By Sudharshan M K, Founder, Frontial Technologies
I started working with Salesforce Marketing Cloud back in the ExactTarget days.
That was a very different Marketing Cloud from what we see today.
Over the years, I have worked on campaigns, SQL, Data Extensions, automations, Journey Builder, integrations, personalisation and a lot of the things that happen behind the scenes to make Marketing Cloud work.
And after all these years, I still see the same problem in many organisations.
They have Salesforce Marketing Cloud, but they are not getting everything they could from it.
Not because the platform isn’t capable.
It is.
The problem is usually somewhere between the technology, the data, the processes and the people using it.
I’ve seen organisations with hundreds of Data Extensions but no clear idea which ones are actually important.
I’ve seen teams manually doing work that could have been automated months ago.
I’ve seen marketers waiting for developers to build simple segments.
I’ve seen journeys become so complicated that even the people who built them struggle to explain how they work.
And I’ve seen perfectly good Marketing Cloud implementations slowly become difficult to manage as the business grows.
That is what made me think about this topic.
What does a good Salesforce Marketing Cloud setup actually look like?
Marketing Cloud Is Not Just an Email-Sending Platform
I still come across organisations that primarily use Marketing Cloud to send emails.
There is nothing wrong with that.
But you are leaving a lot on the table.
Marketing Cloud can be the engine behind much more than email campaigns.
It can help organisations manage:
-
Customer segmentation
-
Marketing automation
-
Journey orchestration
-
Personalisation
-
Customer data
-
SMS and mobile engagement
-
WhatsApp communication
-
Transactional messaging
-
Analytics
-
API-driven experiences
The real value starts when these capabilities work together.
Think about a customer who visits your website, looks at a product, adds it to the cart, doesn’t purchase, later makes a purchase and then becomes a repeat customer.
There are opportunities to communicate with that customer at every stage.
The question is not:
“Can Marketing Cloud send an email?”
The better question is:
“Can Marketing Cloud understand what this customer should experience next?”
That is where things get interesting.
The First Thing I Look At Is Usually Data
Whenever I get involved in a Marketing Cloud environment, one of my first questions is:
Where is the data coming from?
It sounds simple, but it tells you a lot.
If the customer data isn’t reliable, everything built on top of it becomes questionable.
You can build a beautiful Journey Builder journey.
You can create a great email.
You can write perfect AMPscript.
But if the customer data is wrong, the experience will still be wrong.
I’ve seen teams spend a lot of time designing journeys and very little time thinking about the data feeding those journeys.
I think that is one of the biggest mistakes.
Before building anything, I want to understand:
-
Where does the customer data originate?
-
Which system is the source of truth?
-
How does data enter Marketing Cloud?
-
How often is it refreshed?
-
How is customer identity managed?
-
Which Data Extensions should be trusted?
-
How are preferences and consent handled?
-
What happens when data fails?
Good Marketing Cloud starts with good data.
Data Extensions Can Quickly Become a Mess
This is probably one of the most common things I see.
A new campaign comes in.
Someone creates a Data Extension.
Another campaign comes in.
Another Data Extension.
Then another.
Nobody thinks much about it because each individual Data Extension makes sense.
The problem appears two years later.
Suddenly you have hundreds of Data Extensions.
Someone asks:
“Which one should I use?”
Nobody is quite sure.
I’ve seen situations where three different Data Extensions contain almost the same customer information, but each is being used by a different automation.
That creates unnecessary complexity.
Data Extensions are not just tables you create whenever a campaign needs one.
They are part of your Marketing Cloud data architecture.
You need naming standards.
You need ownership.
You need retention rules.
You need to understand which data is reusable and which data is campaign-specific.
A little discipline early can save a lot of pain later.
SQL Is Easy to Write. Good SQL Is Harder.
I enjoy working with SQL in Marketing Cloud.
It gives developers a lot of flexibility, especially when building segmentation and data transformations.
