem to solve first.
  • The offer is the problem. If buyers reject the price or the product, a cleaner checkout does not change their mind.
  • A replatform is under way. Fix measurement on the new store first; leaks found on the old one may not survive the move.
Shweta Agarwal

Written by

Shweta Agarwal

Chief Operating Officer, Vinove

Shweta runs the operating side of the group, holding delivery, process and accountability consistent across the group. Her focus is on the systems that let teams grow without losing the standard they started with.

LinkedIn →
Keep reading

Related
reading.

Rankings on this topic

Scored on public evidence by the PixelCrayons Index. PixelCrayons is scored like every firm: ranked where it has a case study, listed where it has not.

The services behind this piece

Search, paid media, email and conversion, run as one roadmap by the team that wrote it.

Questions

Frequently
asked.

Run the measurement reconciliation monthly and the full nine checks after any major change: a redesign, a new checkout app, a new market or a pricing change. Most leaks start with a change nobody tested end to end.

The one with the largest counted cost on your own store. Count sessions and orders lost at each step, then fix the biggest one you can confirm, rather than the one a generic checklist puts first.

They show what was built and why, not a measured uplift: neither case record gives a revenue or conversion figure. Treat them as worked examples of designing a leak out, and measure the effect on your own store.

Want this applied
to your site?

Send the brief and an itemised, priced proposal comes back. Or start with the free website review: we measure one page of your website and a person reviews the report.

Reply within four business hours (typical) · Free · No retainer required