This editorial appeared in the May 15th, 2025, issue of the Topline newsletter.
Want all of the latest go-to-market insights without the wait? Subscribe to Topline and get the full newsletter delivered to your inbox every Thursday.
The promise of AI revolutionizing GTM teams dangles just beyond reach - tantalizing but not quite fully baked. While the box-solution future might be a year or two away, we're living in that delicious in-between: the goldilocks zone where those willing to get their hands dirty can extract massive competitive advantages.
While most are waiting for the perfect AI solution to arrive gift-wrapped, a small cadre of GTM innovators is already rewriting the rules. They're fusing cutting-edge AI tools with battle-tested GTM wisdom to create playbooks that were science fiction just six months ago.
At the bleeding edge of this transformation stands Jordan Crawford - whose work applying AI to GTM challenges has quietly become the template others are scrambling to copy. Our conversation offers a rare window into how the most forward-thinking GTM leaders are creating tomorrow's advantage today.
What follows is our conversation about what's possible now, and how you can position yourself to dominate in the AI-powered GTM future that's arriving faster than most realize. Enjoy!
1. Asad: A few months ago, I started hunting for the sharpest minds at the AI/GTM intersection. Your name rose to the top of every conversation, and after studying your work, it's clear why. What's the backstory here? How did you become the go-to AI GTM guy while others were still trying to turn their prospecting emails into sonnets?
Jordan: I've been doing growth work since 2016 and specifically GTM engineering since 2020, well before AI became mainstream. This gave me a massive advantage because I already had pre-AI frameworks like PQS (Pain Qualified Segment) that I could apply when AI emerged. My only truly original post-AI concept is the PVP (Permissionless Value Prop) - using AI to combine uniquely valuable data into messages so compelling that prospects find them independently valuable.
Also, I spend more time talking with AI than people, constantly experimenting without the constraints of meetings or organizational barriers, which creates compound interest in my learning.
2. Asad: It seems like we've moved beyond the AI wonder phase into something more exciting – figuring out what actually works. While some use cases like code generation have hit clear product-market fit, GTM applications are at that fascinating early-adoption stage where the pioneers are seeing outsized results.
Since GTM strategies vary so wildly, let's break this down motion by motion. First scenario: I'm selling a transactional product into the SMB space – large addressable market, solid buyer data available. What AI approaches are delivering real ROI today that weren't possible six months ago?
Jordan: For transactional SMB sales, AI brings two major advantages. First, AI agents can now hunt leads with greater precision than tools like ZoomInfo or Apollo, which aren't built for vertical SaaS. Second, and more defensively powerful, AI can leverage your internal data to provide unique value to prospects. The real game-changer is using tools like the MCP server to connect to your database, recursively query it, and weaponize those insights to deliver value no competitor can match because they simply don't have access to your unique data.
Let me give you an example: Imagine a company that's been on Housecall Pro for 20 years, an HVAC company. There are two interesting data sets: all of the other customers' work on specific homes, and tools like Shovels (the permit database mentioned earlier). You can blend these two datasets, and the MCP server can talk with both the internal data and the permit data.
You can ask, "Go find the best three opportunities for Jordan's HVAC company to upsell." The system will autonomously run and say, "What do I have access to? I know every other person that's worked on Jordan's customers' homes, and I have public permit data."
It can search all those things and say, "Hey Jordan, thank you for being a Housecall Pro customer for 20 years. We're working to make sure we're always providing value to you. We have three opportunities you should follow up on: three folks that you installed brand new HVAC systems for in 2004, 2006, and 2008. Our records show that no one else has worked on these systems, there have been no additional public permits, and the systems are ending their useful life. We also couldn't determine that these houses have sold, so we think the owners are still there. And by the way, our records show that your owner information is correct because one of our pros did work on that home six weeks ago. You should follow up with them. This represents $568,000 in HVAC replacement."
The fact that it's possible is mind-blowing!
3. Asad: It sounds like vertical B2B SaaS companies have the AI edge in the SMB space. Does this advantage hold true as you move upmarket? And are horizontal SaaS players fighting an uphill battle here?
Jordan: This is my prediction: a lot of horizontal B2B SaaS companies will have their lunch eaten because what used to be a benefit 10-20 years ago, when there weren't a lot of software options and you could start taking over all these use cases, will now become a gigantic weakness.