But there is a difference between SQL that works and SQL that is maintainable.
I’ve seen the same customer logic written five different ways across five different queries.
For example, a company may have separate queries for:
-
Active customers
-
Engaged customers
-
Recent purchasers
-
High-value customers
-
Lapsed customers
Over time, these definitions start drifting.
One query defines an active customer as someone who purchased in 90 days.
Another uses 120 days.
Another uses email engagement.
All three work.
But which one is correct?
This is where Marketing Cloud architecture matters.
The goal isn’t just to make the query run. The goal is to make the logic reusable, understandable and maintainable.
Your Marketers Shouldn’t Have to Wait for Developers for Everything
This is something I feel quite strongly about.
I’ve seen simple marketing requests turn into long technical processes.
A marketer needs a segment.
They send the requirement to the SFMC developer.
The developer writes SQL.
The marketer reviews it.
There is a change.
The developer updates the SQL.
The query is tested.
Then the campaign can move forward.
For a complex audience, that’s completely reasonable.
But if this happens for every small change, your marketing team becomes dependent on a developer for almost everything.
That is not a people problem.
It is a process problem.
With the right architecture, reusable segments, standardised processes and appropriate self-service capabilities, a lot of this dependency can be reduced.
The developer can then focus on the work where their expertise really matters.
Journey Builder: Just Because You Can Doesn’t Mean You Should
I like Journey Builder.
I’ve spent a lot of time building and troubleshooting journeys.
But I’ve also seen some journeys that made me stop and ask:
“Do we really need all of this?”
A simple customer journey can sometimes become:
Decision Split → Wait → Decision Split → Engagement Split → Wait → SQL → Decision Split → Another Journey
And then six months later, someone asks:
“Why does this customer go down this path?”
Nobody remembers.
The problem isn’t Journey Builder.
The problem is trying to solve every problem inside Journey Builder.
Sometimes the logic belongs in SQL.
Sometimes it belongs in Automation Studio.
Sometimes it belongs in the data layer.
Sometimes the business process itself needs to be simplified.
A good SFMC consultant should not just know how to build a journey.
They should know when not to build one.
Automation Should Actually Automate Something
This sounds obvious.
But I’ve seen plenty of Marketing Cloud environments where people still perform repetitive tasks manually.
Things like:
-
Preparing campaign audiences
-
Updating Data Extensions
-
Running recurring queries
-
Creating suppression lists
-
Preparing reporting datasets
-
Moving data between tables
-
Performing the same validation every week
If someone is doing the same thing every Monday morning, I start asking:
“Why are we still doing this manually?”
Automation Studio exists for a reason.
But automation isn’t simply putting a few SQL activities into a workflow.
You need to think about dependencies, scheduling, failure handling, monitoring and ownership.
A good automation should make someone’s job easier.
Not create another thing for someone to monitor manually.
Personalisation Is More Than First Name
One of the simplest examples is:
Hi %%FirstName%%
Yes, that’s personalisation.
But it isn’t the kind of personalisation that excites me.
The interesting questions are:
What did this customer buy?
What are they interested in?
When did they last engage?
Where are they in the customer lifecycle?
What should they see next?
What shouldn’t they receive?
That requires data, segmentation, content and journey logic working together.
Good personalisation is about relevance, not just inserting a name into an email.
Marketing Cloud Doesn’t Work in Isolation
Another thing I have learned over the years is that Marketing Cloud is rarely successful when treated as a standalone platform.
It usually needs to work with other systems.
Salesforce CRM.
E-commerce platforms.
Websites.
Mobile applications.
Data platforms.
Customer service systems.
Analytics platforms.
External APIs.
The integration between these systems can make or break the Marketing Cloud experience.
If customer data arrives late, your communication may be irrelevant.
If identity isn’t handled correctly, customers may receive duplicate messages.
If preferences aren’t synchronised, you can create compliance issues.
