.svg)
Ignite Your GTM With AI, Chapter 5: Stop Optimizing "Human" Emails and Start Engineering Situational Intelligence
Welcome to the next installment of our deep dive into "Ignite Your GTM With AI." Today, we are dissecting Chapter 5: Market Intelligence & ICP Definition.
In our previous chapters, we explored the structural collapse of traditional SaaS growth models. Now, we are going to look at the specific mechanism that most companies use to try and fix it—and why that mechanism is completely broken.
If you have been on LinkedIn recently, you have seen the advice. It usually looks something like this: "To make your AI outreach feel more authentic, add intentional typos. Make it look like a human wrote it in a hurry."
Jordan Crawford, Founder at Blueprint and the key contributor to this chapter, cuts through this noise immediately: "If you're optimizing your emails so that they are thought to be more 'human' by adding typos, you've missed the plot."
He is blunt because the situation demands it: "People don't think, 'Oh, a typo, great, I'll reply.' That's the dumbest take."
This obsession with surface-level "humanization"—worrying about em dashes, typos, or slang—masks a much deeper strategic crisis.
While leadership teams are wrestling with high-level AI strategy, their market intelligence efforts are trapped in pre-2020 thinking. They are applying band-aids to a broken limb. As Jordan notes, "You're just debating whether it's human garbage or AI garbage. But you've already decided it's garbage."
Chapter 5 is about stopping the debate over garbage. It is about fundamentally reimagining how we define our market, understanding why your "Ideal Customer Profile" (ICP) is likely an illusion, and how to use AI not to write better copy, but to detect the specific situations that make buying your product inevitable.
The Optimization Death Spiral
Most revenue organizations feel simultaneously overwhelmed by new AI possibilities and stuck in declining performance.
The natural reaction to this pressure is to optimize. We ask:
- How do we make our mass outreach feel more personal?
- How do we scale our existing processes faster?
- How do we use AI to do more of what we were already doing?
These are the wrong questions.
Jordan Crawford calls this the "optimization death spiral."
"We keep reaching for the volume knob," Jordan explains. "But people say, 'What if I keep the volume high and just change a few letters?' Or, 'What if I have an AI SDR blast out more so it's cheaper for me to send the thing that never worked in the first place?' That's insane."
The Jevons Paradox of AI Marketing
This behavior is driven by what economists call the Jevons paradox.
Historically, when a resource becomes more efficient and cheaper to use, people use more of it, not less. AI has made the cost of generating "intelligence" (or at least, text that looks like intelligence) effectively zero.
Because the tools are powerful but blunt, and because they are cheap, revenue teams have become "the knife sharpener."
They are trying to turn up the volume of bluntness rather than improving the quality of the knife.
In a world where every customer is overwhelmed by digital noise, doubling down on volume is fighting yesterday's war with obsolete weapons. The companies that survive the next few years won't be the ones that make bad approaches marginally better.
They will be the ones that rebuild their understanding of customers from the ground up.
The ICP Illusion: Why "Who" Does Not Matter
The root of the volume problem usually lies in the data foundation.
Ask most B2B companies to define their ICP, and they will recite a list of firmographics: "Companies between $10 and $500 million in revenue, with 100 to 1,000 employees, in industries like internet and technology."
Jordan points out the immediate absurdity of this definition: "Honestly, that could be everyone reading this book, it’s probably their ICP too."
When you start with a list that includes everyone, you force your team into a "personalization trap."
You have a list of thousands of undifferentiated companies. To make them pay attention, you try to personalize the message based on their persona. You use tools to scrape LinkedIn, find a recent post, or note that they have a podcast.
The resulting message usually sounds like this: "Hey Jonathan, you have a microphone, I have a microphone, we’re basically the same person. Could you please get out your wallet and send me $100,000 in cash?"
The Personalization Fallacy
Chapter 5 challenges the sacred cow of modern demand generation: the belief that better personalization solves targeting problems.
