A marketing team reviews their attribution report. Google Ads shows a 3.8x ROAS. The head of marketing is happy. The CFO asks: “Which campaigns? Which keywords? Where exactly is the return coming from?”
Silence.
Channel-level attribution is where most teams stop. But it’s only one dimension of the picture. Campaign, keyword, source/medium, and touchpoint type each reveal something different, and making budget decisions with only one dimension is like navigating with a map that shows countries but not cities.
The Four Dimensions of Attribution
Attribution data can be sliced across multiple dimensions. Each one answers a distinct question about your marketing performance.
Channel: Where should we invest?
Channel attribution groups all marketing activity by platform or category: Google Ads, Meta, Email, Organic Search, Direct, Referral. It answers the broadest question: which channels drive revenue?
This is the right starting point. If you’re spending across five channels, you need to know which ones earn their keep. Channel-level ROAS, CPA, and revenue share give you the macro view.
Best for: Portfolio-level budget allocation, executive reporting, quarterly planning.
Limitation: Channels are averages. A channel with 3x blended ROAS might contain campaigns ranging from 0.4x to 8x. The average hides the variance.
Campaign: How should we spend within each channel?
Campaign attribution breaks each channel into its constituent campaigns. Instead of “Google Ads returns 3.8x,” you see “Brand Search returns 6.2x, Spring Promo returns 4.1x, Display Retarget returns 0.6x.”
This is where budget optimisation actually happens. You can’t optimise a channel, only the campaigns within it. Moving budget from a 0.6x campaign to a 4.1x campaign within the same channel can dramatically improve overall return without changing your channel mix at all.
Best for: Weekly and monthly budget adjustments, campaign performance reviews, identifying waste.
Limitation: Requires consistent UTM tagging. If campaigns aren’t tagged consistently (or at all), this dimension is unreliable. A UTM manager with naming rules is essential.
Keyword: What are people actually searching for?
Keyword attribution maps organic search terms to revenue. It answers: which search queries bring visitors who actually convert?
This dimension is uniquely valuable because it reveals intent. “Marketing attribution software” has a very different conversion profile from “what is ROAS.” Both might rank on page one, but one signals buying intent while the other signals research intent.
Best for: SEO strategy, content prioritisation, understanding the gap between traffic volume and revenue value.
Limitation: Keyword data is only available for organic search (via search console integration). Paid search keywords are captured via UTM parameters at the campaign level.
Source and Medium: Separating paid from organic
Source/medium attribution combines where visitors come from (source: google, facebook, newsletter) with how they got there (medium: cpc, organic, email). This separates “google / cpc” from “google / organic,” which channel-level attribution sometimes conflates.
Best for: Understanding the true cost of acquisition per traffic type, comparing paid vs organic performance on the same platform.
Limitation: Granularity can become overwhelming. With dozens of source/medium combinations, focus on the top 10-15 that drive meaningful volume.
Why You Need Multiple Dimensions
Each dimension alone gives an incomplete picture. Consider this scenario:
Your channel report shows Email at 5.2x ROAS. Looks great. But drill into campaigns and you find:
| Campaign | Attributed Revenue | ROAS |
|---|---|---|
| Welcome Series | $12,000 | 8.4x |
| Weekly Newsletter | $4,500 | 3.1x |
| Re-engagement | $800 | 0.9x |
| Flash Sale Blast | $6,200 | 6.0x |
The re-engagement campaign is barely breaking even. Without campaign-level data, you’d never know, because the channel average masks it.
Now look at organic search keywords driving the most revenue:
| Keyword | Sessions | Revenue | Conversion Rate |
|---|---|---|---|
| [brand name] | 450 | $18,000 | 4.2% |
| marketing attribution tool | 35 | $3,200 | 9.1% |
| how to calculate roas | 280 | $400 | 0.1% |
| multi-touch attribution software | 22 | $2,800 | 12.7% |
“How to calculate roas” drives 8x more traffic than “multi-touch attribution software,” but the latter generates 7x more revenue. If you’re prioritising content by traffic volume alone, you’re optimising for the wrong metric.
The Hierarchy of Decisions
Think of attribution dimensions as a decision hierarchy:
1. Channel answers: “Where should our budget go?” Quarterly and annual decisions. Which channels deserve investment? Which should we scale back?
2. Campaign answers: “How should we allocate within each channel?” Monthly decisions. Which campaigns should get more budget? Which should be paused?
3. Keyword answers: “What content should we create or optimise?” Ongoing decisions. Which search terms should we target? Where should we focus SEO effort?
4. Source/Medium answers: “Are we getting value from each traffic type?” Periodic analysis. Is paid search cannibalising organic? Is social referral traffic converting?
5. Touchpoint Type answers: “Are we measuring everything that matters?” Structural question. Are phone calls and chat conversations being attributed, or are they invisible?
The Touchpoint Type Dimension
One dimension that’s often overlooked is touchpoint type: page views, form submissions, phone calls, and chat conversations. Each represents a different level of engagement.
A customer journey that includes a phone call converts at a fundamentally different rate than one that’s purely digital. If you’re not tracking calls and chats as touchpoints, you’re missing a significant part of the conversion path, particularly for high-value B2B and services businesses.
34% of converted customers in some industries have a phone call somewhere in their journey. Without call tracking integrated into your attribution, those touchpoints get zero credit, and the channels that drive phone calls look less effective than they really are.
Combining Dimensions
The real power comes from combining dimensions. Filter by channel, then drill into campaigns, then look at which keywords feed those campaigns. This layered analysis reveals:
- Which campaigns on which channels drive the most valuable customers (not just the most conversions)
- Which keywords start journeys that end in high-value campaigns (connecting SEO to paid conversion)
- Which touchpoint types appear in the highest-value journeys (understanding the role of calls and chats)
Getting Started
If you’re currently only looking at channel-level data, here’s how to add dimensions progressively:
Start with campaigns
This delivers the fastest ROI. Clean up your UTM tagging, connect your ad platforms, and run attribution at the campaign level. You will almost certainly find campaigns that are wasting budget within channels that look healthy at the aggregate level.
Add keyword attribution
Connect your search console data and map organic keywords to revenue. Identify high-intent keywords that convert well but have low traffic, then prioritise content around those terms.
Layer in touchpoint types
Enable call tracking and chat attribution. See how offline interactions contribute to conversions. You may find that channels driving phone calls (like local SEO or certain paid campaigns) are significantly more valuable than they appeared.
Review source/medium splits
Compare paid vs organic on each platform. Look for cannibalisation (paid taking credit for organic traffic) and identify undervalued organic sources.
The Bottom Line
Channel attribution is necessary but not sufficient. Campaign attribution is where budget optimisation happens. Keyword attribution guides your content strategy. Touchpoint type attribution ensures you’re measuring everything that matters.
No single dimension gives you the full picture. The teams that outperform are the ones that can move fluidly between dimensions, zooming out for the strategic view and zooming in for the tactical decisions.