Product marketing is the only team without a number
Revenue has a number. Customer success has a number. Marketing has a number. Product marketing does not. I recently talked with an executive from a 400 person company who swore that Product Marketing was useless. He employed multiple ones, but what really annoyed him, was that it wasn't measurable.
That's why product marketing doesn't get the respect it deserves. It's not that the work doesn't matter. It's that there's no number the team is being held to, so nobody can point at it and say "that worked" or "that didn't."
This post is how to fix that. If you're a B2B company, this is written for you. But honestly, you can bend it to fit almost anyone who ships product and has to sell it.
Why it was impossible to measure Product Marketing KPIs until now
I'm not going to pretend people were just lazy. These measurements genuinely couldn't be done. Two reasons:
1. No accountability. Nobody recorded calls. So if a launch happened and a rep either did or didn't mention the feature to a prospect, there was no way to know. It lived and died in someone's head.
2. No AI. The teams that did record calls couldn't do anything with them at scale. Unless a rep said the exact feature name word for word, you weren't finding it. Nobody was going to listen to thousands of hours of calls to check whether a feature came up.
Why it also didn't matter (until now)
Here's the other half of the story. Even if you could have measured this, you probably wouldn't have needed to.
1. Engineering couldn't ship that fast. Product teams just didn't release that much. There wasn't a firehose of features to keep track of.
2. Launches were big or they were nothing. There was no such thing as a small feature launch. You either threw a big launch or you said nothing at all. Small and mid-sized features went unmentioned because nobody could track that at scale.
Both of those things changed. Engineering now ships at a pace that would have been unthinkable a few years ago, which means the number of features that need to be sold has exploded. And calls are recorded, and AI can read all of them and tell you who mentioned what, even when nobody used the "official" wording.
So now you can measure it. And you need to.
The framework: how to set KPIs for product marketing
Think of this as how to tell if product marketing is working, how to tell if you have a good product marketer, and how to set KPIs for your PMM team. Same thing, three ways of asking.
The metrics fall into two buckets: communication (did the word get out?) and revenue impact (did it matter?). Every metric below uses a 30-day window from launch unless I say otherwise.
Part 1: Communication metrics
Sales enablement rate
The question: Did sales and account management actually mention the feature?
The stat: Of the prospects and customers who asked for a feature, what percent were told about it within 30 days of launch?
You can adjust this for meeting frequency too. Instead of counting people, ask: of the meetings with a group that asked for this feature, in what percent was the feature mentioned?
The target: 100%. Seriously. If someone asked for it and you built it, they should hear about it.
What it measures: How good a job you did educating your sales reps and account teams about the feature.
Marketing rate
The question: Did we drive demand for the feature?
The stat: Of the prospects and customers who heard about a feature, what percent asked about it within 30 days of launch?
What it measures: How good a job you did generating demand for a specific feature.
Full communication rate
The question: How effective were sales and marketing communication overall?
The stat: Add the sales enablement rate and the marketing rate together. You can express it as a percentage, an absolute number, or revenue.
What it measures: The total picture of how well a feature got communicated, across every channel.
Part 2: Revenue impact metrics
Everything in this section is about what the feature did to revenue. This is how you tell whether the right thing was built and whether it was sold well. That's a lot of accountability, and it's shared between product and product marketing, which is exactly how it should be.
Deal re-engagement rate
The stat: Of the deals that were lost or went dark, what percent scheduled a meeting after hearing about a given feature? You can also count it as number of deals or revenue.
What it measures: How well you filled the top of the pipeline with SQLs and MQLs who were blocked by a missing feature. A lot of dead deals aren't dead, they're waiting on you.
Deal velocity
The stat: Of the deals in progress (activity in the last 30 days), what percent closed in the 30 days after hearing about the feature? Again, you can use number of deals or revenue.
What it measures: How much faster a feature made a deal happen, or whether it made it happen at all.
I'll be upfront: this is correlation, not causation. Sorry! But a pattern of deals closing right after a feature comes up is a pretty loud signal, and it's a signal you've never had before.
Upsell rate
The stat: What percent of customers scheduled a new meeting after hearing about a feature? And a harder one: what percent of customers added licenses within 30 days of hearing about it? You can calculate this in revenue too.
What it measures: Whether a feature helped grow net revenue retention.
Cross-sell rate
The stat: For companies with multiple products, what percent of customers who heard about the feature added ACV within 30 days because of it, and not because they bought more licenses?
What it measures: Whether a feature pulled customers into another product line.
The part everyone skips: ROI
Here's where this gets fun. Once you have revenue tied to features, you can finally answer the question nobody could answer before: was this feature worth building?
The math:
ROI = the benefits of the feature ÷ the cost to build it
Benefits are the revenue from everything above: re-engaged deals, faster closes, upsells, and cross-sells.
Cost is the time it took to build, plus token cost. We measure time from the ticket being opened to the ticket being closed, and multiply it by an average cost for that time. It's not perfect, but it's consistent, and consistent beats perfect.
Think about what that does. You can now look at two features that shipped the same week and say, with a number, "this one returned 12x and this one returned nothing." That changes what you build next. It changes what you launch loudly and what you let quietly die. And it gives your product marketer something they've never had: a scoreboard.
What this means for your PMM team
If you're hiring a product marketer, or trying to figure out if the one you have is good, stop asking whether the launch felt big. Ask:
- What was our sales enablement rate on the last five launches?
- Did the demand we generated turn into questions from prospects and customers?
- Did dead deals come back? Did live deals close faster?
- What was the ROI?
If they can't answer, it's not necessarily on them. It's that nobody gave them the data. But now that the data exists, there's no excuse for flying blind.
Product marketing has always been asked to move revenue without being given a number. Now it can have one.
Want to see these numbers for your own launches? That's exactly what Arkweaver does. Book a demo.