How do you measure product marketing impact when KPIs can't be soley attributed?

This is one of the trickiest parts of PMM. We generally do not hold a number, but we enable teams that do hold a number (i.e. Demand Gen, SDRs, Sales, etc). 

I do ask my PMM Directors and above to own the top-level KPIs for those other teams as part of their OKRs, even though they don't own all of the areas that impact the results. For example, I believe a PMM Director should have our quarterly MQL and SQL goals as part of their OKRs.

Why? Because it forces them to deeply understand what impacts hitting those goals, and how they can tie their team's work to produce greater impact.

The single biggest value PMM can provide is to really look at a product line's success from TOFU to MOFU to BOFU, identify problems, and attack them cross-functionally. PMMs need to be that overlay business leadership across these teams to impact a product line's overall success.

I agree with this statement. Sometimes when I look at PMM resumes that say something like “Increased sales pipeline by 30%” or “Increased product adoption by 15%”, I’m often skeptical because how much credit can a PMM really take? Did you write all the content? Did you do all the work on campaigns, ads, paid performance, SEO, SEM, digital, video, webinars, and webpages?

PMMs operate through influence, not authority. We’re the strategic center of marketing—defining the strategy, personas, messaging, and execution. That being said, let’s at least start with the stuff we can take credit for:

• Personas - How well are the personas defined and how well does the marketing and sales org understand these personas? What research has been conducted? Which documents can we point to?
• Messaging - Good messaging is highly subjective but the key here is ensuring all messaging has been vetted by sales, customers, and internal experts. Is the messaging easily consumed by other stakeholders like content marketers?
• Sales enablement - If you’re B2B, PMMs are directly responsible for enabling the field on the market, competition, product positioning, messaging, pitches, and demos. Of course, this is all influenced pipeline but is the foundation there? If it’s not, you’ve got work to do.
• Campaign strategy - PMMs should be shaping and directing the themes of campaigns throughout the year and educating the marketing org on why a certain kind of campaign is needed. Campaign runners are responsible for driving those campaigns in market.
• Product launches - PMMs are often the quarterback for launches. How many launches can be accomplished per year? How organized are these launches? Are they reaching their target audience? Was the launch able to drive the expected amount of product adoption?
• Analyst briefings & thought leadership - Just keeping analysts informed and up-to-date is critically important for the business. Spearheading a Gartner MQ is a ton of work. Did you develop thought leadership themes with the comms team?
• Events - Supporting user conferences, tradeshows, and keynotes. How many field events did you support?

There's a lot I missed. Some of the above can be measured quantitatively but most are qualitative. If you take a step back, I would say a PMM can tie themselves to the holistic movement of core KPIs quarter-over-quarter. If you’re doing your job right, you should be able to claim influence on sign-ups, activation, pipeline, and close rates QoQ.

Very true; I've always said that Product Marketing influences everything but owns nothing. 

A few suggestions:

• When Product Marketing runs a launch, an event, or a campaign that wouldn't have existed unless PMM drove it, the engagement (website traffic, social mentions, etc.) can be attributed to Product Marketing. If the campaign also drove revenue or leads/MQLs/pipeline directly from the activities, make some noise about that too!
• If you want to measure the feature adoption impact of a launch, and you have an analytics team that can help, you can look at the adoption or product engagement of those who interacted with the launch (clicked on the email, visited the landing page, etc.) vs. those who didn't — that's Product Marketing's impact! 
• This is trickier but if you do a big sales enablement effort — a new pitch deck, a sales play, a package of enablement assets, or a specific competitive battlecard & training — and most other factors remain relatively constant, you can attribute an increase in win rates (especially competitive win rates, in the case of compete enablement) to Product Marketing. In a fast-moving and dynamic company it's pretty rare to have everything else mostly the same, so you can always survey the sales team on how confident they feel and measure that over time as you do more enablement. 

But my most important suggestion is this: invest in educating cross-functional partners and leadership teams on the fact that Product Marketing influences everything but is notoriously difficult to measure. A team that makes everything more successful but whose impact can't be measured in a cut-and-dry way is still a worthwhile investment to me :)