Your Metrics Look Great. Your Pipeline Doesn’t.

We put a room full of marketing leaders on the spot and asked them to tell the truth about their pipeline and their stack. Here is what they said, and what to do about it.


Before Bill and I said anything, we put one question on the screen and asked the room to be honest.


From the room · Poll 01

Which statement sounds most like your reality right now?

Our metrics look great but our pipeline doesn’t33%
We spend more time managing the stack than running programs33%
We can’t tell the CFO a clean revenue story29%
We’ve added AI tools and pipeline still hasn’t moved5%

Nobody picked a happy answer. The two responses about the metrics-to-pipeline gap and the missing CFO story came in at nearly two-thirds of the room combined.


The top three answers were the same complaint wearing different clothes. Metrics that look good. A stack that eats the day. A revenue story you can’t tell the CFO with a straight face. That is the whole session in one slide. So that is where we started.


More of Everything. Except Pipeline. 

None of these teams are doing less. They have more dashboards, more intent data, more AI, more activity than they did two years ago. The instrumentation keeps climbing. The one number that pays the bills sits still.

67%
say their dashboards show success metrics that never turn into revenue
87%
report intent signals they consider unreliable or inflated
26%
of all intent scores actually convert to qualified opportunities

Read that last one twice. Three out of four intent scores go nowhere. So when we asked the room how much they trust the intent data they are paying for, the answer was not subtle.


From the room · Poll 02

How much do you trust your current intent data?

What intent data?39%
I look at it but verify before acting36%
I mostly ignore it21%
I trust it and act on it4%

Four percent. That is how many will act on their intent data without checking it first.


Everyone else verifies, ignores, or has quietly given up. That is not a problem you fix with a better dashboard. That is a trust problem.

Here is the part worth sitting with. The mirage is not universal. In the same research, the teams that beat it look measurably different from the ones still stuck inside it.


What winners do differently

  Low performers High performers
Attribution coverage42%89% +47 pts
Tools used (avg)13–15+9–10 18% fewer
Budget waste30%23% 7 pts less
Data confidence74%91% +17 pts

Source: 2026 State of Performance Marketing: The Marketing Data Mirage, DemandScience. Survey of 750 senior marketing leaders.


Read down the right column. Fewer tools, not more. Nearly double the attribution coverage. Less waste, and far more confidence in the numbers. The winners did not out-spend the mirage. They out-simplified it.


Same Maze. Smaller Doors. 

Most teams already know the legacy ABM platform is not coming through. Nearly every CMO and VP of demand gen we spoke to who runs on one is not planning to renew.

Here is the trap. They leave the big platform and rebuild the same complexity out of six lighter tools. De-anonymization here. An AI SDR there. Intent, enrichment, content AI, conversation intelligence, all stitched together with workflow automation.

And every one of those tools carries a platform tax. Not just the subscription. The people time to learn it, run it, and get anyone to adopt it. Sales usually does not. So the tool drifts back to being a marketing-only thing, which is exactly the opposite of the point.

AI was supposed to be the relief. So we asked.


From the room · Poll 03

Has AI actually improved your pipeline yet?

Some efficiency, but no pipeline lift38%
Too early to tell27%
It’s added noise, not output19%
Yes, measurably15%

Only 15% had seen AI move pipeline in a way they could measure. Nearly one in five said it had added noise, not output.


That tracks. AI creates leverage when it is designed into a revenue system. Bolt it onto a broken process and all you get is faster noise.


Run Brand and Demand As One Motion.

So what actually works. The answer is not buy more tools. It is one operating motion across brand and demand. Same accounts. Same signals. Same content themes. Same activation logic. Same measurement.

Most teams run these in separate worlds. Programmatic lives in one system. The opt-in list lives in the MAP. Sales outreach lives somewhere else. If a buyer is not on the opt-in list and the outreach does not land, you miss them entirely. You are not even in the room.


Put brand and demand on one account system and the math changes. Teams that unify the two generate around 2.5x more leads inside their ICP.


Same audience. One motion instead of three. The buying system changed. The marketing system didn’t. That is the entire gap.


Five Moves You Can Start This Quarter.


01

Audit

Find the places where your stack creates visibility but not action. The classic tell is an intent report saying a company clicked a topic, with no named contact and no next step. Visibility you can’t act on is just a more expensive way to feel busy.

While you are in there, find where your win rates are highest and model the signals that show up in your closed-won deals.

02

Unify

Put brand and demand on the same account system. Programmatic, your opt-in list, and sales outreach stop living in three separate worlds. One audience, one motion, one set of numbers.

This is the rebuild from a minute ago, turned into a standing operating model instead of a one-time fix. It is also where the 2.5x lift inside your ICP comes from.

03

Re-aim

This is the big one. Knowing who is researching is not the same as knowing who is winnable. Size and industry will not tell you. Winnability shows up elsewhere: the tech they already run, a competitor’s renewal coming due, displacement signals, a complementary tool they just adopted, how they spend.

We have modeled winnable accounts from outside data alone and had CMOs tell us we figured out their plays better than they had. Shift budget toward the winnable subset without walking away from the broader ICP.

04

Trigger

Use visitor identification and engagement signals to drive the next best action, not just another report. When a winnable account shows up on the site or lights up on a signal, that should fire a specific move: a play, a hand-raise to sales, the right piece of content.

A signal that doesn’t trigger an action is just trivia.

05

Measure

Stop reporting activity. Track qualified pipeline. Specifically stage 2, the opportunities sales has given a 20% or better chance of closing. Watch how fast pipeline moves, not just how much of it there is. Tune campaigns every two to four weeks, not every quarter.

The target to chase is engagement from your named accounts climbing from around 5% to 20, 30, and up.


Then we asked the room what they would actually do first.


From the room · Poll 04

What’s your next move after today?

Re-aim from “who’s researching” to “who’s winnable”44%
Audit where my stack creates visibility but not action28%
Unify brand and demand on one account system22%
Rethink how I report pipeline to the C-suite6%

Almost half the room picked re-aim as their next move. They didn’t just nod along. They picked the work.



Build, Buy, or Manage.

Once you accept the old platform is done, the question becomes how you build the new one. Three options, and most teams use a mix.

Build

Build only what is truly unique to you. Your data, your logic, your edge. Don’t rebuild commodities. And know the risk: the day the GTM engineer who built it leaves, you find out how much lived in their head.

Buy

Buy proven capabilities you don’t need to own. Just watch for sprawl. Ten to twenty tools and you get overlap across intent, enrichment, AI SDR, content AI, and conversation intelligence. Your CFO will eventually audit for the duplication. Get there first.

Manage

Hand the motion to a partner who runs it for you. Lower overhead. Time to value in days and weeks instead of quarters. A useful way to keep producing pipeline while you sort out what to build and what to buy.

And yes, boards are now asking companies to report stack spend out loud. The duplication question is coming whether you raise it or not.