# RefundSentry > Fraud and risk intelligence for Shopify orders, returns, refunds and disputes. RefundSentry is an intelligence layer, not a return-management system. It works alongside Shopify native returns, Loop Returns, AfterShip Returns and ReturnGO. ## What it does - Scores order and return behavior without adding checkout latency - Surfaces corroborated evidence and a 0–100 risk score - Can hold risky fulfillment for merchant review - Auto-tags customers by risk level in Shopify - Flags repeat refunders, worn-once returns, refund-method switches, chargeback risk and fraud rings - Provides Shopify Flow triggers and a Shopify Admin order-page extension ## Links - Website: https://refundsentry.com - Features: https://refundsentry.com/features - Pricing: https://refundsentry.com/pricing - Full documentation: https://refundsentry.com/llms-full.txt ## Current offer RefundSentry is generally available for Shopify merchants. - 80+ registered fraud and risk signals across orders, returns, refunds and disputes. - Audit up to 12 months of available history; exact available history is shown after install. - Starter: $49/month, up to 1,000 rolling-30-day orders, 14-day trial. - Growth: $99/month, up to 3,000 rolling-30-day orders, 14-day trial. - Scale: $199/month, up to 6,000 rolling-30-day orders, 14-day trial. - Advanced: $299/month, up to 12,000 rolling-30-day orders, 14-day trial. - Install: https://apps.shopify.com/refundsentry - Example report: https://refundsentry.com/the-report - Support: https://apps.shopify.com/refundsentry#support; email fallback support@refundsentry.com. ## Original research - [Return fraud statistics 2026: 12 months of real Shopify refund data](https://refundsentry.com/research/return-fraud-statistics): 73,174 orders, 2,116 refunds and 861,588 EUR of refunded value from one Shopify store, analyzed and published with null results. Free to cite with attribution (CC BY 4.0). ## Blog (45 published articles) - [Return rate by category: what's healthy, what's a product problem, what's fraud](https://refundsentry.com/blog/return-rate-by-category-benchmark): A 20% return rate on jeans is normal. A 20% return rate on lamps is a crisis. Here's how to read your return rate by category and figure out whether you have a product problem, a fit problem, or a fraud problem. - [Data minimization for fraud scoring (hashes, not PII hoarding)](https://refundsentry.com/blog/data-minimization-fraud-scoring-architecture): You can build effective fraud detection without storing raw email, phone, or address data. Here's the architecture. Hash at the edge, correlate on digests, keep PII only where it legally has to live. An engineering-depth look at how RefundSentry does it. - [Gift returns: the fraud vector nobody designs for](https://refundsentry.com/blog/gift-returns-fraud-vector-nobody-designs-for): When a recipient returns an item they didn't order, your fraud signals go blind. Here's why gift returns are the hardest return category to scrutinize, and the specific patterns that separate legitimate gift exchanges from a playbook abusers rely on. - [Per-merchant machine learning on small data: how 50 refunds a month becomes a working model](https://refundsentry.com/blog/2026-07-07-per-merchant-ml-on-small-data): Eight chargebacks a year is not enough to train an ML model. Fifty refunds a month is. Here is how reframing the label from chargebacks onto refunds, and choosing per-merchant XGBoost over a global pooled model, makes a working ML system on the volume a normal Shopify merchant has. - [70 ways customers commit return fraud and what we look for](https://refundsentry.com/blog/2026-06-30-seventy-signal-catalog): Most fraud-detection tools sell AI without saying what the AI looks for. Our signal catalog is open. Here are 25 of the most distinctive signals across 9 clusters, and the math that holds them together. - [Staff-issued refund leakage: the fraud coming from inside your own customer service](https://refundsentry.com/blog/staff-issued-refund-leakage): Your customer service team issues refunds all day. Most are legitimate, some are policy exceptions, and a small percentage are drifting outside what you'd approve if you were watching. Here's how to find them without turning your support team into suspects. - [Network-effect fraud detection: every merchant makes every other merchant safer](https://refundsentry.com/blog/2026-06-23-cross-shop-network-signals): Most fraud-detection apps treat each merchant as an island. Their data does not flow between merchants. RefundSentry's database is built so it does. Here is the structural choice that makes the network effect real. - [Returns and refunds aren't the same thing, and your fraud model needs to know](https://refundsentry.com/blog/2026-06-16-returns-vs-refunds-ml-label): If your fraud tool only sees Shopify Returns, it is blind to half your refund flow. Loop, Returnly, and AfterShip merchants never produce native return records. Here is why we use refund data, not return data, as the universal fraud label. - [The real cost of a chargeback (the $15 fee is about a tenth of it)](https://refundsentry.com/blog/real-cost-of-a-chargeback): The fee on the dispute is the cheapest part. Here's what a chargeback actually costs a mid-market Shopify store once you count representment labor, CSAT damage, and the monitoring-program risk nobody talks about. - [Return-fraud rings: detecting customers who don't think they're working