I think that the more complicated your solution is – if you sell to multiple personas, multiple product lines, multiple problems – you're going to have a real hard time here because these niche providers can spin up apps and tools in a much shorter period, and they'll just eat your lunch.
4. Asad: What about enterprise sales? Is the real breakthrough just having reps armed with unprecedented research depth in a fraction of the time, or are there more creative applications that enterprise teams are missing?
Jordan: For enterprise sales, the key unlock is deep research. In mid-market, it's about filtering the world; in enterprise, it's about going extremely deep with fewer accounts. This requires more data work – analyzing 10Ks, 10Qs, org charts, blog posts, and deploying AI to understand them fully.
While individual reps are currently using tools like OpenAI's Deep Research, the next evolution will be structuring this data at an organizational level. The even bigger opportunity is focusing on where you have information asymmetry – providing prospects with valuable insights they couldn't easily discover themselves.
5. Asad: Now, let's talk about the AI transformation itself. Who actually owns this? Do we need an AI Czar running point across the company? Should we embed AI specialists in each function – like an AI GTM Engineer specifically for GTM Or is this more about creating a culture where every single person is accountable for finding their own AI leverage points?
Jordan: The AI transition requires two parallel approaches. First, everyone needs to be responsible for using AI – start with ChatGPT, Claude, or Gemini for every problem, even if it takes longer. Second, find your top 1% power users – the people already spending tremendous time with AI tools – and make them your AI committee led by an AI czar.
We also need new frameworks for deploying AI in business. Some of this will come bottom-up from people experimenting. I suspect most of it won't come top-down because no one has any clue how to do this. The problem is that AI is a blank box. It's not like it has edges. You can put anything into that box. So we need to do something we haven't done in a long time: ask the question, "So what?" Just throwing AI at problems doesn't fix them. It can be argued that it makes them worse. You don't have observability because these things are non-deterministic – you could run a prompt a hundred times and get a hundred different answers.
So we need frameworks to say, "What is our goal?" That's the question we have to answer. As far as I know, no one's done this thought experiment. We don't even know where we're marching. We have this tool that is amorphous and ridiculously capable today, and we're buying tools trying to optimize yesterday's processes. It doesn't make any sense.
6. Asad: I've noticed a pattern: companies seeing real AI/GTM results all have CROs who are personally immersed in these tools – constantly testing, learning, and pushing boundaries. Can a GTM organization truly transform if their leader is delegating the AI revolution instead of leading it from the front?
Jordan: No, but for a different reason than you might think. In 9 months or sooner, CROs with clean data and unique data assets will have incredible power. They'll be able to connect their systems of record (CRM), information (enriched data), and action (automation tools) through an MCP server as a system of intelligence. This will allow them to deploy multi-channel campaigns simply by describing what they want to do.
If they do that, you're going to have CRO operators who can deploy things, and their teams can be very lean. They might have layers of QA, but they will be absolutely deadly. I can think of maybe five of them off the top of my head, and Kyle Norton is for sure number one.
7. Asad: There's no shortage of buzzy AI/GTM concepts in the pipeline – AI CRMs, autonomous SDRs, and others still in the conceptual phase. What's on the near horizon (12-24 months) that genuinely excites you? And what do you think is over hyped?
Jordan: The AI SDR has a ton of hype but won't work well for most companies. Not because it's a flawed concept, but because AI SDRs try to structure the entire world and sell outcomes – which is incredibly difficult. The more fundamental problem is that they're built on bad data, and many organizations can't even identify their full customer list properly.
What I'm most excited about is the MCP server and recursive intelligence – tools that can connect to all your systems, fix themselves, and continuously improve. I'm looking forward to having conversations with models that can connect to all your systems through a single interface.
8. Asad: Let's close with something practical. If I'm in GTM today and want to thrive, not just survive, in this AI revolution – what do I need to be doing?
Jordan: Just two things will keep you relevant in the AI-driven GTM future.
First, start every single problem with AI – whether it's ChatGPT, Claude, or Gemini – no matter how small. Second, use Charlie Munger's principle: "Invert, always invert." Assume you have perfect information on the entire world and ask what you could do with it to provide 100x outcomes.
Then work backward to build it, using AI to guide you through each step. Don't just type to AI – talk to it using tools like SuperWhisperer.com. Give it multi-paragraph context and ask complex questions, not simple ones.
