What My fCTO Said in the Hallway
A few weeks ago, one of my fCTOs sat through a meeting at RSA with the CEO of a security vendor. The CEO pitched. My fCTO gave him green lights across the board. The product was great. The positioning was sharp. Every signal a seller hopes for, he gave.
We walked out. He turned to me and said, "I would never buy that."
The CEO is somewhere right now telling his board the meeting went well. My fCTO is telling me the truth he'd never tell the vendor. Both versions of that meeting now exist. Only one of them is real. Guess which one ended up in the CRM.
I've been thinking about this since I read Nathan Herrman's piece on building a Strategic Intelligence function. He's right about a lot of it. Intelligence functions are siloed by design. Competitive intelligence captured by product marketing becomes a proof-point factory instead of a pressure test. Most organizations are intelligence-rich on paper and intelligence-poor in practice. The architectural critique is sound and worth reading.
But there's an assumption sitting underneath the framework that the RSA story breaks. It assumes the signal you're capturing is real signal.
It usually isn't.
Sales conversations are performative on both sides. The buyer performs interest because confrontation is socially expensive and they're often early in their own process. The seller performs confidence because that's the job. Both walk out with a shared fiction, and every system we build to capture field intelligence—CRM notes, call transcripts, AI summaries—captures the fiction with high fidelity. Sophisticated garbage in, sophisticated garbage out.
The real signal in that RSA meeting wasn't in what my fCTO said. It was in what he didn't ask.
No procurement question. No integration concern. No "who else are you looking at?" No pushback on pricing or deployment timeline. A government CIO who's actually evaluating doesn't sit through a 30-minute pitch without those. Their absence was the data. But absence is invisible to a model that's only looking at what's there.
This is where the framework runs out of road. You can't instrument the hallway conversation. The moment you put a microphone on the peer-to-peer exchange between two people with shared context and no performance incentive, you turn it into another sales meeting. The unguarded version evaporates the second you try to capture it. That's not a tooling gap somebody is going to close with better AI. It's structural.
So if capture isn't the bottleneck, what is?
Perspective.
But not perspective the way most people use the word. Perspective isn't experience. Experience is necessary and not sufficient. Perspective is experience plus curiosity plus introspection—the discipline of interrogating what you've learned instead of just accumulating it.
Here's what I mean. Ask a hundred CxOs why you lose your budget if you don't spend it. Most will tell you it's because the organization is punitive, or because budgeting is broken, or because you have to spend it or lose it. They've been operating under that belief for years. They've made decisions based on it. They're confident.
They're wrong.
Your budget isn't a checking account. It's an investment the company made in your function to support the strategy. If you don't invest what you were given, you didn't just fail to do your job—you prevented someone else from putting that capital to work somewhere it could have produced returns. Losing the budget isn't punishment. It's the rational response to your failure to deploy capital you said you'd deploy. Hundreds of experienced executives will get this question wrong because nobody calls it out when you take the job, and nobody asks when they take it. People are so afraid of sounding stupid that they nod along and pretend they understand. They never ask the question that would get them the answer they don't have.
That's what I mean about perspective. If you don't ask, if you aren't obsessively curious, you'll never have a perspective worth anything. You'll just be experienced in ignorance.
Now extend that to the intelligence function we're talking about. AI does in an afternoon what used to take a team of experienced people months—research, synthesis, strategy, the whole production. That used to be hard. It used to be expensive. The friction was the quality control. Now the friction is gone, and the quality control has to come from somewhere else.
It has to come from perspective. From someone at the synthesis layer who knows which questions a real buyer asks, which budget realities actually shape decisions, what political cover looks like in a procurement cycle, what a CISO can and can't commit to without the CFO. From the decider, not the pitcher.
This is the part that should worry people, and it's the part nobody wants to say out loud.
AI makes you sound smart. It does not make you smart. Without perspective, you get fast, great-sounding strategy built on a false premise. The deck is sharp. The framework is clean. The narrative is tight. And it's wrong in ways nobody in the room is equipped to catch, because the same tool that produced the analysis is the one being used to evaluate it.
Then the strategy ships. The product gets repositioned. The campaign launches. The investment thesis gets funded. Six quarters later, somebody is explaining to a board why the numbers didn't move, and the answer—if anyone is honest enough to give it—is that the original read of the market was confident fiction. By then, the people who acted on it are gone. AI doesn't take the blame. It just produces the next deck for the next team.
Speed without judgment is just faster failure.
The obvious objection is: fine, hire former practitioners and put them at the synthesis layer. Problem solved. Most firms will tell you they already do this. The honest answer is that hiring practitioners is the easy part. Protecting their perspective from the pressures that grind it down—utilization targets, vendor relationships, content volume, the gravitational pull of "good enough"—is the hard part. Most firms buy expensive resumes and use them as garnish. The practitioner perspective gets diluted into the same generic synthesis everyone else produces, just with better credentials on the byline.
And even when you find the right people, you have to distill what they know into something that compounds. A working set of instructions, updated with what you learned today, refined the next day, kaizened relentlessly. You won't get it right on the first try. If you try, you'll never ship—perfection is the enemy of progress. So you aim for 80% fast and you work on the remaining 20% forever. The people who can do this work are rare, expensive, and have their pick of places to land. You have to find them, overpay them, and build an environment they don't want to leave.
This is the bet GigaOm is making. It's why ResearchOS exists, and why v1 ships this month.
Every executive reading this has been my fCTO in some hallway, telling someone what they really thought after telling the vendor what the vendor wanted to hear. Some of you, like me, told the truth occasionally. Sometimes it went well. Most of the time, it went nowhere. Nothing improved. Nothing changed. The vendor kept doing exactly what they were doing, wondering why they weren't winning. The Southern translation is "bless your heart, now get out of my office." The business translation is the polite nod that ends the meeting.
The data we're feeding into our intelligence systems is largely composed of that polite nod.
So here's what I want to know—from both sides.
If you've been the buyer in that hallway, I want to know what you didn't say in the room, and why.
If you're on the vendor side, here's a harder question. If a buyer told you exactly why they wouldn't buy—not the polite version, the real one—would you take it back to your team and do something with it? Or would your organization black-hole it because acting on it would be uncomfortable, or expensive, or politically inconvenient?
If the answer is black hole, that's the problem. The world is full of people for whom "good enough" is plenty and status quo is fine. For me, better tomorrow than today is the requirement. No rest for the wicked in that.
But it sure is fun.