Noibu's Alerts feature is immensely flexible, providing more than 15 different criteria you can use to create specific types of error alerts. While you can create any number of alerts, we recommend setting up the following alerts at a minimum. These alerts have proven effective for other eCommerce businesses in identifying high-impact issues.
- Recommended Alert: Closed-Fixed issue is reoccurring
- Recommended Alert: Third-party payment gateway failure
- Recommended Alert: HTTP errors indicating broken functionality
- Recommended Alert: User-caused error in cart
- Recommended Alert: Broken image on key pages
- Recommended Alert: New or existing spiking issues
- Recommended Alert: New verified issues
Note that in addition to these recommended Alerts, Noibu sends weekly emails to notify you of your most important verified issues, priority issues, and performance trends. Learn more here.
🛎️ Recommended Alert: Closed-Fixed issue is reoccurring
This alert notifies you of issues that had been fixed previously and you thought were resolved, but which have started occurring again. Since these issues had previously been reviewed, prioritized, and deemed worth fixing, they should be investigated promptly!
Set the alert to trigger when ALL conditions are met:
Condition | Setting | Value(s) |
State | Is | Closed-Fixed |
Sessions | Greater than | 5 in the last day or whatever threshold makes most sense for your needs |
🛎️ Recommended Alert: Third-party payment gateway failure
This alert helps catch error occurrences specifically related to failing third-party payment gateways. A sudden spike of this type of error could indicate a problem with payment processing. These issues should be investigated promptly to avoid potential order and revenue loss.
Set the alert to trigger when ALL conditions are met:
Condition | Setting | Value(s) |
Error Signature | Contains |
paypal or other third-party providers like adyen, stripe, minicart, etc. |
Sessions | Greater than |
10 in the last day or whatever threshold makes most sense for your needs |
Additional third-party error detection
Note that this same approach can be used to create alerts for failures with other categories of third-party applications. Here are some common examples of terms you might use in the “Error Signature Contains” condition.
Payments | paypal, adyen, stripe, minicart, affirm |
Search | algolia, elasticsearch, searchspring |
Content | wordpress, drupal, joomla, wix, contentful |
Personalization | nosto, barilliance, bloomreach |
Reviews | yotpo, bazaarvoice, trustpilot |
Security | fraudlabs, kount, signifyd |
Support | zendesk, freshdesk, intercom, livechat, helpscout |
🛎️ Recommended Alert: HTTP errors indicating broken functionality
This alert detects recent spikes in several types of HTTP errors. HTTP requests are one of the key ways your site communicates with its various components—like login requests, product catalogues, and more. These types of errors should be investigated to ensure no critical functionality has broken.
Set the alert to trigger when ALL conditions are met:
Condition | Setting | Value(s) |
HTTP status code |
Is |
404, 500, 502, 503, 504 |
Insight tags |
Include Any of |
New |
Conversion Impact |
Is Not |
No Impact, Unlikely |
Sessions |
Greater than |
50 in the last day or whatever threshold makes most sense for your needs |
Note: the Conversion Impact field acts to filter out lower-impact errors. If the alert seems to produce too much noise, you could also consider adding Conversion Impact Is Not Unknown to the condition.
🛎️ Recommended Alert: User-caused error in cart
This alerts catches instances where an issue seems to have been caused by a user interaction on the Cart page.
Set the alert to trigger when ALL conditions are met:
Condition | Setting | Value(s) |
Top URLs | Contains | /cart or whichever URL slug indicates your cart page |
Insight tag | Include Any of | Users Clicked |
Additional user-caused errors
This same approach can be used to discover user-caused errors on other pages too. For example, change /cart to /checkout to be alerted of user-caused errors on the Checkout page.
🛎️ Recommended Alert: Broken image on key pages
This alerts catches instances where an image failed to load on product listing pages or product details pages, which could deter a customer from placing an order.
Set the alert to trigger when ALL conditions are met:
Condition | Setting | Value(s) |
Error type | Is |
Image |
Page type | Include Any of |
Collection, Product |
Sessions | Greater than |
50 in the last day or whatever threshold makes most sense for your needs |
🛎️ Recommended Alert: New or existing spiking issues
This alerts detects issues that have begun to spike in occurrences over the past week. It can be helpful to catch unexpected errors introduced with a recent release, or to detect pre-existing issues that have escalated in severity.
Set the alert to trigger when ALL conditions are met:
Condition | Setting | Value(s) |
Occurrences change | Increased by at least |
20 percent in the last day |
Sessions | Greater than |
20 in the last day or whatever makes most sense for your needs. |
Additional spiking alert criteria
Consider layering additional criteria on this alert to monitor for spiking issues in a more targeted way. For example, you might add a Page Type or Top URL filter to specifically detect spiking checkout issues. Or, you could add the Occurrences Change condition into any of the other recommended alerts above to create a more nuanced alert. Learn more about Alert Conditions.
🛎️ Recommended Alert: New verified issues
While Noibu sends out weekly email updates to let you know about your top Priority Issues to investigate, this alert gives you more timely notification about new issues that have recently been verified by Noibu technical experts.
Condition | Setting | Value(s) |
Insight tags |
Include Any of |
New |
Conversion Impact | Is Any of |
Verified |
Sessions | Greater than |
50 in the last day or whatever makes most sense for your needs. |
Other ways to use Conversion Impact condition
Consider adding a Conversion Impact condition to other alerts in your account to act as a noise filter, rather than adjusting the quantitative thresholds (e.g. sessions, occurrences, revenue loss) you've set. For example, adding Conversion Impact IS NOT Unlikely, No Impact will filter out lower-impact issues that could otherwise trigger an alert.