The budget review was brutal.
Demand gen showed pipeline generated by campaigns. Sales showed quota attainment. Customer success showed retention and expansion revenue.
The CFO asked me, "What pipeline has PMM contributed?"
I'd prepared for this. I pulled up my dashboard: "PMM has influenced 87% of all closed deals in the past quarter."
The CFO stared at me. "87% of all deals?"
"Yes. PMM content was present in 87% of closed opportunities—battle cards used, case studies viewed, product demos attended. Our assets are in nearly every deal."
He asked the question I should've anticipated: "So if PMM influenced 87% of deals, and demand gen says they sourced 65% of pipeline, and sales says they drove 90% of deals... we've got 242% of revenue attributed to three teams. That math doesn't work."
He was right. I'd counted every instance of PMM content appearing in a deal as "influence," regardless of whether it actually mattered.
Presence isn't influence. Just because a prospect downloaded a whitepaper doesn't mean it influenced their buying decision. Just because sales opened a battle card doesn't mean it changed the outcome.
I'd inflated PMM's impact to make our function look valuable. And I'd lost credibility with finance in the process.
I went back to RevOps and rebuilt pipeline attribution from scratch—this time, honestly.
The Influence Inflation Problem
My first attempt at pipeline attribution had used multi-touch attribution logic:
- Prospect downloaded PMM content → +10 points
- Prospect attended product webinar → +15 points
- Sales opened battle card in opportunity → +20 points
- Demo used PMM demo script → +25 points
- Proposal referenced PMM case study → +15 points
Add up the points, divide by total possible points, calculate "PMM influence percentage."
Result: 87% of deals had PMM touchpoints.
The problem: This measured activity, not impact.
A prospect might download a whitepaper and never read it. Sales might open a battle card and immediately close it. A demo might use PMM's script but the prospect was already convinced before the demo started.
Touchpoints don't equal influence. I was claiming credit for being present in deals, not for changing outcomes.
The CFO saw through it immediately.
Rebuilding Attribution: Direct vs. Assisted vs. Noise
I worked with RevOps to rebuild pipeline attribution around a harder question: Can we prove PMM changed the outcome of this deal?
Not "was PMM content present," but "did PMM content make a measurable difference."
We settled on three attribution categories:
Category #1: Direct PMM Attribution (We Created This Opportunity)
These are deals where PMM's work directly generated the pipeline. Clear causation.
Example: Product launches.
When we launch a new product:
- PMM develops positioning, messaging, and launch plan
- PMM runs launch campaigns (webinars, content, email sequences)
- Pipeline is created specifically for the launched product within 90 days
Attribution logic: No launch = no opportunity. PMM gets direct attribution.
How we track it:
Filter Salesforce:
- Product = newly launched product
- Create date = within 90 days of launch
- Source = launch campaign, launch webinar, launch content, or website inquiry during launch period
This gave us clean numbers: "Q3 product launch generated $8.4M in qualified pipeline."
No inflation. We didn't claim we influenced every deal where the new product was mentioned. We only claimed deals where prospects responded directly to launch activities.
Category #2: Assisted PMM Attribution (We Measurably Helped)
These are deals where PMM didn't create the opportunity, but measurably improved outcomes.
The standard: We need to show a before/after difference or a clear correlation between PMM involvement and better outcomes.
Example #1: Competitive Displacement
We track competitive deals where:
- Competitor is tagged in Salesforce
- Sales marked "battle card used" in opportunity
- Deal closed-won
Then we compare win rates:
- Competitive deals where battle card was used: 58% win rate
- Competitive deals where battle card wasn't used: 31% win rate
The 27-percentage-point difference is PMM's measurable assist.
Attribution logic: We didn't create these competitive opportunities. But we measurably improved win rates when sales used our battle cards.
How we track it: "$18.6M in competitive pipeline where PMM battle cards were used. Historical data shows 27-point higher win rate when battle cards are used vs. not used. Estimated PMM contribution: $5.0M in incremental wins."
Not claiming we influenced the full $18.6M. Claiming we contributed the delta between using battle cards and not using them.
Example #2: Sales Velocity Improvement
We track deals where:
- Sales used PMM ROI calculator during demo stage
- Deal progression from demo → proposal
Then we compare:
- Deals with ROI calculator: 14-day avg. time from demo → proposal
- Deals without ROI calculator: 26-day avg. time from demo → proposal
12-day difference × number of deals = time saved × value of faster progression.
