1.1.3 Adobe Marketing Agent for Microsoft 365 Copilot

[Beta]{class="badge informative"}

Prerequisites

In order to follow the steps in this lab as documented below, the following access is required:

Video

In this video, you’ll get an explanation and demonstration of all the steps involved in this exercise.

1.1.3.1 Add Adobe Marketing Agent to Microsoft 365 Teams & Copilot

Open Microsoft Teams and log in using your account details. Once you’re logged in, you should see this.

Click Apps.

Select Manage your apps.

Select Upload an app.

Select Upload a custom app.

Select the manifest file that was given to you by your instructor and click Open.

Click Add.

Click Open with Copilot.

Adobe Marketing Agent is now loaded successfully.

Enter the prompt login and click the send button.

Click Sign in to Adobe Marketing Agent.

A new window will open up, asking you to login using your Adobe account credentials.

After successful authentication, you may need to select the specific instance you want to use. If you see this screen, please select the instance --aepImsOrgName–.

You will then see a similar code being generated. Click Copy to copy the code.

Paste the code in the Adobe Marketing Agent window in Copilot and click the send button.

You should then see something similar to this. You’re now successfully signed in to Adobe Marketing Agent in Microsoft 365 Copilot.

1.1.3.2 Set context in Adobe Marketing Agent

Before interacting further with Adobe Marketing Agent through Copilot, the context needs to be set.

For this exercise, the context needs to be set to use:

To change the sandbox, enter the following command and click the send button.

change sandbox

You should then see something similar to this. Select the sandbox you need to use and click select.

You should then see this. To change the dataview, enter the following command and click the send button.

change dataview

You should then see something similar to this. Select the dataview you need to use and click select.

You should then see this. The context is now set correctly so you can start sending specific prompts next.

1.1.3.3 Start with overall purchase trends to anchor context and zoom into fiber

Intent

Get a toplevel pulse on category demand—Mobile, Landline, Internet, TV, Fiber—specifically for the most recent 60 days. This sets baselines for seasonality, promo effects, and regional variance after the New York rollout.

Enter the following Prompt and click the send button.

Show me purchases by mainCategory over the last 4 months.

You should then see this:

Enter the following Prompt and click the send button.

Show me purchases by mainCategory = Fiber over the last 4 months broken down by week

You should then see this, which drills down into Fiber-specific trends.

1.1.3.4 Correlate Orders with Content Preferences

Intent

Test the hypothesis that a preference for a specific genre (e.g., SciFi, Sports, Drama) predicts broadband upgrade behavior—especially for high bandwidth needs.

First, you need to find out which field is used to store the genre preference.

Enter the following Prompt and click the send button.

Which field is used to store the preferred genre

You should then see this, which shows that the field used for genre is _experienceplatform.individualCharacteristics.preferences.preferredGenre.

With that information, you can start drilling down in the purchase data.

Enter the following Prompt and click the send button.

Show me ordersYTD by preferredGenre for the last 4 months

You should then see this. Click View Data.

You should then see this.

1.1.3.5 Identify Existing Fiber Journeys

Intent

Discover which active or recently concluded journeys include “Fiber” in the title—e.g., “Fiber Upgrade NYC – Sept”, “Fiber Trial – Streaming Bundle”.

Enter the following Prompt and click the send button.

What journeys exist?

You should then see a list of journeys.

Enter the following Prompt and click the send button.

Which of these journeys has 'Fiber' in its name?

You should then see this.

Enter the following Prompt and click the send button.

Show me the details of the journey 'CitiSignal - Fiber Max Launch Promotion'

You should then see this.

1.1.3.6 Validate journey performance via fallout analysis

Intent

You want to understand journey performance fallout to know if there are any nodes or conditions within the journey that are experiencing a large percentage of profiles being dropped. This is helpful in understanding if additional adjustments are needed in the journey.

Enter the following Prompt and click the send button.

Create a fall-out report on the "CitiSignal - Fiber Max Launch Promotion" journey

You should then see this.

Scroll down a little bit more to see observations and recommendations. Click the 3 dots … and then select Journey Details to open the specific journey in Adobe Journey Optimizer.

You should then see this.

You’ve now completed this lab.

Go Back to Agent Orchestrator

Go Back to All Modules