One type of custom metrics returns data about course completion per course. There is no date field in the result:

[
	{
		"username": "user1",
		"email": "person1@abc.de",
		"course_id": "course_1",
		"completion": "0.39"
	},
	{
		"username": "user9",
		"email": "person9@mmm.nn",
		"course_id": "course_2",
		"completion": "0.75"
	},
	{
		"username": "user14",
		"email": "person14@scp.fr",
		"course_id": "course_1",
		"completion": "1.00"
	},
...

I'm parsing this in the Databox metric like this:

(  
  $$.{
	  "value": $number(completion) * 100,
	  "dimension": "EDM (en)"
  }
)

The only function of "dimension" here is to set the default name of the course - I can leave it out and it works the same.

When previewing this metric, the date below provides a sum of the completion and the number of rows, so I can easily calculate the average value myself. For this metric, it is around 4819 / 71 =~ 67.9. In Advanced options -> Metric Definition -> Data aggregation i checked handling multiple values stored for identical date/time. I left the other checkboxes unticked.

However, if I preview the metric results, it displays 81 as result:

It gets even more optimistic in the databoard where I use the metric:

Since I couldn't find out what kind of logic applies here, I tried doing the average in Jsonata itself. Pasting the JSON in the Jsonata Exerciser I arrived at this query:

  $average($.$number(completion)) * 100

which correctly gave me

67.87323943662

I tried to use this in the metric jsonata, fir which I first simplified the query to

 $$.{
	  "value": $number(completion) * 100,
	  "dimension": "EDM (en)"
 }

.. which gives at least the same table of results below. But when I pressed Preview on this one, it suddenly displayed the correct average:

Great! So I saved the metric, purged the old data and refreshed the Databoard:

Now that's quite funky .. I've had several other numbers in between before I started writing this. Not sure what's going on here, but it looks as if databox has trouble handling a metric without a time dimension - maybe because it sets all rows to now ?

Are you aware of any such issues, or better even, how to fix it?

Thanks in advance,

Maike Dulk
Europeana

I’ve been setting up dashboards in databox that pull in daily metrics, but sometimes the latest updates don’t appear immediately, and certain data points seem delayed. I’m trying to figure out whether this is usually a sync issue, api refresh rate, or something else.

I ran into a scenario similar to a telenor quiz today app workflow, where content changes daily and timing matters a lot, and it made me wonder if scheduling updates or caching could be causing the delays.

Has anyone experienced similar issues with databox dashboards? How do you usually troubleshoot delayed or missing data from sources that update frequently?

[ 
  {
     "username":     "learner01",
     "email":        "learner01@domain.edu",
     "course_id":    "IIIF_101",
     "is_active":    1,
     "mode":         "audit",
     "last_login":   "2025-12-02",
     "created_date": "2024-01-29"
   },{
     ...
     ...  (repeated for every course enrolment, for every user)
     ...
   }
]

This feed contains an entry for every course enrolment of all users. It wasn't difficult to create a custom metric that retrieves these data, and build a datablock visualisation with it:

Because there are around 30 different courses available, it often happens that users enrol in more than one course, and therefore we'd like to distinguish between the total number of enrolments and the total number of active users on our course platform.

Unfortunately, the course provider has no API endpoint that provides the total number of users. It shouldn't be hard to derive that from the course enrolments endpoint though, by creating a rolling sum of the number of unique usernames. However, I haven't been able to find a way to accomplish that in Databox, for instance by selecting a field for grouping in the metric.

Then I tried to accomplish this in the JSONata expression. I found that below expression removes all duplicate users when testing in https://try.jsonata.org. It first sorts the feed on created_date, then removes all results where the email address occurs more than once:

(
$reduced := $reduce($sort($, function($l, $r) { 
          $l.$toMillis(created_date, '[Y0001]-[M01]-[D01]') > 
          $r.$toMillis(created_date, '[Y0001]-[M01]-[D01]')}), 
          function ($acc, $item) {(
              $alreadyExistsInAcc := $acc[email = $item.email];
              $alreadyExistsInAcc ? $acc : $append($acc, $item)
              )},[]);

    $reduced{
	    "value": $reduced.is_active,
	    "date": $reduced.created_date,
	    "dimension": $reduced.username
    }
)

I tried using this expression in the metrics editor, but it does not remove the duplicate users there. It doesn't produce an error either - it just gives all results, ignoring the $reduce function altogether.

Still, I can't imagine that something this basic cannot be achieved in Databox, so I'm probably overlooking something. Can anyone point me to a solution?

Much appreciated,

Maike Dulk
Europeana

Step 1: Create Custom Metric 1: Total Ad Spend

Since the standard Facebook Ads connector usually provides this, you might not need to create it, but you'll use the Metric Builder to select and potentially segment it if you're pulling from a complex source or need a specific filter. If it's a standard metric, you can skip to Step 3, but for a true custom pull, follow this:

1. Navigate to Metrics: Go to the Metrics section in your Databox account.

2. Access Custom Metrics: Click on Custom Metrics (or Metric Builder).

3. Create New: Click the + New Custom Metric button.

4. Data Source: Select your connected Facebook Ads (or Facebook Ads) data source.

5. Metric: Select the base metric for Spend (or the most relevant total cost metric).

6. Filter/Dimension (Optional): If you only want to track spend for a specific campaign or ad set, apply the necessary filters/dimensions here.

