Google Cloud Data Integration

🚧

Enabled upon request

At this time, the Google Cloud Data Integration is available upon request. If you don't see the options discussed here in the Conductrics Admin, please contact Conductrics so we can get it enabled for your account.

What You'll Need

First, make sure you have:

• A Conductrics account
• A Google Cloud account
• Some kind of custom visitor data that you'd like to use with Conductrics, stored somewhere in the Google Cloud. For instance, you might store some kind of interesting visitor data in BigTable, Cloud SQL, Cloud Storage, or BigQuery.

Here's a walkthrough of the setup process:

In order to use the integration, you'll need the following, which will be explained here:

• A BigQuery view to power the setup screens in Conductrics, for you and other users in your company. The BigQuery view should contain the fields that you want to share with Conductrics.
• A "service account" that gives Conductrics permission to query the BigQuery view.
• A visitor identifier that you can share with Conductrics, via a "cookie" or passed to our API.

To use the integration in production, you will also need:

• A Cloud Function that Conductrics can use to fetch visitor data for actual visitors in "real time", as they visit your pages or apps that use Conductrics. This is explained below (see "Providing a Cloud Function for Real-Time Visitor Data Access").

Providing a Cloud Function for Real-Time Visitor Data Access

As discussed earlier, the "Hit BigQuery Directly" option isn't appropriate for production use or workloads. When it comes time to go live in production, you'll need to create a Cloud Function that retrieves the actual visitor data from a database of some kind.

The database lookup needs to be reasonably fast - just a few milliseconds preferably, but use whatever performance guidelines you feel are appropriate. Fast-enough performance is generally easy to accomplish using Bigtable, Cloud SQL, Cloud Storage, or similar.

Your Cloud Function should:

• Look for a query parameter called vid
• Based on the vid value, do whatever is needed to retrieve the data for the appropriate visitor data record on your side.
• Return the data as the inputs portion of a simple JSON object.

For instance, here's a sketch of what you might do in Node.js / Express code:

app.get('/visitor-lookup', function (req, res) {

	// the "visitor lookup id" is passed as a query param called "vid"
	var visitor_id = req.query.vid

	// Now do whatever is appropriate to get the appropriate data!
	// For instance, you might execute a SQL query something like:
	//   SELECT * FROM My_Visitors WHERE visitor_id = :visitor_id
	// Here's a "mock record" that illustrates the type of data you might return:
	var visitor_record = {age: 11, rev_total: 1.1, status: 'vip', Cust_type: 'A'}

	// Now return a JSON object, with the visitor data as the "inputs" portion:
	res.json({
		inputs: visitor_record
	});
});

The result when called might look something like this:

{
  "inputs": {
    "age": 11,
    "rev_total": 1.1,
    "status": "vip",
    "Cust_type": "A"
  }
}

📘

In addition to the vid parameter, your cloud function will also be provided with a dt parameter which is the unique identifier for your Deploy Target. In some cases this may be helpful to determine which data you want to return.

🔐

Restricting Access via Basic Authorization

There's an option for requiring HTTP Basic Authorization when specifying the URL for your Cloud Function. You can use that to make sure that the function can't be accessed by just anyone who stumbles upon the URL. It's up to you to look at the Authorization header to get the username and password being presented, and to respond with a 403 Not Authorized if they aren't correct. Feel free to contact Conductrics for help with this!