Just like anything else in Salesforce, every project follows a step-by-step process.
If you follow the process, you'll achieve success every single time.
If you don't follow the process, your projects will fail.
The only problem is... You don't always know what the step-by-step process is.
Especially when you're working with a new product like Salesforce CDP.
In this blog post, I'll show you the 7-step process to get Salesforce CDP up and running as smoothly as possible.
Let's get it started!
Step 1: Connect CDP To External Sources
You need to connect Salesforce CDP to every system it's going to pull data from.
The systems (or data sources) you can connect to CDP are:
Marketing Cloud
Commerce Cloud
Amazon S3
Google Cloud Storage
Interaction Studio
Websites
Mobile Apps
External Activation Platforms
You don't have to connect Salesforce CDP to every single data source above, but if you do plan on using a data source, then you'll want to connect it to Salesforce CDP.
When you connect a data source to CDP, a data stream is created.
A data stream is kind of like an invisible telephone wire between Salesforce CDP and the data source. It's the connection that allows data to flow between your data source and CDP.
For example:
If you connect Salesforce CDP to Marketing Cloud, then a data stream connection is created between Salesforce CDP and Marketing Cloud.
Data can flow from Marketing Cloud into Salesforce CDP through that data stream connection.
So the first step is to decide which data sources will be connected to Salesforce CDP and then set up the data streams to connect the systems.
Step 2: Ingest The Data From Those Sources
The 2nd step is to start ingesting data from the sources you're connected to.
Ingesting data is kinda like when you ingest food into your mouth.
You're taking food and putting it into your mouth with a spoon or fork.
If you can visualize that, that's what CDP is doing.
CDP is taking data from the sources it's connected to, and putting that data into its system just like we do with food.
All the data comes in raw and unfiltered through data streams.
Remember from step one, Data streams are the connection between Salesforce CDP and data sources.
In this 2nd step (the ingestion step), you need to specifically tell CDP which areas to pull data from.
For example:
Say you connected CDP to Marketing Cloud in step one when you were connecting your data sources.
In this second step, you need to tell CDP which data extensions inside Marketing Cloud to grab data from.
This is the step where you get very specific about the EXACT files you're going to pull.
Another example:
If you connected Salesforce CDP to Amazon S3 in step one.
In this 2nd step of data ingestion, you now tell CDP which file to grab inside of the Amazon S3 account you're connected to.
After you ingest data from your data sources, your data will flow into Data Lake Objects inside CDP.
In the next step, you'll take the raw data that's flowing into Data Lake Objects and map them to Data Model Objects.
Data Model Objects are the objects we want our data to flow into.
Data Model Objects are where all the magic happens inside of Salesforce CDP.
Data Model Objects are the prized possession that everyone talks about when they think of Salesforce CDP.
They're what people talk about when they refer to the Customer 360 Data Model.
Salesforce CDP = Customer 360 Data Model
Now that we know that, let's talk about mapping your ingested data to Data Model Objects!
Step 3: Map Your Data To Data Model Objects Inside CDP
Data Source Objects are mapped to Data Model Objects
Data Model Objects live inside a huge Data Lake.
The next step to implementing Salesforce CDP is to take the raw ingested data coming from your data sources to the correct Data Model Objects.
There are 3 Data Model Object types… they are:
Profile
Engagement
Other
"Profile" Data Model Objects represent the characteristics of a person.
Examples are:
A Person's First Name, Last Name, Height, Weight, etc... Profile data represents the characteristics of a person.
"Engagement" Data Model Objects are the behaviors that customers took.
Examples are:
When a customer clicked a link in an email
A customer opening an email
The date a customer purchased a product in store
"Other" Data Model Objects are things that are NOT Profile or Engagement Data.
Examples are:
Store detail information like - Store Number, Location of the store, Address of the store, etc...
The details of a car - The VIN number, car color, wheel size, etc...
Step 4: Unify Your Data
aka Identity Resolution - This is where you map all your data together.
This is where the concept of a “Unified Individual” comes in.
A unified individual is a person that has all data sources compiled together into one single profile.
This is what Salesforce CDP is all about.
You connected all your data sources in step one and now it’s time to pull all those sources together into one unified profile.
A unified profile is a customer profile that has every single touch point that they’ve made with the company linked together.
For example, a customer named Nakia has been on the Nike app, the Nike website, and is signed up for Nike emails, and is in the Salesforce CRM that Nike uses.
Every single touch point that was mentioned above is now connected so Nike knows every single area that Nakia has been to and can connect the dots so Nike doesn’t send overlapping messages to Nakia when Marketing to her in the future.
This is how companies with Salesforce CDP are going to change the Marketing game.
There won’t be any disconnected data sources anymore.
This means more personalized messages in the overall experience for the customer like Nakia.
Step 5: Create Insights On Your Data
There are two types of insights in Salesforce CDP...
They are:
Calculated Insights
Streaming Insights
Let's talk about what they both do.
Calculated Insights
Calculated Insights gather large bulk amounts of data and they answer the following questions:
How much do customers spend with you?
How often do customers spend with you?
How recent have your customers have with you?
How engaged your customers are with your brand (like if they opened or clicked emails)
Your churn rates
Customer Satisfaction Scores (CSAT)
Streaming Insights
Streaming Insights are:
Smaller amounts of data that come in more frequently (in near real-time).
They don't come in large bulk amounts like Calculated Insights do.
Useful to create data actions.
