Why are so many new features going unused after launch?
The short answer: teams are shipping faster than they can communicate. When AI accelerates your engineers' output by 3x or 4x, your go-to-market motion does not automatically scale with it. Features go out the door, a Slack post goes up, maybe a quick Loom gets recorded, and then the feature sits there while customers keep using the same three workflows they already knew about.
This came up in a community thread I was following closely last month, and the responses were striking. One PMM put it plainly: "a ton of companies are shipping a ton of features that aren't getting used at all." Another described the current state as "drinking from a water hose." That is not hyperbole. That is a real description of what product marketing teams are living with right now.
What does the PMM role actually look like when velocity doubles?
Traditionally, a PMM would be involved before a feature shipped. There was a launch plan, positioning work, internal enablement, maybe a coordinated release communication with sales. That rhythm assumed that shipping was the rate-limiting step. It is not anymore.
What I'm hearing from people in the weeds is that the PMM job has become reactive. It's no longer planning launches; it's about marketing what is already out. The planning horizon has collapsed. Features are released before positioning exists for them. Sales does not know what to say about them. Customer success does not know to mention them. And so the feature sits.
This is not a failure of effort. The PMMs I've talked to are not coasting. They are throwing their hands up because the throughput mismatch is structural. You cannot hire your way out of it fast enough, and most companies are not even trying, because headcount requests take quarters to approve while features ship weekly.
How much does this actually cost?
Let's be concrete about the business consequences. If your engineering team ships 40 features in a quarter and only 10 of them get meaningful PMM support, you have 30 features that either go unannounced or get a cursory internal Slack message. Each of those features cost real engineering hours. They have real potential value to customers. They are just sitting inert.
On the sales side, reps are not mentioning features they do not know about. That means deals that could have been accelerated by a relevant capability get closed on older positioning. Upsell conversations that should happen do not happen. A customer churns because they did not know a workflow they needed had already been built.
I have heard the question asked semi-seriously in these discussions: "Are we at the stage of AI where management just doesn't care if the products launched aren't used? Is it just a matter of output for now?" That is a fair question to ask. The answer should be no. But the incentive structures inside a lot of companies right now are rewarding shipping over adoption, and that is a real problem.
Why do smart teams fall into this trap?
The flawed assumption is that releasing a feature is the same as delivering its value. Engineers ship, product closes the ticket, and the mental model in the room is that the work is done. What is actually done is the build. The distribution has barely started.
Before AI accelerated development cycles, this assumption was mostly harmless because the gap between PMM capacity and engineering output was smaller. There was usually enough time to catch up between releases. Now the gap is wide enough that features are stacking up before the previous ones are even understood by customers.
The horse is out of the barn on AI-accelerated shipping. Nobody is going back to slower cycles voluntarily. So the adjustment has to happen on the GTM side, and right now most teams are not making that adjustment in any structured way.
What does a practical adjustment actually look like?
A few things worth considering if you are operating in this environment.
First, triage is now a legitimate part of the PMM job. Not every feature deserves the same treatment. Deciding in advance which tier of release communication each feature gets, based on revenue impact and customer relevance, is a real skill and it deserves a real process. "Every feature gets a one-pager" is not a scalable standard anymore.
Second, the signal that a feature is being ignored is often invisible unless you are looking for it. Feature adoption metrics need to be surfaced regularly alongside release dates so you can see which launches landed and which ones disappeared. If you shipped something eight weeks ago and usage is flat, that is a problem you can still address. But only if someone is watching.
Third, the internal communication layer between product and sales is broken in most companies I talk to. Sales reps should not be learning about new features from a customer asking about them on a call. That is embarrassing and it costs deals. Making that handoff clear and consistent, even with lightweight tooling, is worth prioritizing before you hire your next PMM.
The goal is not to slow down shipping. It is to make sure that what gets built actually reaches the customers it was built for.
FAQ
Why are features going unused even when teams are shipping more than ever?
Feature adoption falls because product marketing capacity has not scaled with AI-accelerated development. Features ship before positioning, sales enablement, or release communication catches up, so customers never learn about them.
What is the business cost of features that go unannounced?
Unused features represent wasted engineering hours, missed upsell conversations, and deals closed on outdated positioning. At high shipping velocity, this can affect dozens of features per quarter.
How should PMM teams handle launch planning when velocity has doubled or tripled?
Triage is necessary. Not every feature warrants the same launch treatment. Tiering features by revenue impact and customer relevance, then assigning corresponding communication effort, is a more realistic model than uniform launch planning.
What is the first thing to fix if feature adoption is consistently low?
Surface adoption metrics alongside release dates so you can see which launches landed. Then look at your internal handoff process between product and sales, since reps cannot sell features they do not know exist.