If important behavioural events don’t reach Marketing Cloud, your journeys can’t react to them.
So when someone tells me:
“We need a Marketing Cloud implementation.”
My next question is usually:
“What does the ecosystem around Marketing Cloud look like?”
If I Had to Audit an Existing SFMC Environment
If I joined an organisation that had been running Marketing Cloud for several years, I wouldn’t immediately start building new journeys.
I’d first try to understand what is already there.
I’d look at:
Data Architecture
Is the data model clean and scalable?
Data Extensions
Which ones are actually being used?
SQL
Is the logic reusable and maintainable?
Automations
Are there duplicate or unnecessary processes?
Journeys
Are they still relevant to the business?
Integrations
Is data flowing reliably?
Governance
Does someone actually own the environment?
Deliverability
Are sending practices and engagement being managed properly?
Documentation
Could another developer take over tomorrow?
Business Outcomes
Are we measuring what matters to the business?
Only after understanding these areas would I start recommending changes.
Sometimes the answer is a rebuild.
But quite often, it isn’t.
Sometimes you just need to clean up what you already have.
What I Think the Future of Marketing Cloud Looks Like
I think the next few years are going to be particularly interesting for Marketing Cloud.
We will see more AI.
More natural-language interfaces.
More real-time data.
More sophisticated personalisation.
More connected customer journeys.
And, hopefully, less repetitive manual work.
But I don’t think AI will make Marketing Cloud expertise irrelevant.
Quite the opposite.
When technology becomes easier to use, understanding what should be built becomes even more important.
AI can generate a query.
But someone needs to know whether the query is correct.
AI can suggest a segment.
Someone needs to understand whether the segment makes business sense.
AI can help create a journey.
Someone still needs to understand the customer experience.
Technology can accelerate execution. Experience provides judgement.
Where We Help at Frontial
At Frontial, we work with organisations across the Salesforce Marketing Cloud lifecycle.
Sometimes the requirement is a completely new implementation.
Sometimes it is a migration.
Sometimes the Marketing Cloud environment already exists and needs to be cleaned up.
Sometimes the team needs additional technical capacity.
And sometimes the organisation simply wants an experienced team to take care of Marketing Cloud operations.
Our work includes:
-
Salesforce Marketing Cloud implementation
-
SFMC migration
-
Marketing Cloud architecture
-
Data Extensions and data modelling
-
SQL and segmentation
-
Automation Studio
-
Journey Builder
-
Email development
-
AMPscript
-
CloudPages
-
API integrations
-
Marketing Cloud Intelligence
-
Data Cloud
-
Personalisation
-
Campaign operations
-
SFMC optimisation
-
Managed services
We don’t believe every organisation needs the same solution.
The right Marketing Cloud setup depends on the business, the data, the team and the customer journey.
If Your Marketing Cloud Environment Feels Complicated, Let’s Talk
If you’ve been using Salesforce Marketing Cloud for a few years and things have started getting complicated, you’re probably not alone.
Maybe you have too many Data Extensions.
Maybe your SQL has become difficult to maintain.
Maybe your journeys are more complicated than they need to be.
Maybe your marketing team depends heavily on developers.
Maybe campaigns take longer to launch than they should.
Maybe you have a lot of Marketing Cloud capability that you’re simply not using.
These are all solvable problems.
And you don’t necessarily need to start again.
Sometimes you need the right people to look at the environment with fresh eyes.
After spending many years working with Marketing Cloud, I still enjoy these problems.
Not because of the technology itself.
But because when Marketing Cloud is designed properly, it can genuinely change how a marketing team works.
Less manual work. Better data. Faster campaigns. More relevant customer experiences.
That, to me, is what a good Salesforce Marketing Cloud implementation should deliver.
If you’re planning a Salesforce Marketing Cloud implementation, looking to optimise an existing environment, or need experienced SFMC resources to support your team, I’d be happy to have a conversation.
