Ritu Kapoor is the CMO of Observe.AI, she's managing two completely different product lines - a mature conversation intelligence platform and a one-year-old AI agents business, each requiring its own GTM strategy. This is her third stint as a CMO, and she's spent her career moving from product marketing to demand to revenue leadership.
In our latest Moneyball CMO conversation, she said something that made me reimagine the marketing funnel:
"Funnels track averages. Buyer journeys are nonlinear. We've built systems to track averages, then we punish people for not following the path."
Here's what that actually means in practice:
A lead qualifies as SQL and Sales engages with it. But then the lead goes quiet or pushes back the timeline. It gets moved back to MQL or marked as "recycled." Your dashboard shows it as a loss. Marketing gets blamed for poor lead quality. Sales gets frustrated about wasted time.
Everyone treats it as failure. In reality? It's learning. It's a signal that you now have a better estimate of where this account actually is in their journey. Maybe they need more time. Maybe a different stakeholder needs to be engaged. Maybe the timing isn't right yet.
But our systems punish learning instead of capturing it as context.
The Funnel Problem We've All Been Ignoring
Ritu told me she recently spoke with one of the founder. His take:
"I created this funnel for a specific purpose. The way it's used in marketing is not how I envisioned it."
According to her, what most people miss is that the funnels were designed to track averages across populations, not predict individual buyer behaviour.
On average, your ICP goes from MQL → SQL → Opportunity → Closed Won. That's the macro motion. But no individual buyer follows that exact path.
Yet we've built entire operations around the assumption that they will.
"If people don't follow that specific path, we either disqualify them, which is strange, or we say they've digressed from path, there is no path."
The consequence? We're skipping buyers who are genuinely interested but don't fit our linear progression model.
Think about how you actually research and buy software. Maybe you check out a few vendor websites one afternoon. Then work gets busy and you forget about it for two weeks.
Someone mentions the problem in a meeting, so you're back researching. A colleague shares their experience. You finally look at pricing, realise budget approval needs to happen next quarter, and the whole thing goes dormant for months.
That's not a broken buyer journey. That's just how buying works.
But if you were being tracked in a traditional funnel, you'd look like a "problem lead."
Ritu's point: Buyer journeys are ambivalent by design. Funnels capture the average macro motion.
The tension between these two is where the real signals live.
"We're so used to doing things at scale, so used to doing things in buckets and understanding things only in a certain progression, that there's a lot of buyers in the market that we will just miss."
The stages we created to operate at scale are now working against us.
Attribution Is not About Credit - It is Institutional Memory
The Old Way :
Most companies treat attribution like a courtroom. Every channel is fighting to prove it deserves budget. Marketing and sales are in constant debates about what "counts."
The New Way :
Ritu reframes this entirely. She gave example of a ebook which was part of 10 buyer journeys. It's not like that the ebook converted this particular customer from S1 to S2. It was part of 10 buyer journeys. Something must be working.
Attribution shouldn't be an audit mechanism.
It should be a system for preserving institutional memory of what actually happened across the account.
When you think about it this way, every interaction becomes context. Every touchpoint is part of the story. You're not assigning credit - you're building a timeline of engagement that helps you understand momentum.
This is critical for enterprise deals where 6-15 people are involved in the buying decision, each with different roles, different concerns, and different moments of engagement.
The question isn't "which channel gets credit?"
The question is "what sequence of interactions moved this account forward, and how do we replicate that pattern?"
The Chandelier Metaphor That Changes How We Think About Brand
I've heard dozens of explanations for why brand and demand should work together. Ritu's is the best I've heard.
"Think about brand as a chandelier in the home that's shining a light on everything. Then you suddenly turn off the lights—you still remember the room because you still had the chandelier. Now I take a flashlight and flash it on one specific point in that room. Brand showed me what the room looked like. Demand allowed me to focus on that one ant on the wall."
Here's the insight: If you only do demand without brand, you'll spend 3X the money shining flashlights in the dark.
Why would someone open an email from a company they've never heard of? They won't. But if they've seen your content, received a thoughtful gift, or read about you in an industry publication, suddenly that demand email has context.
The emotional piece matters too. People need to relate to you.
They have to see you multiple times before they'll even open your email. "How do I get you to open that demand email? That's a brand exercise."
The challenge is the Measurement.
As Ritu put it:
"It's hard to measure because they do go together. You start measuring them separately and it's just hard - that is where tools like revsure.ai comes in"
Because we all know. The things we can't measure are often the things driving revenue.
The CMO Role Is Evolving (Or Disappearing)
Ritu made a prediction that will make a lot of people uncomfortable:
"I'm guessing if I look past the next 10 years, this becomes one role. Your demand gen and your sales execution - that becomes one leader. Because today, more and more CMOs are being held accountable to revenue."
Two people owning revenue on the same executive team? That's never a sustainable model.
She's not saying sales takes over marketing. She's saying the future might be a "President of Field Operations" who owns the entire motion, whether that person came up through marketing or sales doesn't matter.
We're already seeing early signs:
CMOs now own SDRs and BDRs
Stage 2 pipeline is becoming a marketing metric
Revenue accountability, not just pipeline generation
The difference between VP Marketing and CMO?
It's not the title. It's whether you have a seat at the table and whether you own revenue outcomes.
"Often you would see everybody else would have a seat at the table. You'd be on the exec team. And then the head of marketing would be a VP. I don't see that anymore."
The AI Tool Question Nobody's Asking
Everyone's racing to adopt AI SDRs, AI campaign tools, AI everything.
Ritu's using them too but in "really small portion sizes." Her question before scaling any AI tool: "Does my team know how to use the output?"
She's testing AI SDRs in a brand new market specifically, so her team can learn how to work with the output before deploying it in the US - where the stakes are higher and the conversations more nuanced.
"It's actually not the technology, it is the people using the technology that we have to focus on."
The Billboard on 101
Ritu shared an example that every CMO will feel in their bones: "If most of my revenue is now coming from this billboard on 101 and I cut that because I don't have any data on it, I just broke the buyer journey."
This is the impossible position CMOs are in:
Boards demand measurable ROI on everything. But buyer journeys are increasingly dark, nonlinear, and unmeasurable. So we're pressured to cut the things we can't prove, which are often the things that matter most.
The channels that are easiest to measure aren't always the channels that drive revenue. And the channels that drive revenue aren't always the channels we can attribute cleanly.
Ritu's point: It takes intuition, experience, and courage to know where to invest when the data is incomplete. That's the art side of marketing that nobody wants to talk about.
What This Means for You
If you're a CMO or marketing leader:
Stop punishing leads for not following your funnel. Your stages are averages, not absolutes. If a lead moves backwards- that's a signal, not failure.
Reframe attribution as memory, not credit. Build a timeline of engagement across accounts. Focus on patterns, not individual touchpoint ROI.
Invest in brand to make demand cheaper. You can't afford to only do demand. Brand makes every demand dollar 3X more effective.
Own revenue, not just pipeline. If you want a seat at the table, you need to speak the language of business outcomes - not marketing metrics.
Ask if your team can use the output Before adopting any AI tool, make sure your organisation has the capability to act on what it produces.
The tools will keep changing and the access will democratise, AI will make more things possible.
But the ability to think from first principles, to see systems instead of channels, and to have the courage to invest in right channels when the data says something opposite to gut feeling. That stays rare.
Watch the full conversation with Ritu Kapoor here:
Moneyball CMO is a series featuring CMOs who challenge conventional B2B marketing wisdom with data, candour, and real transformation stories. Know a CMO who should be on the show? Reply to this email.