7. Name: Name the metric clearly, e.g., "FB Ads - Total Spend".

8. Save: Click Save.

Step 2: Create Custom Metric 2: Cost Per ThruPlay

Similar to Spend, this is likely a standard field, but you may need to use the Custom Metric builder to select and filter it.

1. Navigate to Metrics: Go back to the Custom Metrics area.

2. Create New: Click the + New Custom Metric button.

3. Data Source: Select your connected Facebook Ads data source.

4. Metric: Select the base metric for Cost Per ThruPlay.

5. Filter/Dimension (Optional): Apply the same filters/dimensions as you did for the Spend metric to ensure both metrics cover the same data set.

6. Name: Name the metric clearly, e.g., "FB Ads - Cost Per ThruPlay".

7. Save: Click Save.

Step 3: Create the Calculated Metric: Number of ThruPlays

This step uses the metrics created (or verified) in the previous steps to perform the reverse calculation.

1. Navigate to Calculated Metrics: Go to the Metrics screen and then select the Calculated Metrics tab.

2. Create New: Click the + New Calculated Metric button.

3. Enter the Formula: You will build the equation:

   Number of ThruPlays = FB Ads - Total Ad Spend \ FB Ads - Cost Per ThruPlay

   • Variable A: Select your "FB Ads - Total Spend" Custom Metric.

   • Operator: Select the Division (/) operator.

   • Variable B: Select your "FB Ads - Cost Per ThruPlay" Custom Metric.

4. Name: Name the resulting metric clearly, e.g., "Calculated ThruPlays".

5. Format: Ensure the metric format is set to Number (not Currency or Percentage) with the desired decimal places, as this represents a count, e.g., "#,###".

6. Preview and Save: Review the data preview on the right side to ensure the results look accurate, and then click Save.

Your new "Calculated ThruPlays" metric is now available to use on any Databoard, complete with historical tracking based on the underlying metrics.

You can select all Databoards included in a Loop and copy them to a client account in bulk, saving you time and repetitive work.

How to Copy Multiple Databoards to a Client Account

1. Open the Looped Databoards view in your Agency account.

2. Click the blue Add to Loop button.

3. From the dropdown, select Copy to client account.

4. Choose the destination client account.

5. All selected Databoards will be copied at once.

Instead of opening and copying each Databoard individually, this method allows you to transfer entire sets of reports in just a few clicks which is ideal for onboarding new clients or replicating reporting templates.

For more on this feature, visit the How to Use Databoard Loops guide.

Fortunately, GA4 provides a few built-in ways to filter out this noise, which will also clean up your Databox reports.

Steps to Exclude Bot and Spam Traffic in GA4

1. Automatically Exclude Known Bots

   • Go to Admin > Data Collection and Modification > Define Internal Traffic.

   • Create a rule for internal IPs or enable the built-in bot filtering option (usually auto-enabled).

   • This helps GA4 automatically filter out known bot activity.

2. Block Spam Referrals

   • In Admin > Data Streams, select your web stream.

   • Scroll down to More Tagging Settings and open List Unwanted Referrals.

   • Add known spam or bot domains to exclude them from future reports.

3. (Optional) Exclude Traffic in Databox via Custom Metrics

   • When building Custom Metrics in Databox, review the Source/Medium or Referral Dimensions.

   • Add filters to manually exclude known spam or bot sources (e.g., suspicious domains or irrelevant sources).

Filtering bot traffic ensures your Databox Metrics reflect real user behavior, leading to more reliable decision-making. Clean data is especially important for Metrics like Sessions, Bounce Rate, and Conversion Rate, which are easily distorted by non-human visits.

For more help, visit the Google Analytics 4 Integration guide.

Databox offers flexible sharing options so stakeholders can access the insights they need without requiring a login.

Two Ways to Share Databoards Without Adding Users

1. Public Live Link

   • Share a live URL that anyone can access without logging in.

   • Ideal for executives, clients, or team members who just need view-only access.

   • They can bookmark the link to check metrics anytime.

2. Scheduled Snapshots (PDF)

   • Automate recurring email reports with PDF snapshots of your Databoard.

   • Each email can also include the public live link for quick access.

   • Useful for regular updates to non-Databox users.

When to Add Users

If someone needs to build Databoards, edit Metrics, or manage Integrations, you’ll want to add them to your Databox account and assign an appropriate user role.

Learn more about user roles here.

1. Configure the dataset

Ensure that you have the following views and columns included in your dataset. Read more about how to create a dataset.

a) Include these columns in the Campaign Reports Dataset:

b) Include these columns in the Campaigns Dataset:

d) Merge Datasets together:

Read more about how to merge datasets.

2. Create the Custom Metric

1. Navigate to Metric Builder and create a new Custom Metric by selecting the dataset created in the previous step. Fill out the fields as shown in the table below.

2. Read more about how to create a custom metric

Check out the Schema Explorer to learn more about all supported views and columns/properties: Mailchimp Schema Explorer