Sift Science Fraud Prevention App (English)

With AdVision's Sift Fraud Prevention app, any time a new order is placed, updated or paid for in your webstore, the order information is forwarded to your Sift account for a fraud analysis. Sift uses it's unique fraud prevention system to analyize your customer's risk level (how likely is this purchase fraudulent) and returns a score to indicate the risk of fulfilling the order.


Follow instructions in Lightspeed eCom to purchase the Sift Science Fraud Prevention app from the App Marketplace.  During install, you'll be forwarded to our app management site to either login to an existing account (if you've purchased a Lightspeed eCom app from us before) OR create a new account with us to install and manage this and any future apps you may purchase from us for Lightspeed eCom.

Once you've created an account or logged into your existing account, the app installs automatically, after which you'll be forwarded to the shops list, which displays a list of Lightspeed eCom shops registered with your account and any apps installed with those shops.  To manage this app, click on it's name "Sift Science Fraud Prevention)" under the appropriate shop.

Before you can start using the app, you need to connect it to your Sift account, which you can do under the Setup tab. 

Under Setup, so you'll be directed to either provide the connection information for your existing Sift account, or you can fill out a form in the app and we'll create a new Sift account for you.  In the Setup form, you fill out the following:

  1. Already have a Sift Account?: To use this app, you're  required to have an account with Sift. 

    If you already have one, select "I already have a Sift Account", after which you need to fill in:

    a. Sift Account ID: Login to your Sift account, go to the profile page and get your Production Account ID: https://siftscience.com/console/account/profile
    b. Sift API Key:  Once logged into your Sift account, go to Developer>>API Keys and copy the 'REST API Key' value: https://siftscience.com/console/developer/api-keys?abuse_type=payment_abuse

    Once those fields are completed, click "Save Sift Account Info" and your app will be setup to connect to your Sift Account.  You can then continue to the Settings tab.

    If you don't have a Sift account yet, select "I don't have a Sift Account", after which you need to fill in:

    a.Site Domain: Enter the domain name for your webstore. Example: advision-ecommerce.webshopapp.com
    b. Password: Enter the password you'd like to use for your Sift account.
    c. Site Email: Enter the main email address for your webstore
    d. Analyst Email: Enter the email you'd like to receive reports from your Sift account.

    Once you've filled out those fields, click "Create Sift Account" and the app will attempt to create an account with Sift Science for you.  If an account is successfully created, your app will be setup to connect with that new Sift Account, so you can continue to the Settings tab.

Once the app is connected to your Sift account, you can complete the settings form to setup a Javascript snippet in your webstore, indicate your Risk Threshold score, and set the base currency being used in your webstore.  In the settings, fill in the following:

  1. JS Snippet Key *: The Javascript Snippet Key is used to install some Javascript on your webstore that allows for additional reporting and analysis from Sift.  Login to your Sift Account to get your Javascript Snippet Key, available at the link below (once logged in). https://siftscience.com/console/developer/api-keys?abuse_type=payment_abuse

  2. Risky Threshold: Set your risk threshold score. Any event (order created/paid) above this score, you'll be notified by email.  If you're unsure what score to set, consult Sift to determine what maximum score you're comfortable with.

  3. Base Currency Code: Set the base currency code that is being used in your Lightspeed eCom Shop (must match for accuracy).

If you make any changes to any of these fields, be sure to click "Save Settings" button to save those changes.


*If this is your first time setting the JS Snippet Key, or if you're changing the JS Snippet Key, after you click "Save Settings", you'll need to ensure the webstore is updated with new Javascript.  To do so, after you save your settings, find the button below the form that will say "Finalize Install Now" (if this is your first time setting up the app) or "Check Install/Publish Changes" (if app has been settings have previously been setup).  Click this button to ensure the Javascript is properly installed/updated in your webstore.

Once you've completed the settings form, you're done! Future new, updated and paid order events in your webstore will be scored by Sift and any that exceed your risk threshold will be logged in the app (with a link to a detailed report in your Sift account) and you'll be notified via email. 

You can check the Summary tab for a quick view of recent events and alerts, or for a full list of events and alerts in your app, use their respective tabs: "Events" and "Alerts". 

Under Events, the app will log any events that exceed your Risk Threshold (events under the risk threshold will be not be logged in the app), and with each event you can click the "View Customer in Sift" to get a detailed report and decide if you want to proceed with fulfilling that customer's order.  You can also search the Events log by Order Id or Customer Id (system generated numeric order and customer Id's from eCom).

Under Alerts, the app will let you know if something failed any why.  Commonly errors are logged here if an event fails to be scored by Sift.  The alerts in the app are just to let you know a problem occured, so you can decide if you want to go into your Sift account to manually score an order.  If you're unsure about an error being reported, please check our FAQs and if you can't find an answer there, you can contact us at info@advision-ecommerce.com

Sift Science App: Frequently Asked Questions

Q. Is the app built into our Lightspeed dashboard or is it its own separate account on another website where we would access from Also will the report Sift generates show up in the orders section of our Lightspeed dashboard?

Ans. No, it doesn't show in any Lightspeed dashboard.  They can see alerts and basic info of those alerts in the app dashboard (external from LS) or more details reports in their Sift account.

Q. With the automations and triggers that we set, is there any option for declining the transaction/order automatically, or does it simply highlight and rank the orders for us to more easily find in Lightspeed?

Ans. No, it won't auto decline, just alerts if the order exceeds fraud score threshold.  Nothing shows "in Lightspeed".

Q. Do you offer support with setting up the snippets and settings in the account if we are less than tech savvy?

Ans. Yes, we can help setup the app with you.

Q. Does the app retroactively pull data from previous orders to help determine what may or may not be fraudulent?

Ans. It only works for orders that are created, updated or paid after the app is installed.  So if a previously placed order is updated or paid for after the app is installed, it would detect that and check it's risk.  But if the previously placed order doesn't create any events after app is installed, app will never be aware of them and neither will Sift.

Q. How many factors are involved in the risk report or is that depend on how we set up our account?

Ans. That's a better question for Sift as they handle the risk calculation.  Our app just creates the connection and logs/notifies when an event exceeds risk score.

Q. How sensitive is the ranking system? How likely is it likely we would cancel orders that are not fraudulent?

Ans. I'd suggest you speak with Sift and review how they calculate risk.  How sensitive our app is depends on the "risk threshold score" you set.  Determining what risk threshold you're comfortable would best be decided by playing with Sift's fraud tool and see what scores you get back for different order info, then choosing a score that ignore orders that you know are OK but alerts you for orders that may be fraudulent.