together](https://refundsentry.com/blog/2026-06-09-fraud-rings-graph-detection): Three customer accounts, two addresses, one operator. The pattern that small DTC merchants see every quarter and don't have language for. Here is the graph-shaped data model that makes them visible. - [Why small orders are where card-testing fraud hides](https://refundsentry.com/blog/2026-06-06-card-testing-rings-small-orders): Multiple small orders, different stolen cards, one freight-forwarder address. Per-order fraud scoring misses it. Here is the pattern, why it slips through, and what the engine actually scores when you look at the address instead of the order. - [We rebuilt our risk engine, here's what was wrong with v1](https://refundsentry.com/blog/2026-06-02-rebuilding-the-risk-engine): The first version of our scoring engine produced a dashboard where eight fraud chargebacks all sat in the LOW zone. The problem was not the model. It was the absence of a per-merchant baseline. Here is what we tore out, what we kept, and how the calibrated v2 formula works. - [Bracketing: when buying five sizes is normal and when it's abuse](https://refundsentry.com/blog/bracketing-five-sizes-normal-vs-abuse): Your apparel customer just ordered the same dress in sizes 6, 8, and 10 with plans to return two. That's either a smart shopper or a drain on your margins, and the difference depends on what happens next. - [Day one of using a fraud tool shouldn't be a blank dashboard](https://refundsentry.com/blog/2026-05-26-day-zero-backfill-insights): Most fraud tools install in five minutes and then ask you to wait three months for any data to accumulate. Here is how RefundSentry backfills 12 months of historical orders, refunds, and chargebacks at install time so the first dashboard view is the merchant's own historical fraud picture. - [Why we store every webhook for a year and what we do with it](https://refundsentry.com/blog/2026-05-19-raw-event-store-365-day-retention): A 365-day raw event store sounds like overkill until you try to investigate a fraud chargeback that landed three months ago. Here is what we keep, why we keep it, and what we do with it. - [The hidden cost of labeling fraud twice](https://refundsentry.com/blog/hidden-cost-labeling-fraud-twice): Your fraud tool asks you to mark returns as fraud. Your Shopify workflow already requires you to decline, close, or dispute the same returns. That's the same decision, asked of you twice, and it quietly breaks every feedback loop on the market. - [The anatomy of a repeat return fraudster](https://refundsentry.com/blog/2026-05-12-anatomy-of-a-repeat-return-fraudster): A walkthrough of one merchant's actual investigation. Five customer accounts, two addresses, three months. What the data showed and which signals would have caught it earlier. - [Fraud ring detection beyond the obvious signals](https://refundsentry.com/blog/2026-05-09-fraud-ring-detection-beyond-obvious-signals): A small fraud ring is invisible to per-customer signals by design. Here is how velocity windows, address clusters, and confirmed-fraud propagation expose them. - [Why your Shopify chargeback data is not enough](https://refundsentry.com/blog/2026-05-07-shopify-chargeback-data-not-enough): Shopify hands you the dispute. It does not connect the dots between disputes that share an address, email, or phone. Here is what that gap costs and what closes it. - [Email and phone recycling in return fraud](https://refundsentry.com/blog/2026-05-04-email-phone-recycling-return-fraud): A new Shopify customer account resets every history counter. The email hash and phone hash do not. Here is how cross-account pivots catch fraud the customer record cannot. - [How fraudsters reuse addresses after a chargeback (and what your store can do about it)](https://refundsentry.com/blog/2026-05-01-fraudsters-reuse-addresses-after-chargeback): A chargeback ends one customer's relationship with your store. The address keeps shipping. Here's the pattern and the signal that catches it. - [Wardrobing detection: how to catch worn-and-returned items](https://refundsentry.com/blog/wardrobing-fashion-invisible-fraud): Return reasons can't separate a customer who didn't love the fit from one who wore the dress to a wedding. The signals that detect and prevent wardrobing without rejecting real customers. - [How to audit six months of return fraud without hiring a data team](https://refundsentry.com/blog/audit-six-months-return-fraud): You suspect fraud is hiding in your past returns but have no way to look backwards. Here's what a real retrospective fraud audit requires, and how to skip the SQL. - [Signals v3: 13 new fraud signals for identity, network, and cashout abuse](https://refundsentry.com/blog/signals-v3-identity-network): RefundSentry's biggest scoring release yet. 13 new return-fraud signals covering multi-account identity laundering, shared-address clusters, gift card cashout, and review-then-return abuse. - [Why is my Shopify return rate so high? A diagnostic](https://refundsentry.com/blog/why-is-my-shopify-return-rate-so-high): A 5-step diagnostic for Shopify merchants staring at a return rate that keeps climbing. Walk through fit, product, seasonality, and fraud before you decide what to change. - [Why some Shopify customers order then return constantly (and how to spot the pattern)](https://refundsentry.com/blog/shopify-customers-ordering-and-returning): A small fraction of customers drive most of your return losses. Here's how to identify