A Head of Commerce Assurance for mid-market ecommerce brands

Your checkout passed every test. The order still didn’t ship.

Commerce stacks fail between systems: storefront to PSP, PSP to OMS, OMS to warehouse, warehouse to carrier. Commerce Assurance is one senior person who owns that risk end to end. The journeys, the Business UAT, and the sign-off that says an order makes it from checkout to doorstep, and back again when it’s refunded.

REPLAY INCIDENT: ORDER CA-58291
  1. STOREFRONTcheckout.confirmedOK
  2. PSPpayment.capturedOK
  3. OMSorder.createdOK
  4. OMS→WMSexport.attempt_1TIMEOUT
  5. OMS→WMSexport.attempt_2TIMEOUT
  6. OMS→WMSexport.attempt_3TIMEOUT
  7. WMSorder CA-58291NO RECORD
CUSTOMER SEES: “ORDER CONFIRMED” WAREHOUSE SEES: NOTHING
FIG. 1 ORDER PIPELINE CA-58291
THE RETURN AND REFUND PATH: SAME SYSTEMS, REVERSED, TESTED LESS STOREFRONT checkout confirmed PSP payment captured OMS order created WMS nothing received CARRIER awaiting manifest
  1. STOREFRONTcheckout confirmed
  2. PSPpayment captured
  3. OMSorder created
  4. WMSexport timeout ×3, stranded, no alert fired
  5. CARRIERawaiting manifest
Order CA-58291 was confirmed at checkout, charged at the PSP and created in the OMS, but it never reached the warehouse. Every individual system reported green. The customer found out before anyone else did.
0 Systems in a typical mid-market stack
0 Handoffs where an order can vanish
0 Senior owner who closes the gap
§01

Where orders die

Your QA team tests screens. Your agency tested the build. Your platform vendor tested their platform. Nobody tests the seams, and the seams are where revenue leaks. Orders don’t die on the page. They die in the handoffs. Four entries we see over and over:

  1. No. 1 PSP → OMS

    Payment captured. Order never created.

    A webhook misfires, or a retry fires twice. The customer is charged, sometimes twice, and the OMS either has no order or has two. Support finds out first. Finance finds out at reconciliation.

    Charged twice
  2. No. 2 OMS → WMS

    The warehouse never saw it.

    An export job times out, a field mapping drifts after a release, a SKU fails validation silently. Orders sit in “processing” for days. No alert fires, because no system believes it failed.

    Never shipped
  3. No. 3 Refund path

    Refunds work, for some payment methods.

    Cards refund fine. Klarna, PayPal and gift-card splits fail halfway: the OMS marks the order refunded, the PSP never moves the money. Each one is a complaint, then a chargeback, then a trust problem.

    Never refunded
  4. No. 4 WMS → Carrier

    Tracking emails for parcels that never left.

    Labels generate, emails fire, and the manifest never reaches the carrier. Cross-border and split shipments make it worse, because the second parcel is the one that vanishes. Peak season makes it unmissable.

    Lost in transit

These failures share one trait: every individual system reports green. Only the journey is broken, and the journey is nobody’s job.

§02

Leadership, not labour

We don’t sell tester hours. You get a senior owner who designs and runs your order-lifecycle assurance function, using the QA, engineering, agency and ops teams you already pay for.

[ + ] What you get

  • A senior owner for order-lifecycle risk, from checkout to doorstep and back through refunds and returns.
  • A living map of your revenue-critical journeys across storefront, PSP, OMS, WMS, carriers and comms.
  • Release and peak-season test plans with named owners. Your people execute; we direct.
  • Structured Business UAT built on real orders and real failure modes, not screen walkthroughs.
  • Monthly leadership summaries: what broke, what’s fixed, what’s still risky.

[ − ] What we are not

  • Not a bench of manual testers running page-level checks.
  • Not a test-automation outsourcing project for generic UI coverage.
  • Not a tooling vendor. We work in your ticket system and your dashboards.
  • Not a replacement for your QA team. We make them effective at cross-system risk.
  • Not project-scoped consulting that ends at go-live and leaves nothing behind.

“If performance teams keep the site up, we make sure every legitimate order has a safe path from checkout to its final state: delivered, refunded, or returned.

