The VP of Sales asked me a simple question in our QBR: "What revenue did product marketing actually contribute this quarter?"
I opened my mouth to answer and realized I had no idea. I knew we'd launched two products, created competitive battle cards, trained the sales team, and produced dozens of enablement assets. But revenue contribution? I had no concrete number.
"We influenced a lot of deals," I said weakly. "Sales uses our materials all the time."
The VP nodded politely and moved on. I could see exactly what he was thinking: If you can't measure it, how do you know it matters?
That conversation haunted me. Product marketing was doing important work—I knew it, sales reps knew it, our customers benefited from it. But I couldn't prove it in the language executives understand: revenue impact.
So I spent the next six months building PMM attribution models. I tried three different approaches, made every mistake possible, and eventually landed on something that actually worked. Not a perfect system—those don't exist—but one that survived scrutiny from both our CFO and our CRO.
Here's what I learned about measuring PMM's impact without losing your mind or your credibility.
The Vanity Metrics Trap
My first attempt at PMM attribution was embarrassingly bad.
I built a dashboard tracking:
- Number of battle cards created (24 last quarter!)
- Sales enablement sessions delivered (8 trainings, 156 attendees!)
- Product launches completed (3 major launches!)
- Marketing collateral produced (47 assets!)
I presented this to the executive team feeling proud. The CFO looked at it for exactly four seconds.
"That's nice. But which of these activities generated revenue?"
I didn't have an answer. I'd built an activity tracker, not an attribution model. I was measuring outputs, not outcomes. The number of battle cards I created meant nothing if sales wasn't using them to close deals.
The fundamental problem: I was tracking what PMM produced, not what PMM achieved.
This is the trap most PMMs fall into. We measure what's easy to count—trainings delivered, content created, launches completed—because it feels like progress. But executives don't care how busy you are. They care whether customers bought more because of what you did.
I deleted that dashboard and started over.
The Multi-Touch Attribution Rabbit Hole
My second attempt went to the opposite extreme. I decided to build a sophisticated multi-touch attribution model that would track every PMM touchpoint in the buyer journey.
I created a system that tracked:
- When prospects downloaded PMM-created content
- When they attended product demos (using PMM demo scripts)
- When sales used battle cards in competitive deals
- When they engaged with launch messaging
- When they viewed pricing pages (created by PMM)
- When they consumed case studies (written by PMM)
I assigned weighted scores to each touchpoint. Battle card usage in a deal got 15 points. Demo attendance got 10 points. Content downloads got 5 points. I built formulas that calculated PMM's "contribution percentage" to every closed deal.
The result: PMM had influenced 87% of all closed deals last quarter!
I brought this to our VP of Revenue Operations. He pulled up the raw data and started asking questions.
"This deal—you're claiming 35% PMM attribution because the prospect downloaded three whitepapers and attended a webinar. But they were already in active conversations with sales before any of that happened. How do you know PMM influenced the outcome?"
"This competitive deal—you're taking credit because sales opened the battle card. But they didn't actually use it. The prospect asked about the competitor, sales looked at the battle card for ten seconds, then closed the tab and won based on pricing. Did PMM really influence that?"
"This enterprise deal—you're claiming attribution because they attended a product demo. But they attended demos for six different features over eight months. Which demo actually moved the deal forward?"
He was right about all of it. My model was attributing correlation, not causation. Just because PMM assets were present in a deal didn't mean they influenced the outcome.
The fatal flaw: I was tracking PMM touchpoints, not PMM impact.
Multi-touch attribution sounds sophisticated, but it's almost impossible to implement honestly. You end up either inflating your numbers (claiming credit for touchpoints that didn't matter) or building such a complex model that no one trusts it.
I needed something simpler and more honest.
What Actually Worked: The Three Attribution Buckets
After two failed attempts, I finally built an attribution model that survived executive scrutiny. The secret was admitting what I could and couldn't measure.
I stopped trying to track every PMM touchpoint and instead focused on three types of measurable impact:
Direct Attribution: Clear PMM-to-Revenue Connection
These are deals where the connection between PMM's work and revenue is undeniable.
Example 1: Product Launch Pipeline
When we launched a new product, I tracked every opportunity created where:
- The opportunity was created within 90 days of launch
- The primary product in the deal was the newly launched product
- The opportunity source was campaign response, webinar attendance, or launch-related content
This gave me a clean number: "The Q2 product launch generated $8.4M in qualified pipeline."
No ambiguity. No inflated claims. A new product launched, we ran launch campaigns, pipeline was created for that product. PMM gets attribution.
Example 2: Competitive Displacement Wins
I worked with our sales ops team to tag competitive deals in Salesforce. For deals tagged as "competitive," I tracked:
- Which competitor we were against
- Whether sales marked "used battle card" in the opportunity record
- Whether we won or lost
Then I calculated: In competitive deals where sales used battle cards, we won 62%. In competitive deals where they didn't, we won 34%.
The difference—28 percentage points—represented PMM's impact. Applied to our average competitive deal size, that translated to $3.2M in competitive displacement wins I could attribute to PMM.
Clear causation. Battle card usage correlated with significantly higher win rates.
Example 3: Sales Productivity After Training
When we launched new messaging and trained the entire sales team, I measured:
- Average deal size 60 days before training
- Average deal size 60 days after training
- Win rate before and after
Average deal size increased 18% after training. Win rate increased 6 percentage points. I couldn't prove the training caused all of that improvement, but I could say: "Sales performance improved materially after PMM delivered new messaging and training, worth an estimated $2.1M in additional revenue."
These three buckets gave me $13.7M in direct, defensible attribution for the quarter.