Jordan offers a litmus test. Imagine you had 100x better personalization. Imagine you could say, "Hey Jonathan, I know you woke up at 8:55, five minutes before this interview, and you were barely awake."
That is hyper-personalized. It is also completely true.
But, as Jordan notes, "It has nothing to do with whether I want to buy your AI tech. It has nothing to do with whether I’m in a situation where I need the product."
The hard truth is that "You can’t personalize your way out of a targeting problem."
The Cannonball Way: From Demographics to Situations
If demographics are dead and personalization is a trap, what is the alternative?
Jordan Crawford proposes a methodology he calls the Cannonball Framework.
The core shift is moving from Demographic Intelligence (who they are) to Situational Intelligence (what is happening to them).
"First, you need to identify your best-performing segment," Jordan advises. "And by segment, I don’t mean mid-market. Throw away all of your garbage tool thinking."
Instead of looking for companies of a certain size, you must look for companies in a certain situation.
The Three-Question Discovery Process
To find these situations, you cannot rely on call transcripts or CRM notes. You have to talk to your customers. But you have to ask the third question.
- Level 1 (Basic): "Can you do Discover card?" (Feature validation).
- Level 2 (Why): "Why is this useful to you?" (Surface-level benefits).
- Level 3 (Situational): "What changed in your organization? What made you need this now?"
When you ask the third question, you uncover the specific event or vulnerability that drove the purchase.
Jordan illustrates this with a client in the secure file transfer space. By interviewing customers, they discovered people weren't buying because they loved "secure file transfer" as a category. They were buying because they had specific security vulnerabilities.
Even more specifically, they found a competitor product—maintained by a single guy who lived on a boat in Lake Tahoe—that was riddled with unpatched security holes.
The segment wasn't "Enterprise Healthcare Companies." The segment was "Companies using the Lake Tahoe competitor who are currently vulnerable."
This leads us to the two most powerful concepts in Chapter 5: Pain Qualified Segments (PQS) and Permissionless Value Propositions (PVP).
Pain Qualified Segments (PQS): The Diagnostic Approach
Once you have identified a specific situation (like the vulnerable competitor software), your messaging changes completely. You no longer need to be "clever." You just need to be accurate.
A Pain Qualified Segment (PQS) message works by re-describing the prospect's situation with precision.
It follows a specific structure:
- Situation: "My research indicates you're using [Competitor X]..."
- Problem: "You've probably noticed [Specific Pain Point]..."
- Insight: "What you might not know is [Context explaining why the problem exists]..."
- Validation: "Other companies in your situation told us they felt trapped. Does that resonate?"
Notice what is missing.
"Notice what's missing: any mention of my product," Jordan points out.
When you know the prospect is in pain, you don't need to pitch. You just need to demonstrate "tactical empathy." If you can describe their problem better than they can, they will automatically assume you have the solution.
This is the inversion of traditional marketing. You aren't saying "Companies like yours often struggle with X." You are saying "You are experiencing X."
And because you targeted them based on the situation, you are right every time.
Permissionless Value Propositions (PVP): Giving Before Getting
The second concept, Permissionless Value Propositions (PVP), addresses prospects who fit the segment but haven't felt the pain yet.
A PVP is a message "so good that people would pay to receive it." It is independently valuable, regardless of whether they ever buy your product.
This isn't about sending a whitepaper or a Starbucks gift card. It is about delivering high-value intelligence that is specific to their business.
The Anatomy of a PVP
To build a PVP, you combine two types of data:
- Foundational Data (The "Who"): Regulatory filings, patent filings, or tech adoption databases.
- Dynamic Data (The "What's Changing"): New vulnerabilities, regulatory changes, or competitive movements.
Jordan gives the example of regulatory changes. Most mid-market companies don't have the resources to monitor every new compliance requirement.
If you can tell a prospect: "A new regulation was published last week that specifically impacts the software stack you are running. Here is the financial implication for your company, and here is how you can prepare," you have delivered massive value.