Attribution logic: ROI calculator didn't create these opportunities. But it measurably accelerated progression through the sales cycle.
Category #3: Noise (PMM Content Present, No Measurable Impact)
These are deals where PMM content appeared but we can't prove it mattered.
Example: Prospect downloaded a whitepaper 8 months before becoming an opportunity. Did that whitepaper influence them? Maybe. Maybe not. We can't prove it.
Our approach: Don't claim it.
If we can't draw a clear line from PMM activity to measurable outcome change, we don't count it as influence.
What we do instead: Track it as "PMM content engagement" without claiming pipeline attribution.
Report it as: "PMM content was engaged with in 73% of closed deals. We assisted 31% of these measurably (Category #2), and directly created 8% (Category #1). The remaining 62% had content engagement but no proven impact."
This honesty made the CFO trust our numbers instead of dismiss them.
The Framework That Survived CFO Scrutiny
After rebuilding attribution, I presented new numbers to the CFO:
PMM Direct Attribution: $12.4M pipeline
- Product launches: $8.2M
- Competitive content campaigns: $2.4M
- ICP-specific positioning campaigns: $1.8M
Methodology: Opportunities created directly from PMM-led initiatives within 90 days.
PMM Assisted Attribution: $24.1M pipeline
- Competitive displacement (battle card usage): $18.6M (estimated $5.0M incremental wins based on 27-point win rate improvement)
- Sales velocity (ROI calculator usage): $5.5M (12-day acceleration on avg = reduced pipeline leakage)
Methodology: Measurable outcome improvement (win rate, velocity) correlated with PMM asset usage.
PMM Content Present (No Claimed Attribution): $31.8M pipeline
- Deals where PMM content was engaged but impact not provable
Methodology: Content tracked in opportunity, but no before/after or correlation data to prove influence.
Total Tracked Pipeline: $68.3M
- Direct attribution: $12.4M (18%)
- Assisted attribution: $24.1M (35%)
- Content present, no claimed impact: $31.8M (47%)
The CFO studied this for two full minutes.
Then he said, "This is the first pipeline attribution I've seen from marketing that I actually believe. The fact that you're not claiming credit for the 47% where you can't prove impact makes me trust the 18% and 35% where you do claim it."
Approved.
The difference between my first attempt (87% influenced!) and the honest framework:
First attempt: Claimed everything → Lost credibility Honest framework: Claimed only what we could prove → Gained credibility and budget approval
The Data Requirements for Honest Attribution
Building honest attribution required RevOps to track data I'd never asked for before:
Data Point #1: Campaign Source Tagging at Opportunity Level
For direct attribution to work, we needed clean data connecting opportunities to PMM campaigns.
RevOps built:
- UTM tracking on all PMM campaign URLs
- Campaign tagging in Salesforce for launch webinars, content downloads, and campaign responses
- Opportunity source field mapping back to specific PMM campaigns
This let us filter: "Show me all opportunities created within 90 days of Product X launch where source = launch campaign."
Clean attribution.
Data Point #2: Content Engagement Tracking in Opportunities
For assisted attribution, we needed to know which PMM content sales used in which deals.
RevOps integrated:
- Salesforce with sales enablement platform (tracked which battle cards/materials sales opened)
- Custom fields in opportunities for "battle card used" (yes/no)
- Content engagement data linked to contact records (tracked which prospects viewed which content)
This let us compare win rates in deals where battle cards were used vs. not used.
Data Point #3: Before/After Baselines for Every PMM Initiative
For every major PMM initiative, we started tracking baseline metrics before we launched the initiative.
Example: Before launching new competitive positioning against Competitor A:
- Baseline win rate: 51%
- Baseline sales cycle: 76 days
- Baseline discount rate: 18%
After launching (90 days later):
- New win rate: 62%
- New sales cycle: 68 days
- New discount rate: 12%
This gave us clean before/after comparison to claim impact.
Without baselines, we couldn't prove whether new positioning actually improved outcomes or if other factors changed (better product, easier competitive landscape, different sales reps).
The Uncomfortable Questions This Framework Forced
Honest pipeline attribution raised uncomfortable questions about which PMM work actually mattered.
Question #1: What Percentage of PMM Work Has Measurable Impact?