Small streams of data that allow you to use APIs to automatically pull data into CDP
Data Actions are common with Streaming Insights because you're getting your data pretty much in real-time.
This basically means that when something happens (like a customer enters a geo-fence of a Nike store), you can take immediate action.
So for example:
A customer is in the mall and comes within 100 feet of a Nike store, you can trigger a notification to go to their phone and let them know that they can get 10% off their next pair of shoes.
This entices that person to go to the Nike store and check out the shoes.
Who knows what that person will end up buying when they're in the store, but at least the streaming insight was able to get them into the store when they otherwise might not have even thought about going there in the first place.
And this is just the tip of the iceberg on the data actions you can take with Streaming Insights!
Are you starting to see how you can harness the power of CDP?
This is going to really change business as we know it forever!
Now that we know about Insights, let's talk about Segmentation.
Step 6: Segment Your Data
Segmentation is the process of grouping people into certain buckets so you can send VERY specific messages to them.
The reason you want to segment your customers is because...
A personalized message that relates to someone is more powerful than random generalized messages.
For example:
Let's say you are Nike and you want to highlight your shoes to your customers in the United States.
What if you had snow boots and you sent an email to ALL of your subscribers in the United States and said "Hey, check out our newest shoes"... And it was a picture of some snow boots...
How many people would be interested in that?
Not many...
People in a warm state like Florida would be wondering why in the world Nike would send them something so unrelated to their everyday lives.
But change that to a message that's more personal and sent to a segmented audience...
Now in this new example:
Nike segments their audience and finds all the people that live in Miami Florida and that like AirForce Ones and sends those people a message that says:
"Hey (FirstName), check out our new all-white Air Force Ones. These would go well with a pair of black Nike shorts!
How much more personalized is that message compared to the first one?
WAY MORE PERSONALIZED!
And people in Miami Florida would be much more likely to buy those shoes compared to a random message about snow boots that have nothing to do with them.
So now that we know why segmentation is important, let's talk about things you can segment in Salesforce CDP.
You can create segments with:
Calculated Insights that you've already created in the last step
Customer Data from Data Extensions in Marketing Cloud
Customer Data from CRM data in Sales Cloud or Service Cloud
Customer Data from your website
Customer Data from your Mobile App
And so many other places
Once you create a segment in Salesforce CDP, you'll have a certain number of people that match the criteria that you set.
For example:
Let's say you segment the following people:
All the men and women that live in Miami Florida, are between the ages of 21 and 31 years old and make between $50,000 to $75,000 per year.
There is going to be a certain number of people that fall within that category based on the customer data your company has.
When you find that number of people, you want to save that segment and then publish it.
Once you publish your segment in Salesforce CDP, it's time to ACTIVATE it!
So let's go to the final step!
Step 7: Activate Your Data
The last step in the CDP implementation process is to Activate the published segments that you created in step 6.
Activation is the process of taking your customer segments and sending them back to the places you originally ingested your data from.
For example:
You ingested your data from Marketing Cloud, Sales Cloud, Commerce Cloud, your Mobile App, your Website, Amazon S3 buckets, etc...
And now that you consolidated all that data together in identity resolution and segmented that data, it's time to send your published segment back out to Marketing Cloud (or whichever platform you want to send that segment to).
So in Marketing Cloud, you can create journeys in Journey Builder.
A journey is basically a pre-determined path of emails, text messages, and app notifications that you can set up one time so that when someone enters into a journey, all of those messages that you already designed into the journey are automatically sent to that person based on the schedule you setup.
You can take the people that are in your published segments and put them into a pre-defined journey in Journey Builder and then the journey will take care of all the messaging you want to send to them.
For Example:
Nike can create an "Abandoned Cart Journey' in Marketing Cloud Journey Builder for anyone that was about to checkout on their website but for some reason didn't actually complete the purchase and pay for the items in their cart.
If that happens, Nike is already ingesting data into CDP and can create a segment of the people that were about to check out but didn't and then publish that segment and put that segment into their Abaonded Cart Journey to send them a series of messages that will entice them to pay for the items in their cart.
So it would look something like this:
Kaelan goes to the Nike website and looks at a pair of all-white Air Force Ones
Someone knocks on Kaelan's door and Kaelan goes to see who it is and get's distracted and goes out with some friends for the night (forgetting all about those new shoes he was about to buy)
Salesforce CDP ingests data from the Nike website and puts Kaelan's information into a data model object
Identity resolution rules are being run behind the scenes and Kaelan is mapped to the email address that was ingested from Marketing Cloud.
A segment is running that finds people who were about to check out in the last 24 hours but didn't purchase.
The people in the segment are pushed back to Marketing Cloud and put into a journey in Journey Builder
A series of 7 messages get's sent to those people over a 2-week time frame to entice them to come back to the website and finish their purchase.
If they purchase, then they get kicked out of the journey so they don't have to receive any more messages.
And CDP has done its job!
Conclusion
Thats' Salesforce CDP in a nutshell for you!
Now that you know the 7 step process to implementing Salesforce CDP, you're ready to tackle any project you encounter.
All you have to do is follow this 7-step process and you'll be ready for whatever is thrown your way.
Of course, you'll need to learn more about the intricate details inside of each step, but that's the fun part of it all.
The 7-Step Salesforce CDP Process
(Learn More)
Additional Resources
Here are additional resources for you to learn more about Salesforce Marketing Cloud
