Why Everyone’s Talking About MCP Connections, and Why You Should Care
Written by Tara Patoile, Adobe data & insights practitioner
The Bite-Sized Breakdown
MCP Is the “Magic Glue” MarTech and RevOps Weren’t Asking For — But Absolutely Needed
Hot take (because when have I ever not got one?):
I’m willing to bet most of you reading this have some complicated way of connecting your data and trying to keep up with the management of it is a burden. And, you probably haven’t used MCP yet to try to solve it.
Not the sexy, buzzword-friendly “we need AI to save us!” problem. The you can’t stitch revenue and journey data without breaking a sweat problem. The why-is-the-pipeline-question-still-a-slog problem.
Enter MCP connections between data sources — quietly evolving into the unsung hero of modern analytics stacks. If it sounds like just another acronym du jour, stay with me. This one actually matters. And yes, I’m going to explain it like your fellow MarTech pro who knows you’re tired of jargon.
So What the Heck Is MCP?
In plain human terms:
MCP — Model Context Protocol — is the universal adapter for your data sources.
It lets disparate systems talk to each other without needing custom duct tape, MacGyver-style API hacks, or another heroic weekend of data engineering.
Instead of hard-coding a new connector every time:
CRM data
Web behavior
Product usage
Media impressions
Pipeline signals
…all magically become discoverable, understandable, and queryable by analytics and BI tools in a governed, consistent way.
Imagine if your reporting tools could peek into every data silo, but with manners and respect for governance. That’s MCP.
It’s like:
“Hey systems… show me what you’ve got and how to use it securely, please and thanks.”
No more guessing where the CRM lives. No more “which schema did we use last sprint?” guilt spirals.
Why You Should Get Fired Up About MCP
Because for years, marketing and RevOps teams have been wrestling not with data, but with connections.
You’ve lived this:
“We’ve got Adobe Analytics … but revenue lives in Salesforce. And CDP profiles live in another universe. And oh yeah, the BI team is queuing another ETL.”
Sound familiar? That’s the connection tax. Until recently, no one was talking about how to get rid of it, just how to paper over it.
MCP is one of the first standards that actually tries to solve the root cause:
Multiple tools speaking different data languages
Dashboards that disagree just for fun
Analysts begging for clean joins like they’re in a fantasy novel
It doesn’t replace pipelines or data engineering, but it lets tools share context about your data so your business queries aren't always “Is it in Adobe? Or Salesforce? Or Kafka? Or somewhere Larry from BI forgot to document?”
That’s a win.
Why This Matters for Adobe Analytics, CJA, and BI Users
If you live in Adobe Analytics, Adobe Customer Journey Analytics, or a BI tool like Power BI, Looker, or Tableau — MCP changes the game board:
You Get Fewer Custom Connectors
No more bespoke scripts that break every time someone renames a field or changes a schema. MCP holds context so tools know what they’re querying.
You Get Consistent Definitions
Revenue in CJA finally means the same thing as revenue in the BI dashboard… not “close enough” or “depends who asked.”
You Can Ask Better Questions
Not:
“Where is the data?”
But:
“What actually happened between touchpoint A and pipeline creation last quarter?”
That shift from location drama to insight story is powerful.
The Real Impact (and the Part No One Wants to Admit)
Let’s cut the fluff: MCP doesn’t make your legacy dashboards obsolete.
Your analysts still matter. Your governance still matters. Your data cleansing still matters.
What MCP does do, and why I think it’s getting traction, is this:
It turns your stack from a bunch of siloed chatterboxes into a coordinated orchestra.
Not perfect harmony yet, but no more kazoo solos in the violin section either.
What You Should Be Thinking About Right Now
If your org is investing in:
Cross-channel analytics
Unified customer journeys
Revenue attribution models
AI-driven insights
BI democratization
Then MCP shouldn’t be a “nice to have”… it should be part of your data strategy conversation.
Why?
Because every modern reporting layer, every analytics engine, and every emerging AI insight mechanism fundamentally depends on context, not just raw logs.
MCP gives tools that context.
Final Mic Drop
You have a data connection problem — the sort that makes analytics teams grind their teeth and sends MarTech folks to the craft beer aisle early.
MCP doesn’t fix everything, but it fixes the thing that’s been slowing you down the longest.
And that?
That’s worth paying attention to.
Reference Links for Light Reading
Want to get started setting up an MCP connection between your Adobe agents and ChatGPT? Follow this step-by-step video to get started: Adobe Marketing Agent with ChatGPT