them in your Shopify data without building anything, and what to do once you find them. - [What's a normal chargeback rate on Shopify? (And what triggers the 1% suspension)](https://refundsentry.com/blog/shopify-chargeback-rate-benchmark): Benchmarks for Shopify chargeback rates by category, what the 1% Shopify Payments threshold actually means, and what to do if you're trending above it. - [Why returns are eating 15-30% of DTC margin (and the costs merchants don't count)](https://refundsentry.com/blog/returns-eating-dtc-margin): A full-cost breakdown of what a return actually costs a Shopify merchant. Product, shipping, refund, labor, restocking, resale loss, and why the 'refund amount' is less than half the real cost. - [Protecting DTC margin from returns: 5 levers most merchants don't pull](https://refundsentry.com/blog/protecting-dtc-margin-from-returns): Return rates climb, margins shrink, and most merchants reach for the same two levers (tighten policy, raise prices). Here are five others that move the number without hurting conversion. - [How to reduce returns on Shopify (without hurting conversion)](https://refundsentry.com/blog/how-to-reduce-returns-on-shopify): Four levers Shopify merchants can pull to cut return rates without killing the conversion lift free returns bought you. Pragmatic, ranked by effort and impact. - [5 Return Analytics Shopify Doesn't Give You (How to Get Them)](https://refundsentry.com/blog/return-analytics-shopify-missing): Shopify shows what was returned, not why. The 5 return analytics it's missing, what each blind spot costs you, and how to get them without switching platforms. - [The ultimate guide to Shopify return fraud in 2026](https://refundsentry.com/blog/shopify-return-fraud-guide): A comprehensive breakdown of return fraud tactics, real-world costs, and proven detection strategies for Shopify merchants. - [Pre-ship order scoring: stop fraud before you fulfill](https://refundsentry.com/blog/pre-ship-order-scoring): How multi-signal order scoring lets Shopify merchants hold risky fulfillments and prevent losses before they happen. - [6 pre-built Shopify Flow templates for return fraud prevention](https://refundsentry.com/blog/shopify-flow-templates-fraud): RefundSentry now ships 6 ready-to-use Shopify Flow templates that automate return fraud response, from holding risky orders to alerting your team on Slack. Install in one click. - [Customer intelligence: know your riskiest customers at a glance](https://refundsentry.com/blog/customer-intelligence-page): How RefundSentry's Customer Intelligence page and Shopify Admin risk card give merchants a complete fraud profile for every customer: return history, chargebacks, fraud rings, and blocklist status in one view. - [Chargeback prevention on Shopify: predict disputes before they happen](https://refundsentry.com/blog/chargeback-prevention-shopify): How Shopify merchants can reduce chargebacks by predicting disputes before shipping, catching double-dip fraud, and building evidence before claims arrive. - [Inside organized return fraud rings](https://refundsentry.com/blog/fraud-rings-explained): How professional fraud networks target e-commerce merchants, the tactics they use, and the specific signals that expose them. - [Disposable emails and fake customers: the simplest fraud signal you're ignoring](https://refundsentry.com/blog/disposable-email-fraud): How disposable email addresses like Mailinator and Guerrilla Mail are the easiest fraud indicator to detect, and why most Shopify merchants miss it entirely. - [Address Red Flags: 4 Geographic Signals of Return Fraud](https://refundsentry.com/blog/geographic-fraud-signals): Billing vs. shipping mismatches, reshipping warehouses, and sudden country changes: the geographic patterns that expose return fraud other signals miss. - [Emerging return fraud patterns you need to know](https://refundsentry.com/blog/return-fraud-patterns): A data-driven breakdown of the newest return fraud tactics, how to identify them, and which defense strategies actually work. - [Return reason clustering: why your return data is hiding the real problems](https://refundsentry.com/blog/return-reason-clustering): How AI-powered reason clustering groups free-text return reasons into actionable categories (sizing issues, quality defects, coordinated fraud) so Shopify merchants fix root causes instead of chasing individual complaints. - [How AI is changing fraud detection in e-commerce](https://refundsentry.com/blog/ai-fraud-detection): A practical breakdown of how machine learning detects return fraud that rules miss, and why the shift matters for Shopify merchants. - [Refund method tracking: how to measure your fraud prevention ROI](https://refundsentry.com/blog/refund-method-tracking-roi): How tracking whether refunds are issued as cash, store credit, or exchange lets Shopify merchants measure the real revenue impact of fraud detection, and prove ROI. - [How to Flag Risky Customers in Shopify (Auto-Tagging Guide)](https://refundsentry.com/blog/auto-tag-risky-customers): Flag high-risk customers in Shopify automatically: risk scores, tagging rules, Shopify Flow and Klaviyo workflows, and when a tag should become a block. - [How Much Does a Return Actually Cost? $30 to $60, Not $7](https://refundsentry.com/blog/return-policy-costs): A line-item breakdown of what each return really costs ($30 to $60 for most stores), and when a no-questions-asked policy helps conversion vs. destroys margin.