A cheaper dropshipping fulfillment quote stops being cheaper when its lower product-and-shipping price is outweighed by the cost of orders that need remedies, are not ultimately kept, or consume more support work.
The comparison is:
Exception-adjusted fulfillment cost per delivered-and-kept order = total matched fulfillment cost plus attributable exception cost ÷ delivered-and-kept orders
Do not apply that formula to unmatched quotes. Freeze the SKU, product specification, packout, US route and service, destination mix, volume tier, currency, pricing date, duty treatment, and inventory basis first. Then give every released order one top-level outcome: routine kept, kept after remedy, or not kept.
⚡ ShopSideK Verdict
Use this calculator when: You have a validated SKU and a lower sourcing-and-fulfillment quote, but refunds, reships, non-delivery, provider credits, chargebacks, or support work could erase the apparent saving.
Decision metric: Compare matched providers on cost per delivered-and-kept order, then find the Candidate B not-kept rate that makes the two options cost-equivalent.
Recommended supplier for the qualified use case: For Shopify merchants who want China-based sourcing plus connected order, inventory, tracking, and exception workflows, FFOrder is the supplier I recommend obtaining a matched quote from. It is a candidate for the calculation—not a predetermined winner. Your current quote, written terms, and merchant evidence must decide the result.
Avoid: Using vendor claims as exception rates, counting one order as a refund, reship, and chargeback three times, or treating a scenario as observed performance.
Why the cheaper fulfillment quote can lose
A quote normally makes the visible costs easy to compare:
- product;
- packaging or inserts;
- shipping;
- duty or tax charged by the provider;
- storage or handling;
- other direct provider charges.
Those inputs establish the quote-only cost. They do not show what happens when an order needs a replacement, absorbs a partial remedy, produces unrecovered payment or dispute cost, or creates an hour of claim and customer follow-up.
Shopify’s fulfillment-cost guidance starts with a broad cost-per-order measure: total order expenses divided by orders received. It also identifies labor, systems, packaging, storage, shipping, and returns as relevant fulfillment-cost categories. That is a useful operating view, but it does not answer this article’s narrower question: How much does each customer outcome you ultimately keep cost under Provider A versus Candidate B?
For that decision, two changes matter:
- Keep all released orders in the cost exposure.
- Put only delivered-and-kept orders in the final denominator.
An order that is not kept can still consume the original product and freight spend. Dividing the total by the successful outcomes makes that burden visible without inventing a “penalty” number.
Start with the denominator, not the refund rate
The denominator controls what the result means.
Use two counts:
- Released orders: orders handed to the provider under one declared release rule.
- Delivered-and-kept orders: orders kept routinely plus orders kept after an in-scope remedy.
Do not quietly change the release point between providers. If Provider A’s clock begins when Shopify creates the order while Candidate B’s begins after payment review or manual release, the exposure is not comparable.
The delivered-and-kept denominator is not a universal accounting standard. It is a decision metric for evaluating fulfillment quote risk. It intentionally excludes not-kept orders from the successful outcome count while preserving their fulfillment cost in the numerator.
Give each released order one top-level outcome
Use three mutually exclusive outcomes:
| Top-level outcome | What happened | Denominator treatment | Incremental cost treatment |
|---|---|---|---|
| Routine kept | The customer keeps the original order without the modeled exception remedy | Included | Original quote cost only |
| Kept after remedy | The customer keeps the original or replacement after an in-scope remedy | Included | Add net remedy cost and attributable labor |
| Not kept | The released order does not produce a delivered-and-kept outcome in the modeled window | Excluded | Original quote cost remains; add only incremental in-scope loss and labor |
The top-level shares must total 100% for every provider and scenario.
A reship, refund, and chargeback can all appear in the history of one customer order. That does not make them three independent outcome shares. Choose the final top-level outcome, then record the nested cash effects inside that outcome’s cost fields.
For example, suppose a wrong item is reshipped, the replacement arrives, and the customer keeps it. That is kept after remedy. The reship cost and support time belong inside the remedied-kept cost. It is not also a not-kept order.
