Data Cloud Part 6: Practical (Ingest Physical Store's Customer details and Sales details)

Hope you have already created your Data Cloud Org and also downloaded the Customer's details and Sales details from previous posts. 

Now it's time to ingest that data in our Data Cloud Org. So, first login in your Data Cloud Org and go to App Launcher and search for "Data Cloud" App.

It will open the "Data Cloud" App and you will get multiple data cloud related Tabs like belwo screen but in this post we will concentrate on "Data Streams" and "Data Lake Objects" Tabs.

Now our aim is to ingest data from external sources to this data cloud Org and here as we need to load data from direct excel files, so we don't need to connect with any ORG and we will directly ingest data from excel to the Data Cloud by Data Stream's File Upload options. What is this? - let's do some paractical to get more idea on this.

Click on the "Data Stream" Tab and it will open the Data Stream List View. 

Now, click on the "New" button to create the new Data Stream.

It will open the below screen and select the "File Upload" icon and then click on "Next" button.

Now it will open the file selection screen as below,

Click on the "Upload Files" button and then select your "OfflineSalesDetails" file that you have downloaded for customers' offline sales data.

System will load the file and open the below screen. In the below screen you could see that System set default Dala Lake Object Label and API. So, it is always better to update the name based on your need. Update it as mentioned in the below screen and on left side you could find the Category selection option and as we are going to create individual profile, so select here "Profile" option.

Next we need to select the Primary Key (PK) - but if you noticed our data in OfflineSalesDetails sheet then you will found there is no unique value to set as PK. We have "Cstomer Id" but we can't able to use it as PK as multiple records with same "Cstomer Id". So, how we can get the PK?

Here we have to Create a Formula Field to make PK.

Save it and then select this custom formula field as PK.

Now click on the "Supported Fields" Tab and verify the data type.

In this POC every thing looks good, so click on the "Next" button. Select the Data Space. In this POC, we don't have any extra Data Space ccreated, so we will set the OOB "default" data space.

Click on "Deploy".

After deploy, it will redirect to below screen. 

It will take some time and after some time it will show status as "Success" and details will be display in "Refresh History". We can also noticed total records ingested = 5.

Same way create another data stream to ingest "OfflineCustomerDetails" data. Here we will use "Customer Id" as a PK as there is no dupicate id on records and also update Email fied type as "email".

After deploy it will take some time and after some time it will show status as "Success" and it will load 3 records.

Next Step: Now our next step will be checking the data of Data Stream, Data Lake Objects. 

Data Cloud Part 5: Practical (Understand and analyse the data)

 Now, let's start the practical implementation and set up the Data 360 configuration.

Let’s assume that ABC Company has provided us the below Excel file containing customer purchase data from its physical store.

Click Here to Download Offline Sales Details

As a Data 360 Consultant/Architect, the first responsibility is not to immediately start configuring Data 360. Instead, we first need to understand and analyse the data.

So, let's take a look at the Excel file and identify:

• What data is available?
• What does each column represent?
• Who is the customer?
• What information can uniquely identify a customer?
• Are there any duplicate or missing records?
• How can this data be mapped to the Data 360 data model?

Remember: Before configuring Data 360, understanding the source data is the first and most important step.

Our Analisys:

- In the above offline Sales Details sheet, we could found that for Customer Id 1001 and 1002, we have multiple or we can say dulicate record for same customer ids.

- We have also noticed that in the above sheet there is no customer data like customer's name or email or phone number etc. which are very much needed as identifier to maintain unique customer.

- There could be many more as such analysis, but for simplicity let's start with these only.

Our Next Steps:

- As we need customer's name, email etc. so request ABC Company to share these detials from their Offline Physical Store.

- Now there could be 2 possibilities,

    - ABC Company will share you again a considated sheet with customers' details.

    - Or else ABC Company will share you a seperate sheet with only customers' details.

Let's say  ABC Company has shared the below seperate sheet with only customers' details,

Click Here to Download Offline Customer Details

So, our Offline data is ready and we have 2 set of data,

1. Offline Sales Details

2. Offline Customer Details

Note: For this POC, I am only using very very minimal data to make it very easy to understand but in real scenario these data volume will be very large.

Next Step: Now our next step will be ingesting these data in our Data Clould Org.

Data Cloud Part 4: Understand your POC requirement.

Before learning any topic, I believe it is always better to first understand these three simple questions in very very smple ways:

What? → Why? → How?

• What is it? — Understand the concept and what it actually means.
• In previous post, we already undestand What is Data Cloud?
• Why do we need it? — Understand the problem it solves and its real-world value.
• In previous post, we already undestand Why do we need it?
• How does it work? — Understand how to implement or use it in practice.
• Now we will learn very simple way How to implement it?

After this POC we will get a very good understanding of Data Clould and then we can explorer more by ourself by mulitiple Data Cloud realted lerning documents/trainings.

POC: Check the below diagram and this will be our 1st Data Clould Project in very simple way.

🎯 Project: Identify ABC Company's High-End Customers

🏢 Business Scenario

Let's say ABC Company sells products through two different platforms:

1️⃣ Physical Store
Customers purchase products from ABC's physical stores.
Their sales/order information is maintained in Excel.

2️⃣ Online Website
Customers purchase products from ABC's website.
Their sales/order information is stored in Salesforce.

The problem is: ABC doesn't have one complete view of a customer's total purchases.

For example:

Customer	Physical Store	Online Website	Total Purchase
Rahul	₹10,000	₹15,000	₹25,000
Amit	₹2,000	₹15,000	₹17,000
Priya	₹1,000	₹5,000	₹6,000

Rahul may look like a ₹15,000/- customer in Salesforce, but when his physical-store purchases are added, ABC discovers that he has actually purchased ₹25,000 worth of products and he is a High End Custoper for ABC Company. 

🏢 Solution

Now we will use Salesforce DATA Cloud Platform as sohown in below diagram and then perform the below action items like we learn in previould post.

1. External Data Sources (Salesforce, Excel).

2. Establish connections from Data Cloud to external Salesforce.

3. Ingest data by creating Data Streams both from external Salesforce (using connecctor) and from Excel (using file upload option).

4. Data Lake Objects (Auto + Manual based on our needs).

5. Use of Data Explorer / Query Editor to check/validate the data.

6. Data Transforms (Join the mulitple DLOs created from 2 different external sources in a single Talbe or single DLO).

7. Organize data into a meaningful structure by mapping with proper DMOs.

8. Identity Resolution - removed dulicate customer or in other teram identify the unique customer.

9. Calculated Insights - group the customer based on the total purchase amount.

10. Segmenting customers based on the total purchase amout.

11. Activation - Send data/audiences.