Validation Monitors
Validation Monitors run checks on your systems every few minutes, flagging the IDs of entities that require your attention. Once identified, the monitors provide a complete record of the entity, including the reason it was flagged and links to the system where you can make the appropriate changes.
How Validation Monitors Work
At their core, Validation Monitors are API queries that represent your business rules and/or configuration requirements that must be adhered to in your systems such as an ad server.
For example, when you create a campaign in your ad server they might have to be targeting a specific country or region. Failure to do so may result in the campaign not delivering. A simple mistake to make that can have disastrous outcomes.
With Validation Monitors you simply create a query that represents this requirements, save it and the Watching That system will run the check every few minutes alerting you to any campaigns that don’t meet this requirement.
NOTE: You must have your system API connected to your Watching That account. Please contact the Support team to ensure this is the case.
Creating a Validation Monitor
To create a Validation Monitor you start in the Monitoring Module.
Select Add Validation Monitor
In the Create Validation Monitor screen you need to:
Select the data source to use
Select the dimension you want to validate against (*note this will be unique to the data source)
Create a filter that represents your query / validation
Give it a name and description
Add a notification channel to receive alerts
Validation Monitor Example
In this example we’ve created a monitor to flag any Ad Creatives that are active in our Freewheel MRM ad server that don’t have an External Id set.
For AMCE+, not having external ids set on the ad creatives can cause significant issues in ad delivery and performance. However there are 1,000s of creatives flowing in and out of the ad server every day.
SO this Validation Monitor will take the heavy lifting off the operators shoulders and flag any creatives in the haystack that need attention.
Managing Validation Monitors
All your active Validation Monitors are displayed in the list page.
From here you can see:
The monitor name
The dimension/endpoint it is targeted to
Who, in your organisation, owns it
When the monitor last run
It’s current status
Quick actions you can perform
The number of active monitors your account is entitled to.
The actions bar
Allows you to
Edit
Duplicate
Deactivate
Delete
Your monitors
NOTE: deactivated monitors do not count towards you usage entitlment.
Viewing a Monitor’s Details Page
When a monitor is in the Alert state it means it has flagged at least one entity.
To view this you will look at its Details Page
In this example our Monitor is set to flag any Advertiser in Freewheel’s MRM ad server that doesn’t have an Industry value set. At this moment it has flagged 4 that need our attention.
From this page you can :
Click the ID to be taken straight to the source system, in this case Freewheels' MRM
View the record details
Manually clear the record from the list but it might get flagged again
Manually filter the record so the monitor excludes it from all
Viewing the Record Details
Creating Monitor Queries
Creating the right query can be an iterative process since there can be a lot of nuance in your validation requirements and the available fields you can use are dependent on the system that is being checked.
Follow this process as a starting point:
Describe the check you need carried out in terms of examples you’ve seen in your business operations. (sometimes it helps to answer the question: what would the list of dos and don’ts be for a new hire to your team.)
Next head to the Add Validation Monitor screen and select the Dimension you think best represents this requirement.
Then, in the filter list, scroll the available fields you can use to craft your query. NOTE this list can be VERY LONG so use the search function to zero in.
As you can see in this example, for Ad Unit on Freewheel MRM, there is a very long list of options that include the Placement attached to the Ad Unit.
Tip: Create the monitor as close as you can and then let it run. You can edit the filter query based on the results do or don’t see.
Tip: Create an entity in the source system that should flag your monitor. This way you can verify it is catching what you want it to catch.
If in doubt reach out to the Watching That Support team who will be more than happy to help you craft the perfect query.
100 Results Limit
Validation Monitors call system APIs which, in turn, can be rate limited. This means the Watching That platform can only call the API at a certain frequency.
This limit can be hit if you have numerous active monitors and/or they are complex checks.
Should this limit be hit the results will be capped at 100 until such time the results are cleared.
During the time that the monitor has reached it cap ALL NEW breaches of the query will be ignored.
Tip: If you are are experiencing a Validation Monitor that flags more that 20 results it is likely your query is too general or there is a deeper systemic issue that needs to be addressed.
FAQ / Troubleshooting
I’ve created a monitor but no results are being returned when I think they should
Ensure your API is connected to the Watching That platform. Contact Support to confirm.
The run times of the monitors seem to be longer and longer
The more complex your query the longer it can take to complete. Especially if the query flags 100s of entities in one run. Try and keep Monitors focused and finely tuned.