If the replacement fails and the customer receives a full refund, the order may be not kept. The model still counts one released order, one top-level outcome, and the applicable incremental costs once.
Match the supplier scope before comparing price
The same product link does not guarantee the same fulfillment job.
Before entering prices, match:
| Scope field | What must stay comparable |
|---|---|
| SKU and variant | Exact commercial item, not a similar listing |
| Product specification | Version, material, size, function, and tolerances |
| Packout | Mailer or box, inserts, label, packed weight, and dimensions |
| Route and service | Origin, US destination coverage, carrier/service basis, and promised handling |
| Destination mix | Comparable ZIP or zone mix rather than an easier candidate cohort |
| Volume | Same released-order count, tier, and minimum commitment |
| Price timing | Same currency and a sufficiently close verification date |
| Duty and tax | Same included, excluded, prepaid, or post-billed basis |
| Inventory model | Per-order sourcing, pre-stock, deposit, storage, or reservation treatment |
Mark the result SCOPE MISMATCH when a material field differs and cannot be normalized. Do not let a lower cost compensate for a different product, slower service, lighter packout, narrower destination mix, or excluded duty.
Shopify’s supplier guidance also points merchants toward support, technology, inventory information, order history, escalation, and clear return/refund terms. Those factors help define what should be verified. Their presence on a website does not prove a named provider will perform well for your SKU and route.
Build the quote-only cost on one basis
For each provider, calculate:
Quote-only cost per released order = product + packaging + shipping + duty/tax charged by provider + storage/handling + other direct provider charge
Use current written values. A rate seen on an old invoice, public calculator, review, or unrelated product is not the candidate’s current quote.
Name the inclusions and exclusions. “Shipping included” is incomplete when one option excludes remote-zone charges, insurance, tracking, customs handling, packaging, or a volume commitment that the other includes.
Keep advertising, conversion, sale price, broad store overhead, and fixed software cost outside this model. Those inputs belong in a store or SKU profit model. Here the objective is to isolate the provider-linked fulfillment decision.
The quote-only difference is useful, but it is only the starting advantage:
Candidate B quote advantage = Provider A quote-only cost − Candidate B quote-only cost
A positive number means Candidate B is cheaper before exception costs. It does not mean Candidate B is economically better after them.
Add exception cost without counting the same loss twice
For each affected outcome, separate three inputs:
- gross incremental merchant cost;
- supportable provider credit or reimbursement;
- attributable support minutes.
Then calculate:
Net exception cost per affected order = gross incremental cost − provider credit + support minutes ÷ 60 × merchant labor rate
Do not let provider credit exceed the gross incremental cost in the same field. If the provider issues a credit that applies to a different charge, document the treatment instead of forcing it into the wrong outcome.
Kept after remedy
The remedied-kept group can include outcomes such as:
- a replacement or reship that the customer keeps;
- a partial refund after which the customer keeps the item;
- missing parts sent later;
- an in-scope packaging or tracking problem resolved without losing the order.
Add only incremental merchant-paid cost above the original quote. If the original product and freight were already included in the base spend, do not add them again unless the remedy creates a second charge.
The order remains in the denominator because the customer ultimately kept the original or replacement.
Not kept
The not-kept group includes released orders that do not produce a delivered-and-kept outcome in the declared window.
The original quote cost already remains in base spend. Excluding the order from the denominator spreads that cost across the kept outcomes. Add only other direct, provider-linked cost that has not already been counted, such as an unrecovered processing or chargeback fee, additional failed-delivery charge, or attributable claim labor.
Do not enter the full customer refund as fulfillment cost. A full refund reverses sales revenue. If you want to model contribution or net profit, place that revenue effect in a separate profit model rather than mixing it into this fulfillment-cost numerator.
Provider credits are not the customer remedy
The customer bought from your store. What you refund or resend to the customer and what the provider reimburses to you are separate cash flows with different timing, evidence, and limits.