Alex Boydell Founder & Head of Commerce Assurance
§03

Two ways to engage

OPTION AFixed scope · two to four weeks

The Commerce Audit

A short, intense engagement. We map your stack, run real orders through the highest-risk journeys, and hand your team a plan they can actually run.

  • Revenue-critical journeys and every system handoff, mapped
  • Real orders through the high-risk paths: mobile, cross-border, discounts, split shipments, click and collect, returns
  • Storefront → PSP → OMS → WMS → carrier → comms validated, refund path included
  • One structured Business UAT cycle with your cross-functional stakeholders
  • A prioritised risk map and a repeatable assurance plan, handed over in a workshop

A fixed, scoped piece of work. It pays for itself the first time it catches a significant incident before peak season does.

Scope an audit
OPTION BRetainer · one to three days a week

Continuous Assurance

The role itself, ongoing. An alternative to hiring a full-time head of QA, sized to your stack rather than your headcount.

  • A living journey and risk map, kept current as your stack changes
  • Per-release and seasonal test plans: what must be tested, by whom, before what date
  • Business UAT cycles planned and facilitated around major changes
  • Incidents, support tickets and refund data folded back into the risk map
  • Monthly leadership summaries; quarterly assurance reviews

Tiered by complexity, not headcount: Lite for a single brand and country, Standard for multi-region, Plus for multi-brand.

Discuss a retainer

Areas of expertise

  1. SVC-01

    Review

    An outside read of where you stand. Test coverage, automation strategy and order-lifecycle risk assessed and prioritised, so you can see what’s exposed before you commit to anything bigger.

  2. SVC-02

    Test strategy

    The end-to-end approach, designed for your stack. Journey maps, data sets, Business UAT scripts and release-gating checklists your team can run without us in the room.

  3. SVC-03

    Management

    Ongoing ownership of the assurance function. Deciding what gets tested each release, directing your QA, engineering and ops teams, and reporting risk to leadership.

  4. SVC-04

    Training

    Coaching your QA and engineering teams to design cross-system tests themselves, so the capability stays in-house long after we step back.

  5. SVC-05

    Omnichannel

    Click and collect, ship-from-store, marketplaces and store returns. The journeys that cross channels cross the most systems, and they fail the most. We test the promises your channels make to each other.

§04

How an audit runs

  1. 01

    Discovery

    Remote sessions to map systems, integrations, payment methods and logistics. We pull incident history and support pain points, and pick your top three to five revenue-critical journeys.

  2. 02

    Test design

    End-to-end journeys defined across systems, with data sets that match reality: products, discounts, payment methods, addresses, edge cases. UAT scripts and checklists prepared.

  3. 03

    Execution

    Real orders through each journey while we watch behaviour and logs across every system. Business UAT sessions facilitated with your cross-functional stakeholders.

  4. 04

    Analysis

    Observed failures and weaknesses mapped to revenue and operational impact, then prioritised by risk and effort, not by who shouts loudest.

  5. 05

    Workshop & handover

    Findings walked through with leadership and teams. You keep the journey map, the UAT scripts and the test cadence. A function your team can run, not a PDF that gathers dust.

§05

Is this you

Built for brands whose stacks have outgrown checkout testing. Mark each line that sounds familiar. Three or more and we should talk.

Packing slip № PS-2026-001 Items: 7
  • Shopify Plus, Adobe Commerce, BigCommerce, SFCC or custom headless, plus a dedicated OMS or WMS
  • Multiple PSPs and payment methods: Stripe, Adyen, PayPal, Klarna, wallets
  • Support tickets about missing orders, duplicate charges, or “you never sent it”
  • An incident where the warehouse or carrier never saw orders checkout called successful
  • Click and collect or ship-from-store promises that depend on store stock being right
  • Business UAT that’s ad-hoc, screen-focused, and skipped when deadlines bite
  • A knot in your stomach before big releases, migrations and Black Friday
Three or more ticked → proceed to §06
§06 · Correspondence

Run one real order through your stack with us watching every hop.

Thirty minutes is enough to tell whether your handoffs deserve an audit. If they don’t, we’ll say so plainly and you’ll have lost half an hour.

hello@commerceassurance.co →

Replies within one working day