Quickly grasp these 4 essential data concepts for GTM teams
Also - Multi-touch Attribution versus MMM versus Incrementality
In this edition:
Modern Data Concepts Explained Simply
Multi-Touch Attribution Versus MMM Versus Incrementality
Exclusive For Paid Subscribers: What Being a Manager of Managers Taught Me About Leadership
Imagine gaining 5 hours back each week.
UTM.io makes tracking UTMs simple and efficient, so you can spend less time on admin tasks and more time optimizing campaigns.
Try UTM.io and get your time back.
Interested in reaching over 5,500 marketing ops and revops professionals? You can also get in front of 50,000 marketers on Linkedin and garner 100,000’s of impressions for your brand. Fill out this form to learn more about The Marketing Operations Leader sponsorships.
Modern Data Concepts Explained Simply
Data is core to everything we do now.
But the concepts can be pretty confusing.
Here’s a simplified breakdown of key modern data concepts GTM teams use today.
🔷 Zero-Copy Data/Data Sharing
Tools access the data warehouse directly—no duplicating, no syncing nightmares.
Breakdown: Standard tech stacks typically involve multiple copies of your database for CRM, MAP, reporting, and other purposes. If tools can access the central data warehouse directly and use that as their database, then there are “zero copies.” Imagine the difference between exchanging multiple word docs versus collaborating on Google Docs.
🔷 Warehouse-Native Processing
Logic and data transformations happen inside the warehouse.
Breakdown: Let’s say you want to target new users and former users of your product. You’d have to wait for data to come from your warehouse and other sources into your email tool, and then you can build a list (and it might already be out of date). A warehouse-native approach would be to build the list directly in the data warehouse and push to your email tool in real-time (faster, cleaner, more accurate).
Examples: Reverse ETL, dbt, AI agents, or analytics tools working directly in BigQuery, Redshift, Databricks, etc.
🔷 Reverse ETL
Data is pushed from the warehouse to your business tools, such as Salesforce, Hubspot, Tableau, etc.
Breakdown: When you take all the data from your business tools and store it in a data warehouse, that’s called ETL (extract, transform, load). Doing the reverse of that is actually very different and sometimes challenging, we call it Reverse ETL. Imagine - it’s pretty straightforward to take all the files from 20 different offices and put them into one cabinet in a head office. But it’s a completely different job to take all files from one head office and distribute to 20 different offices and make sure everything is in the right cabinet.
Example: Hightouch, Census, or GrowthLoop syncing updated segments into your CRM or MAP.
🔷 Composable Architecture
Modular systems where tools are loosely coupled and easily swapped.
Breakdown: If you use one platform for various functionality (email, forms, lead scoring, personalization) that is called monolithic. But what if you like the email but not the forms or the scoring? Picking separate tools/functionality and having them work effectively together is called composable.
Multi-Touch Attribution Versus MMM Versus Incrementality
I recently participated in a debate on which measurement model is better: multi-touch attribution, incrementality, or market mix modeling.
Here’s what I took away:
1.
Attribution isn’t dead. Misuse is.
The problem isn’t that attribution doesn’t work—it’s that most teams are using it for the wrong reasons or expecting it to do more than it’s capable of.
You need to match the tool to the question you’re asking:
Multi-Touch Attribution (MTA) = What influenced the journey?
Marketing Mix Modeling (MMM) = Where should I allocate my budget?
Incrementality Testing = What actually caused a result?
Each has its place. Don’t use a wrench to do a hammer’s job.
2.
Attribution without alignment is pointless.
You can have the most sophisticated attribution model in the world…
But if your sales team doesn’t trust it, your CFO doesn’t understand it, and your CMO can’t report on it—you’ve failed.
Analytics should tell a story. Not start a debate.
Operational clarity comes from aligning on a few shared truths. For example:
Where do our opportunities actually come from?
What activities are repeatable and scalable?
What should we stop doing?
If you can’t answer those questions together, you don’t need another dashboard—you need alignment.
3.
Don’t mistake complexity for accuracy.
This one’s spicy:
Single-touch attribution might be the best model for you.
Not because it’s “right,” but because your team can understand it, rally behind it, and act on it.
If your MTA report is 12 layers deep and still no one knows what to do next… you’ve built an elegant trap, not a strategy.
Sometimes, clarity beats precision.
4.
MMM is powerful, but not for everyone.
Marketing Mix Modeling can help you understand macro impact—great for orgs spending millions across lots of channels.
But don’t expect magic if:
You’re not spending at scale
You don’t have 2+ years of historical data
You can’t control for confounding variables
MMM is a scalpel, not a sledgehammer. Use it wisely.
5.
Incrementality is the holy grail—but it’s hard.
Incrementality testing gives you causality. It’s the closest thing to a scientific method in marketing.
But in enterprise orgs with 6–12 month sales cycles, you can’t afford to wait a year for closed-won data. You need leading indicators:
Meetings booked
Product signups
Engagement lift
Creativity matters here. You don’t always need a clean holdout group—sometimes pulsing spend or increasing frequency is enough to see a directional lift.
6.
Your job isn’t to build a report. It’s to tell a story.
Reports don’t drive results. Clarity does.
The data doesn’t have to be perfect—but it does have to be believable.
One of my favorite lines from the session:
“Analytics shouldn’t start a debate. They should tell a story.”
In other words: If your CFO sees the same number you do and still doesn’t trust it, you’ve lost before the conversation even starts.
7.
Final word: Measurement is a means to an end.
You’re not running a measurement model to impress your peers. You’re doing it to answer core business questions:
Is marketing driving revenue?
Where should we invest next?
What’s working—and what’s not?
If your attribution model doesn’t help you answer those, it doesn’t matter how technically sound it is.
Build systems your team can use. Tell stories your org can trust. Use the model that matches the moment.
Keep reading with a 7-day free trial
Subscribe to The Marketing Operations Leader to keep reading this post and get 7 days of free access to the full post archives.