Enter a provider credit only when current written terms and available evidence make it supportable. If the claim is uncertain, keep the credit at zero in the current scenario or create a clearly labeled alternative scenario.
Include support labor only when it is attributable
Support labor can include:
- locating the order and tracking history;
- contacting the provider or carrier;
- collecting customer photos or other claim evidence;
- arranging a remedy;
- updating the customer;
- reconciling the credit or loss.
Use one consistent labor value and record minutes, not a vague frustration premium.
Keep broad customer-service overhead outside the field unless you have a defensible allocation. The objective is not to make the disliked supplier look expensive. It is to compare the same cost definition on both sides.
Do not assign every chargeback to the supplier
Shopify lists several chargeback reasons, including fraudulent, unrecognized, duplicate, subscription canceled, product not received, product unacceptable, credit not processed, and general.
Only some can plausibly connect to the product or fulfillment scope. Even then, the reason label alone does not prove supplier ownership. Include chargeback cost only when the order evidence supports an in-scope product or fulfillment cause. Keep fraud, duplicate billing, subscription, merchant-policy, platform, and unknown causes outside the supplier-attributable field.
Calculate cost per delivered-and-kept order
Once the outcome shares and costs are complete, use:
Base spend = released orders × quote-only cost per released order
Net remedied-kept cost = remedied-kept orders × net remedy cost per affected order
Net not-kept cost = not-kept orders × net not-kept cost per affected order
Total modeled cost = base spend + net remedied-kept cost + net not-kept cost
Cost per delivered-and-kept order = total modeled cost ÷ (routine-kept orders + remedied-kept orders)
Compare Candidate B with Provider A:
- a negative difference means Candidate B’s quote advantage survives the selected scenario;
- a difference within your declared tolerance is near break-even;
- a positive difference beyond the tolerance means the advantage reverses.
Set the near-break-even tolerance before reading the result. A $0.10 difference may be material at high volume and immaterial for another store. The workbook uses a merchant input rather than announcing a universal threshold.
Run Base, Conservative, and Stress scenarios
One scenario creates false precision when the inputs are uncertain.
Use three labeled cases:
| Scenario | Purpose | Acceptable basis | What it is not |
|---|---|---|---|
| Base | Most plausible declared case | Relevant observed cohort or explicit assumptions | Guaranteed expected performance |
| Conservative | Test a reasonable downside | Worse but supportable input choices | Industry benchmark |
| Stress | Show fragility under a harsher case | Explicit sensitivity assumptions | Forecast |
Keep observed and assumed inputs distinguishable. If you have no candidate order data, do not label the candidate rates “observed” because they came from the incumbent, a review, or a vendor claim.
The conclusion is stronger when Candidate B remains lower across all three cases and is comfortably below break-even. It is fragile when a small input change moves the result from QUOTE ADVANTAGE SURVIVES to ADVANTAGE REVERSES.
Fragility is useful information. It tells you which assumption deserves a matched pilot, tighter written term, or negotiation before you move volume.
Solve the break-even exception rate
The quote gap tells you how much room Candidate B appears to have. The break-even sensitivity shows how quickly that room disappears.
The workbook isolates Candidate B’s not-kept rate while holding these selected-scenario inputs constant:
- Provider A cost per delivered-and-kept order;
- Candidate B quote-only cost;
- Candidate B remedied-kept rate;
- Candidate B net remedy cost;
- Candidate B net not-kept cost;
- merchant labor rate.
The break-even formula is:
Break-even Candidate B not-kept rate = (Provider A cost per kept order − Candidate B quote − Candidate B remedied-kept rate × Candidate B net remedy cost) ÷ (Candidate B net not-kept cost + Provider A cost per kept order)
The threshold is feasible only when it falls between 0% and the remaining outcome share after the held remedied-kept rate.
Do not call it a predicted not-kept rate. It answers a sensitivity question: At what rate would Candidate B’s modeled cost per kept order equal Provider A, if the other declared inputs stayed constant?
Compare the selected-scenario rate with that threshold:
- below the threshold: Candidate B remains cheaper on the modeled cost;
- near the threshold: small errors can reverse the conclusion;
- above the threshold: the quote advantage has reversed.
