Payment Disputes and Chargebacks: A Merchant's Response Guide

The first dispute notification we ever received showed up in our merchant dashboard on an ordinary morning, flagged in red, with a response deadline that felt uncomfortably close. I handle payments and backend infrastructure for a cross-border e-commerce operation shipping into the Philippines, and up to that point we had processed a healthy volume of transactions without a single formal dispute. My first instinct was panic — fraud, misunderstanding, or a customer angling for a free product? I had no framework for answering any of that, and I burned half a day just figuring out what the notification meant before I could even start a response.

That scramble is the reason this article exists. If a dispute or chargeback just landed on your desk for the first time, the goal here is to shortcut the panic: what a dispute actually is, how the lifecycle works, what evidence genuinely changes outcomes, where merchants lose cases they should have won, and what to fix so this happens less often.

What a Dispute Actually Is (and Why It Isn't the Same as a Refund)

A refund is something you initiate: a customer asks, you agree, money moves back through the same channel, and the record closes cleanly. A dispute — commonly called a chargeback when it runs through a card network, or simply a "dispute" when it runs through an e-wallet like GCash or Maya — is initiated by the customer's bank, card issuer, or wallet provider on the customer's behalf, without you being consulted first.

That difference matters because a dispute is adversarial by design. The issuer or wallet provider acts as an intermediary between you and your customer, and by default the money moves against you unless you prove otherwise. You start from a position where funds have already been provisionally pulled or flagged, and your job is to win them back with evidence.

The Dispute Lifecycle: From Complaint to Final Decision

Once you understand the shape of the process, it stops feeling like an ambush. The lifecycle generally runs through five stages, though exact terminology and timing differ depending on whether you're dealing with a card network, a bank, or an e-wallet:

• The customer or their bank raises the dispute, contacting their card issuer or wallet provider directly — not you — with a stated reason: item never arrived, transaction wasn't authorized, service didn't match the description.
• You get notified, usually through your payment gateway or acquiring partner, with a reference number, a reason category, and a deadline.
• You have a response window to submit evidence. This is typically a matter of business days rather than weeks, though the exact length varies by network, wallet provider, and case type, and can change over time. Treating the deadline as "whenever I get to it" is the single most common way merchants lose.
• The issuer or wallet provider reviews what you submitted. This is a documentation review, not a hearing. Whoever assembles the clearer paper trail tends to win, regardless of who was actually telling the truth.
• A decision comes back. Either the dispute resolves in your favor and you keep the funds, or it resolves against you and the reversal — plus, in many cases, a processing fee — goes through.

The part that catches first-time merchants off guard is stage three. There is no grace period for "we were busy." If routing these notifications to someone who can act on them is slow or nonexistent, you can lose a case you had every right to win simply because nobody responded in time.

The Evidence That Actually Moves the Needle

Not all documentation is equally persuasive. Three categories of evidence have consistently done the heavy lifting for us:

Proof of delivery or service fulfillment. For physical goods, that means courier tracking showing delivery to the order's address, ideally with a timestamp, signature, or photo. For digital goods, it means system logs showing activation, download completion, or account provisioning. The fulfillment timestamp needs to make logical sense — delivery after the dispute was filed, or to a mismatched address, actively hurts you.

Customer communication that shows the transaction was legitimate. Support tickets, chat transcripts, or emails where the customer confirms receipt or asks a product question are strong counter-evidence against an "unauthorized transaction" claim. A customer who messaged support asking when their order would arrive is going to have a hard time later claiming they never made the purchase.

Matching transaction details. This sounds basic, but it trips up more merchants than it should: the order ID, amount, currency, and timestamp in your evidence package need to line up exactly with what the issuer has on file. A mismatch — even an honest one caused by a system logging local time instead of your gateway's timestamp — can get your submission thrown out on a technicality before anyone evaluates the substance of your case.

We route every dispute case through a documented evidence trail rather than an ad hoc email thread, partly because SMARTPAY operates under BSP-regulated OPS and EMI licensing, and a licensed payment partner's dispute workflow expects exactly this kind of structured, auditable record rather than a screenshot pasted into a chat.

