Every Shopify store owner has dealt with this frustrating situation: you get an order that seems suspicious, but you’re not sure if it’s legitimate or fraudulent. Do you fulfill it and risk a chargeback? Do you cancel it and potentially lose a good customer? If chargebacks do occur, knowing how to submit proper evidence can help you recover lost revenue.
Shopify’s fraud analysis helps by flagging orders as low, medium, or high risk, but then you still have to manually decide what to do with each risky order. This creates a daily headache, especially if you get several suspicious orders.
Shopify Flow can automate these decisions for you, but most store owners don’t realize it exists or don’t know how to set it up properly. Here’s how to use it to handle high-risk orders automatically, so you can focus on growing your business instead of constantly worrying about fraud.
What Shopify Flow actually does
Shopify Flow is a free automation app that can automatically take actions based on what happens in your store. For fraud management, it can automatically capture payments on safe orders, cancel risky orders, and even track customers who frequently return items.
The key benefit is that it works with Shopify’s built-in fraud analysis. When an order comes in, Shopify automatically analyzes it and assigns a risk level. Flow can then take different actions based on that risk level, without you having to manually review every order.
Flow doesn’t replace Shopify’s fraud analysis – it just automates what you do with the results.
Setting it up correctly
Before you create any workflows, you need to change one important setting in your Shopify admin. Go to Settings > Payments and set your credit card authorization to “Manually capture payments” instead of “Automatically capture payments.”
This is crucial because Flow’s payment capture workflows only work if you’re set to manual capture. If you’re on automatic capture, Flow can’t control when payments are taken.
You also need to understand that Flow workflows for order risk should use the “Order risk analyzed” trigger, not the “Order created” trigger. Fraud analysis takes time to process, so workflows that start immediately when an order is created won’t have the fraud analysis results yet.
The most useful workflows for fraud management
Automatically capture safe payments
This workflow captures payment for low and medium-risk orders while leaving high-risk orders uncaptured. This means you get paid immediately for safe orders, but suspicious orders wait for your manual review.
To set this up, go to Settings > Apps and sales channels > Shopify Flow > Create workflow > Browse templates > Risk > “Capture payment if order is not high fraud risk.”
This workflow is probably the most valuable one for most stores because it saves you from manually processing every single order while still protecting you from obvious fraud.
Cancel and restock high-risk orders
This workflow automatically cancels orders that are flagged as high risk and puts the inventory back in stock. It also tags the order and customer, and sends you an email notification.
This is more aggressive than the previous workflow because it actually cancels orders instead of just holding them for review. Only use this if you’re comfortable with potentially canceling legitimate orders that happen to be flagged as high risk.
Track customers who return too much
This workflow cancels orders from customers who have returned items five or more times in the last six months. This helps catch customers who might be abusing your return policy.
This one is useful if you have a generous return policy that some customers might be taking advantage of.
What to expect after setting up these workflows
These workflows will reduce the number of orders you need to manually review, but they won’t eliminate fraud decisions entirely. You’ll still need to review some orders, especially medium-risk ones that don’t get automatically captured or canceled.
The workflows also generate email notifications, so you’ll know when they take actions. You can customize these emails or turn them off if you find them too frequent.
Don’t expect these workflows to catch every type of fraud. Sophisticated fraudsters can often get past automated systems, so you should still pay attention to your orders and chargebacks.
Important limitations to understand
Flow workflows are based on Shopify’s fraud analysis, which isn’t perfect. Sometimes legitimate orders get flagged as high risk, and sometimes fraudulent orders slip through as low risk.
If you set up the workflow that automatically cancels high-risk orders, you will occasionally cancel legitimate orders. You need to be comfortable with this trade-off.
The workflows also don’t work for orders placed through some payment methods like PayPal, which have their own fraud protection systems.
Getting started
Start with the “Capture payment if order is not high fraud risk” workflow. This gives you the biggest benefit with the lowest risk of problems. You’ll still capture payments on most orders automatically, but suspicious orders will wait for your review.
After you’ve used that workflow for a few weeks and you’re comfortable with how it works, you can consider adding the other workflows.
Remember to monitor your chargebacks and order cancellations after setting up these workflows. If you notice any concerning patterns, you can always adjust or turn off the workflows.
The bottom line
Shopify Flow can significantly reduce the time you spend manually reviewing orders for fraud, but it’s not a complete solution. You’ll still need to use your judgment for some orders, and you’ll need to monitor the results to make sure the automation is working well for your business.
The biggest benefit is that it handles the obvious cases automatically – capturing payments on clearly safe orders and flagging clearly risky ones. This frees up your time to focus on the borderline cases that actually need human judgment.
If you’re currently manually reviewing every order, these workflows will save you significant time. Just make sure you understand how they work and monitor the results regularly.
Author