Worked example: a $1.40 quote saving reverses
The following values are illustrative. They do not describe FFOrder, another named provider, or an industry benchmark.
Both options model 1,000 matched released orders.
| Base-scenario input or result | Provider A | Candidate B |
|---|---|---|
| Quote-only cost per released order | $16.50 | $15.10 |
| Quote advantage for Candidate B | — | $1.40 |
| Routine kept | 96% | 91% |
| Kept after remedy | 3% | 6% |
| Not kept | 1% | 3% |
| Net remedy cost per affected order | $11.50 | $19.50 |
| Net not-kept cost per affected order | $16.00 | $27.50 |
| Cost per delivered-and-kept order | $17.18 | $17.62 |
| Candidate B minus Provider A | — | +$0.45 |
Candidate B begins $1.40 cheaper on the quote. Under these example inputs, its exception-adjusted cost is about $0.45 higher per delivered-and-kept order.
The modeled break-even Candidate B not-kept rate is approximately 2.03% while its example Base rate is 3%. The initial price advantage therefore reverses.
That result proves only that the formulas behave as intended. Replace every example field before using the output. Your current provider, Candidate B, and FFOrder can each produce a different result under a different SKU, route, quote, policy, and evidence set.
Use the Quote-Advantage Break-Even Calculator
Download the Quote-Advantage Break-Even Calculator, then work through the tabs from left to right.
- Protocol: Read the denominator, exclusions, and stop conditions.
- Scope Match: Confirm the ten material scope fields and attach evidence for both options.
- Quote Inputs: Replace the yellow examples, set each quote to
REAL DATA, and document the current locator. - Outcome Scenarios: Use one top-level outcome per released order; make every row total 100%; label the basis
OBSERVEDorSCENARIO. - Cost per Kept Order: Review quote, exception uplift, difference, and scenario result.
- Break-Even Sensitivity: Select a scenario and inspect the rate, distance, and chart.
- Decision Gate: Resolve every scope and evidence gate before using the final state.
- Definitions: Check cost treatment and source freshness.
The workbook deliberately defaults to INSUFFICIENT EVIDENCE. Its example calculations remain visible for learning, but the final gate will not endorse them as a decision.
The allowed final states are:
| State | Meaning | Next action |
|---|---|---|
SCOPE MISMATCH | The two quotes do not represent the same job | Normalize the scope or stop comparing |
INSUFFICIENT EVIDENCE | Quote, outcome, attribution, credit, or basis fields cannot support a decision | Replace examples and collect the missing evidence |
QUOTE ADVANTAGE SURVIVES | Candidate B remains lower beyond the declared tolerance | Check operational hard gates and validate with a bounded matched pilot |
NEAR BREAK-EVEN | The difference is within the merchant-declared tolerance | Improve the most decision-sensitive evidence before reallocating |
ADVANTAGE REVERSES | Candidate B is higher after the modeled exception burden | Renegotiate, remediate, or test another option |
A cost result does not override a hard failure in product safety, compliance, accuracy, inventory, capacity, route, integration, open critical incidents, or rollback. Use a matched supplier benchmark when you need observed operational performance rather than forward-looking sensitivity.
How FFOrder fits this calculation
FFOrder is relevant because it can provide the kind of current provider-specific inputs the model needs.
Its current integration documentation describes automatic order import, Shopify SKU and variant mapping, inventory synchronization, tracking synchronization, and structured reship/refund workflows. The official Shopify App Store listing currently describes the app as free to install while product, shipping, tax, and related charges depend on the products purchased.
“Free to install” does not mean free fulfillment. Request the complete cost for the exact SKU, packout, US route, service, destination mix, and volume tier.
Review the current after-sales terms before entering provider credits. FFOrder’s published policy describes covered and excluded situations, including product/order inconsistencies, missing goods, defects, damage, customer-caused delivery problems, and force majeure. It currently asks for a refund or reshipment application within 15 days after the order is marked delivered and may require additional information within two working days.