Influenced Attribution: Assisted, Not Owned
These are deals where PMM clearly helped, but wasn't the primary driver.
I stopped trying to calculate precise influence percentages. Instead, I tracked binary signals:
- Did sales use PMM content in this deal? (Yes/No)
- Did the prospect engage with PMM-created assets? (Yes/No)
- Was this a vertical/persona that PMM had trained sales on? (Yes/No)
Then I simply reported: "PMM assets were used in 76 deals worth $24.3M in closed revenue. We assisted these deals but don't claim primary attribution."
This was honest. PMM helped, but so did sales, SDRs, customer success, the product itself, and probably competitive missteps. I wasn't going to pretend I could calculate our exact percentage.
RevOps appreciated the honesty. The CRO said, "This is more credible than if you'd claimed you influenced 80% of that revenue."
Strategic Impact: Measurable But Indirect
Some PMM work has clear business impact but no direct revenue attribution.
Example: ICP Refinement
I analyzed two years of customer data and discovered our highest-performing segment was mid-market healthcare companies—much better unit economics than other segments. I built a business case that convinced the exec team to reallocate 40% of marketing budget toward healthcare-specific campaigns.
Result: Healthcare pipeline increased 3x within six months. I couldn't claim revenue attribution (demand gen ran the campaigns), but I could claim strategic impact: "PMM analysis drove ICP refinement that generated $12M in new healthcare pipeline."
Example: Pricing Strategy
I ran competitive pricing analysis and customer willingness-to-pay research that informed a pricing restructure. The new pricing increased average deal size 14%.
I didn't negotiate those deals. Sales did. But PMM's research enabled the pricing change. Strategic impact: "PMM pricing research drove strategy change worth $4.8M in increased deal value."
These three buckets—Direct Attribution, Influenced Attribution, and Strategic Impact—gave me a complete story of PMM's contribution without inflating numbers or claiming credit I hadn't earned.
The Dashboard That Convinced Executives
I rebuilt my dashboard around these three buckets. Here's what I showed in quarterly business reviews:
Direct PMM Attribution: $13.7M
- Product launch pipeline: $8.4M
- Competitive displacement wins: $3.2M
- Sales productivity improvement: $2.1M
PMM-Assisted Revenue: $24.3M
- 76 closed deals where sales used PMM content
- (No percentage claimed—just the assist)
Strategic Impact:
- ICP refinement drove $12M in new segment pipeline
- Pricing research drove $4.8M in deal value improvement
Total Measurable Impact: $54.8M
The CFO studied this for a full minute. Then he said, "This is the first PMM metrics deck I've seen that I actually believe."
What made it credible:
- I wasn't claiming credit for every deal where PMM assets existed
- I separated clear attribution from assisted attribution
- I admitted when I couldn't measure precise impact
- Every number had a clear methodology behind it
The RevOps team loved it because the data was clean and auditable. The CRO loved it because it showed PMM's value without absurd claims that we influenced 90% of all revenue.
What I Stopped Trying to Measure
Building an honest attribution model meant admitting what I couldn't measure:
I stopped trying to track content consumption as attribution.
Just because a prospect downloaded a whitepaper doesn't mean it influenced their buying decision. Maybe they read it, maybe they didn't. Maybe it changed their perspective, maybe it confirmed what they already thought. I couldn't know.
Now I track content consumption as an engagement metric, not an attribution metric.
I stopped trying to attribute brand awareness to revenue.
PMM does work that builds long-term brand perception—category positioning, thought leadership, analyst relations. This work matters, but I can't draw a straight line from "we're positioned as a leader in Gartner" to "we closed $10M more revenue."
Now I track brand metrics separately and don't pretend they're revenue metrics.
I stopped trying to measure "influence" with decimal precision.
I deleted all formulas that calculated things like "PMM influenced 23.7% of this deal." That's false precision. I have no idea if PMM influenced 23.7% or 15% or 40%. I just know we helped.
Now I track binary influence: Did PMM assist? Yes or no.
Letting go of these unmeasurable things made my attribution model more credible, not less.
The Hard Truth About PMM Attribution
After six months building attribution models, here's what I learned:
Most PMM work can't be precisely attributed to revenue.
You can measure activities. You can measure outcomes. But connecting the two with mathematical certainty? Almost impossible.
Product marketing creates the conditions for revenue—better positioning, clearer messaging, more confident sales reps, more informed buyers. But sales closes the revenue. The product delivers the value. Marketing generates the leads. Success requires all of it.
Trying to calculate PMM's exact percentage contribution to every deal is a fool's errand. You'll either inflate your numbers and lose credibility or build such a complex model that no one understands it.
But you can measure enough to prove PMM's value.
You don't need perfect attribution. You need defensible attribution.
Track the deals where PMM's contribution is undeniable—product launches, competitive wins with battle card usage, measurable sales productivity improvements.
Track the deals where PMM clearly assisted without claiming ownership.
Track the strategic decisions where PMM's analysis drove high-impact changes.
Add it up and you'll have a number that's honest, credible, and substantial enough to prove PMM's value to even the most skeptical CFO.
The best attribution model is one executives trust.
I've seen elaborate attribution models that claimed PMM influenced 85% of all revenue. No one believed them.
I've seen simple attribution models that claimed PMM directly contributed to $12M in revenue and assisted with $30M more. Everyone believed those.
The difference wasn't sophistication—it was honesty.
Build an attribution model that admits its limitations, focuses on clear causation, and separates direct impact from assisted impact. RevOps will trust it, executives will believe it, and you'll finally have an answer when someone asks, "What revenue did PMM actually contribute this quarter?"
The answer might be smaller than you hoped. But it'll be real.