When I mapped all PMM activities over a quarter to pipeline attribution, the breakdown was sobering:
Activities with direct attribution:
- Product launches: 18% of PMM time → $8.2M direct pipeline attribution
- Competitive content campaigns: 8% of PMM time → $2.4M direct pipeline attribution
- ICP-specific campaigns: 6% of PMM time → $1.8M direct pipeline attribution
Total: 32% of PMM time → $12.4M direct attribution
Activities with assisted attribution:
- Competitive battle cards: 12% of PMM time → $5.0M assisted attribution (incremental wins)
- ROI calculators: 4% of PMM time → estimated $1.8M velocity improvement value
- Sales enablement training: 9% of PMM time → 18% faster ramp (calculated value: $2.2M)
Total: 25% of PMM time → $9.0M assisted attribution
Activities with no measurable attribution:
- Generic content creation: 22% of PMM time
- Internal stakeholder management: 11% of PMM time
- Ad-hoc sales requests: 10% of PMM time
Total: 43% of PMM time → No measurable pipeline impact
43% of PMM effort couldn't be connected to revenue outcomes.
This didn't mean that work was useless—stakeholder management and ad-hoc sales support might be necessary. But it forced the question: Should we reallocate effort from low-impact activities to high-impact activities?
Decision: Reduce generic content creation (low impact) and ad-hoc sales requests (reactive, low leverage). Increase time on launches, competitive programs, and sales enablement (high measurable impact).
Question #2: Should We Only Do Work We Can Measure?
The CFO asked a provocative question after reviewing our attribution framework: "If you can only prove impact on 53% of your work, should you stop doing the other 47%?"
I thought about this for a week before answering.
My response: "No. Some valuable PMM work isn't measurable with current data infrastructure, but that doesn't mean it doesn't create value."
Examples of valuable but hard-to-measure work:
- Brand positioning (creates long-term perception, no direct pipeline link)
- Thought leadership (builds market awareness, no attribution model)
- Analyst relations (influences market category, indirectly affects pipeline)
But— I committed to tracking the ratio of measurable-impact work vs. non-measurable work, and not letting non-measurable work exceed 40% of total effort.
If more than 40% of PMM time went to unmeasurable work, we were probably doing too much low-leverage activity disguised as "strategy."
How to Build This Without RevOps Infrastructure
If you're at a company without sophisticated RevOps infrastructure, you can still build honest attribution—it just requires more manual work.
Start with One High-Impact Activity
Don't try to attribute everything. Pick one PMM activity with clear before/after measurement:
- Product launches (easiest—clear time boundary, specific product)
- Competitive battle cards (measurable if sales tracks usage)
- Sales velocity improvements (measurable if you track time-in-stage before/after initiatives)
Build attribution for that one activity first. Prove PMM can measure impact honestly. Then expand.
Use Manual Tagging If You Don't Have Automation
If your CRM doesn't automatically track campaign source or content engagement:
- Ask sales to manually tag which battle cards they used in key deals (top 20 opportunities per quarter)
- Manually filter opportunities created during launch windows
- Survey sales quarterly on which PMM materials were most useful in closed deals
Manual tracking isn't perfect, but directionally correct data beats no data.
Calculate Impact Conservatively
When in doubt, underestimate PMM's impact.
If win rate with battle cards is 58% and without is 31%, you could claim the full 27-point delta. But safer to claim half of it (13.5 points) to account for other variables.
Conservative estimates build trust with finance. Inflated estimates destroy credibility.
The Real Value: Credibility With Finance and Exec Team
The honest attribution framework did something more valuable than justify PMM's budget.
It gave PMM a seat at strategic planning conversations.
Before honest attribution, PMM was seen as a creative function. We made messaging and content. Hard to measure, hard to value.
After honest attribution, PMM was seen as a revenue function. We could quantify impact:
- "Product launches generate avg. $8M in qualified pipeline per launch"
- "Competitive programs improve win rates 15-20 percentage points in 6-month timeframes"
- "Sales enablement reduces ramp time 25%, worth $2.1M in productivity gains"
The CRO started asking me:
- "What's the expected pipeline from next quarter's launch?"
- "How much competitive risk do we have based on recent win rate trends?"
- "What's PMM's ROI if we invest in another headcount?"
Those are revenue strategy questions, not marketing questions.
Honest attribution transformed PMM from "we make content" to "we drive measurable revenue outcomes."
And it started with admitting that presence in deals isn't the same as influence on outcomes.
Track what you can prove. Be honest about what you can't. Finance will trust you more, not less.