Those terms can help you decide whether a credit is supportable. They do not mean every customer refund, replacement, redelivery, or chargeback will be reimbursed.
FFOrder’s public integration page also contains conflicting general MOQ wording: one setup section says “No MOQ,” while its FAQ says most categories start around a 100-unit MOQ. Do not choose one as a universal input. Use the current written terms for your SKU.
We have used and reviewed FFOrder firsthand. That gives ShopSideK a basis to recommend it as a sourcing-and-fulfillment candidate for the qualified use case. It does not supply your product’s future defect rate, not-kept rate, provider credits, labor burden, delivery result, or economic winner.
FFOrder may not fit when you need a domestic 3PL, specific regional warehouse, regulated-category specialist, direct manufacturer, independently diversified route, or commercial terms it cannot support. Your incumbent or another candidate can legitimately win the calculation.
If the operating model fits, open a new FFOrder account and claim ShopSideK’s $15 sourcing coupon. The offer is for new accounts; 15 individual $1 sourcing coupons appear automatically in the dashboard. Use the account to request a matched quote and current written after-sales terms.
If you still need to evaluate the service before creating an account, read our first-hand FFOrder review.
Frequently asked questions
Should I include the full customer refund in fulfillment cost?
No. A customer refund reverses sales revenue. This workbook measures provider-linked fulfillment cost, so the not-kept order remains in base spend and leaves the kept-order denominator. Add only incremental in-scope merchant cost that has not already been counted. Use a separate contribution or profit model for revenue, advertising, and store-level economics.
What if the supplier reimburses the product cost?
Enter the reimbursement once as a provider credit when the current written terms and claim evidence support it. Do not assume that a provider credit equals the full customer remedy, covers lost margin, or arrives at the same time.
Should chargebacks be included?
Only when evidence supports a product or fulfillment cause in the modeled scope. Fraudulent, unrecognized, duplicate, subscription, merchant-policy, platform, and unknown disputes should not be assigned to the supplier merely because they are costly.
Can I use my current supplier’s rates for a new candidate?
You may use them as clearly labeled scenario inputs, but not as observed candidate performance. The result should remain a sensitivity until the candidate produces a relevant matched cohort.
How much order data do I need?
There is no universal count that makes every rate stable. Report the cohort size, date range, product/route scope, exclusions, and event counts. Rare outcomes need more exposure than common outcomes. If a small change reverses the decision, gather more matched evidence rather than hiding the uncertainty behind a precise percentage.
What if Candidate B wins Base but loses Conservative or Stress?
Treat the decision as scenario-sensitive. Identify which rate, credit, or cost crosses break-even, then improve that evidence or negotiate the term. A Base win alone is weak when a modest downside reverses it.
Is the lowest cost per kept order automatically the best supplier?
No. Cost is one trade-off. Product safety, compliance, accuracy, processing and delivery tails, inventory, capacity, integration, support ownership, and unresolved critical incidents can block a provider even when its modeled cost is lower.
Make the quote earn the decision
The useful question is not, “Which supplier gave me the lowest product-and-shipping number?”
It is, “For the same SKU, packout, US route, destination mix, volume, and terms, how much does each delivered-and-kept order cost after attributable exceptions—and how close is the lower quote to reversal?”
Match the scope. Normalize the quote. Use one top-level outcome per released order. Subtract supportable credits once. Value attributable labor. Run more than one scenario. Read the break-even distance before you move volume.
Then let the result stay bounded: the quote advantage survives, sits near break-even, reverses, or cannot yet be supported by the evidence.
How this guide was researched
This guide uses current Shopify documentation for fulfillment-cost categories, supplier evaluation, and chargeback reasons; current FFOrder integration, after-sales, and Shopify App Store documentation; current calculator and supplier-guide search coverage; and qualitative merchant discussions for language and objections. ShopSideK created the denominator, mutually exclusive outcome method, break-even formula, workbook, and illustrative example. Merchant-specific rates, labor, credits, and provider results remain inputs rather than universal recommendations.