Dispute Reason and Evidence Matrix

Different dispute reasons call for different evidence, and the realistic odds of winning shift with how clean your records are for that reason:

Dispute Reason	Typical Evidence Needed	Win Likelihood Factors
Item or service not received	Courier tracking, delivery confirmation, fulfillment system logs with timestamps	Stronger when delivery is confirmed to the order's address before the dispute was filed; weaker when tracking shows delivery after the dispute date or to a mismatched address
Unauthorized transaction / "I didn't make this"	Device and IP fingerprint at checkout, account login history, prior order history from the same account	Stronger when the transaction fits the customer's normal purchase pattern and there's a verified account login; weaker for guest checkouts with no identity trail
Duplicate charge	Transaction logs showing two distinct order IDs and timestamps, idempotency key records	Strong when your system logs clearly separate two distinct orders; weak if your own system actually double-billed the same order
Product or service not as described	Product listing archive at time of sale, photos, pre-dispute customer messages	Depends heavily on whether listing language was specific and whether support messages show the customer had accepted the stated terms
Credit or refund not processed	Refund transaction record, processing timestamp, channel confirmation	Strong when the refund timestamp predates the dispute filing; weak if the refund was genuinely delayed past what the customer was told to expect
Subscription cancelled but still charged	Cancellation request record, renewal terms shown at signup, billing history	Strong with clear opt-in to auto-renewal and a timestamped cancellation workflow; weak if the cancellation flow is confusing or undocumented

Notice the pattern: the win-likelihood factor almost always comes down to whether your records were clean before the dispute existed, not how persuasively you argue after the fact. Evidence assembled retroactively is always weaker than evidence already sitting in a well-organized system.

Submitting Evidence: What the Process Looks Like on the Backend

If you're integrated with your payment provider's API rather than working entirely through a web portal, dispute evidence submission is usually its own endpoint, separate from the transaction and refund APIs. Here's a representative shape of what that call looks like:

curl -X POST https://apix.example.com/api/dispute/evidence \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "merid": "MCHXXXXX",
    "dispute_id": "DSP20261003000123",
    "tno": "ORDER20261003000456",
    "amt": "100.00",
    "evidence_type": "proof_of_delivery",
    "evidence_files": [
      "https://cdn.yourstore.com/evidence/pod_order456.pdf",
      "https://cdn.yourstore.com/evidence/chat_log_order456.pdf"
    ],
    "notes": "Courier tracking confirms delivery to billing address; customer support chat shows customer acknowledging receipt prior to dispute filing.",
    "sign": "..."
  }'

A few implementation details matter more than they look like they should:

• The tno (transaction number) must match exactly what the issuer has on record. Pull it directly from your transaction table rather than letting a support agent retype it.
• Evidence files need stable, accessible hosting for the review period, not a temporary link that expires. Exempt dispute evidence from any storage retention policy until the case closes.
• The notes field is where you tell the story — don't leave it blank. A short, factual narrative tying the files together gives the reviewer a map to follow.

Why Merchants Lose Disputes They Should Have Won

The losses that sting the most are the ones where the merchant was factually in the right but still lost on process. The recurring culprits:

Missing the response window. The single biggest cause. Notifications land in a shared inbox or a generic portal alert nobody configured to route anywhere urgent. By the time someone notices, the window has closed and there's no appeal for a late submission.

Submitting incomplete evidence. A merchant assembles a delivery screenshot and calls it done, without matching customer communication or a clearly cross-referenced transaction ID. Reviewers don't go hunting for context you didn't provide.

Evidence that doesn't match on paper, even when true in reality. A timestamp logged in the wrong timezone, a typo'd order ID, an amount off by a rounding difference — small inconsistencies that get a submission rejected on a technicality rather than evaluated on merit.

No single owner for the process. When dispute handling is "whoever happens to see it," cases fall through the cracks. Merchants who consistently win assign a specific person or team to own the response, with a checklist and a deadline tracker.

Generic, boilerplate responses. The same templated paragraph pasted into every dispute, without tailoring the notes field to the specific reason and evidence attached, reads as low-effort and tends to get deprioritized.