You are essentially doing the work of a high-priced consultant for free.
The message structure for PVP is:
- Hook: "Have you seen [Specific Change]?"
- Analysis: "Based on your setup, this means [Concrete Implication] for you."
- Guidance: "Here are your options to handle it..."
Again, there is no product pitch.
The goal is to prove you are an expert in their problem space. When they engage, they aren't doing it to be polite. They are doing it because you have established credibility.
The Real Role of AI: Code, Not Copy
This brings us to the critical misunderstanding of AI in 2025.
Most marketers are using ChatGPT to write the email. They paste in a persona and say, "Write me a witty 100-word email."
Jordan argues this is a waste of the technology. "The purpose of these models isn’t to produce text, it’s to produce code. Or to go search for context."
AI is the engine that makes PQS and PVP scalable.
In the secure file transfer example, Jordan's team didn't use AI to write the emails. They used ChatGPT to find the data sources.
- They asked: "How do I find companies using this specific vulnerable competitor?"
- The AI suggested Shodan, a search engine for connected devices that the marketing team had never heard of.
- They then used AI to write the code to query the Shodan API, cross-reference it with the CVE (Common Vulnerabilities and Exposures) database, and build the list of vulnerable companies.
"The most valuable tokens that ChatGPT, Claude, Gemini and other LLMs produce aren’t about your customer," Jordan says. "They’re the tokens that help you find the data showing your customers are in that situation and structure that data with code."
This is the "Intelligence Architecture" approach. AI handles the data processing and pattern recognition that would overwhelm a human. Humans design the targeting logic and the strategy.
The Orchestration Engine for Situational Intelligence
The strategies outlined in Chapter 5—PQS, PVP, and Situational Intelligence—are undeniably powerful. But they present a massive operational challenge.
How do you actually do this at scale?
How do you monitor the "Who" data and the "What's Changing" data simultaneously? How do you connect the Shodan API to your CRM? How do you trigger the specific PQS message the moment a vulnerability is detected?
If you try to build this with a patchwork of spreadsheets, Zapier zips, and manual SDR research, you will fail. The complexity is too high.
Jordan notes that "The complexity doesn't disappear, it shifts from execution systems to research and design." You need a system that can handle that complexity in the background so your execution remains simple.
This is why we built Momentum.
Momentum is the infrastructure layer that turns the "Cannonball Way" from a theory into a repeatable revenue engine. We are the "Intelligence Architecture" that Jordan advocates for.
We Solve the Data Discovery Problem: Jordan talks about using AI to find and structure data. Momentum’s AI agents can be configured to monitor these specific signals—whether it’s a change in a website’s tech stack, a new regulatory filing, or a hiring surge in a specific department. We capture those signals in real-time.
We Solve the Orchestration Problem: Chapter 5 emphasizes that "The message IS the list." When a specific situation is detected, the message should be automatic. Momentum allows you to build workflows where a "Situation Detected" signal instantly triggers the correct PQS sequence in your engagement platform, updates the CRM with the relevant context, and alerts the account owner—all without human intervention.
We Solve the "PVP" Delivery: Delivering Permissionless Value requires combining multiple data sources to generate a custom insight. Momentum acts as the central brain that pulls data from your "Who" sources and your "What's Changing" sources, synthesizes the insight using enterprise-grade AI, and delivers it to the right channel.
Jordan Crawford’s insight is clear: "You can’t filter your way to a good audience. You need to find a good audience."
Momentum is the tool you use to find them.
Stop trying to optimize your way out of a strategic crisis. Stop asking your AI to write "human" emails with typos. Start building an architecture that detects the truth about your customers' situations.
The insights in this blog post are just a fraction of what is available in the full chapter. To see the specific message templates, the detailed diagrams of the Cannonball framework, and the full breakdown of the Shodan/CVE case study, you need the book.
Get your copy of "Ignite Your GTM With AI" today.