Proactive Practices That Reduce Your Dispute Rate

Winning disputes after the fact is damage control. The more durable fix is reducing how often they happen. A few practices that have measurably helped our dispute volume:

• Make your billing descriptor recognizable. Many "unauthorized transaction" disputes aren't fraud — they're a customer who didn't recognize your store name on their statement and panicked. A clear, consistent descriptor matching your storefront name cuts this down.
• Send order and delivery confirmations automatically, with enough detail that the customer has a paper trail in their own inbox, not just in your system.
• Keep support responsive before a dispute gets filed. Most customers dispute because they feel ignored, not because they're defrauding you. A fast reply to "where's my order" resolves the frustration before it escalates.
• Use 3D Secure authentication on card transactions where supported, which shifts liability for certain fraud-related disputes away from the merchant when the authentication step was completed.
• Build a payment risk control framework rather than reacting case by case, flagging mismatched shipping/billing, unusually large first-time orders, or rapid repeat attempts before fulfillment.
• Apply consistent fraud detection rules at checkout instead of manual review, so screening doesn't depend on who's on shift.
• Keep clean, centralized payment records and reconciliation data so pulling the matching transaction, delivery, and communication records takes minutes, not a frantic afternoon.

When It's Worth Fighting (and When It Isn't)

Not every dispute is worth contesting, and treating all of them identically wastes time you could spend on cases you can actually win. A few questions I ask before deciding how much effort to put in:

How complete is the evidence, honestly? If delivery tracking is ambiguous, there's no customer communication, and the order came through a guest checkout with no identity trail, the realistic odds are weak no matter how well you write the notes field. Spend effort where the paper trail is actually strong.

Is this isolated or part of a pattern? A single dispute from a long-standing customer is probably a genuine misunderstanding, sometimes faster to resolve directly than through the formal process. A cluster from new accounts or similar order patterns is a different problem — a fraud or account-takeover vector, and the fix belongs in your dispute and refund procedures and risk rules, not in contesting each case individually.

What does losing cost versus fighting? For low-value transactions, staff time spent assembling a thorough evidence package can exceed the amount in dispute. That's a legitimate business call — just be deliberate rather than defaulting to "always fight" or "never bother."

Frequently Asked Questions

How long do I actually have to respond to a dispute?

It depends on the network, wallet provider, or bank, and the window can change over time, so there's no single fixed number that applies everywhere. What matters is that the window is short enough that "I'll deal with it next week" isn't viable — treat every notification as needing a same-day triage decision.

What happens if I miss the response window?

In most cases the dispute resolves automatically against you, regardless of how strong your evidence would have been. There's typically no late-submission appeal, which is why routing notifications to a responsible owner matters more than almost anything else here.

Can I still contest a dispute after I've already refunded the customer?

If you've already refunded, the dispute is usually moot — you won't win back money you voluntarily returned. The more common trap is the reverse: a customer disputes a transaction you'd already refunded through a different channel, which is why matching transaction details and clean refund records matter so much.

Does working with a BSP-licensed payment partner actually help my case?

Indirectly, yes. A provider operating under BSP OPS and EMI licensing is expected to maintain structured transaction records, proper KYC data, and auditable logs — the same raw material that makes your own evidence package stronger when a dispute lands.

Does informal customer communication, like a chat app message, count as evidence?

Generally yes, as long as it's exportable with a timestamp and clearly tied to the order. Screenshots without timestamps or order references are weaker — build a habit of exporting and archiving chat conversations against an order ID.

Should I always fight a dispute, or is it sometimes smarter to let it go?

Sometimes letting it go is the right call, particularly for low-value transactions with thin evidence. Reserve effort for cases where records are genuinely strong and for patterns suggesting a recurring fraud vector worth shutting down at the source.

Do e-wallet disputes work the same way as card chargebacks?

The underlying logic — complaint raised, merchant gets a window to respond with evidence, decision follows — is similar across e-wallets and card networks, but reason categories, documentation formats, and response windows differ by provider. Treat each channel as its own workflow.

What's the single highest-leverage thing I can do starting today?

Assign one specific person or team as the owner of every dispute notification, with an alert that actually reaches them when a case opens. Nearly every avoidable loss we've seen traces back to a notification that sat unseen, not evidence that genuinely wasn't good enough.

If you're building out your dispute-handling process and want a second opinion on how your evidence workflow should plug into your payment integration, reach out — @LouisSMARTPAY on Telegram or [email protected] by email.