Pipes Before the Chocolate
Author(ish): Nathan Thompson
How to architect an organic content factory that powers GTM AI.
Built for sales, not for spend.
My name is Nathan, and I’m currently the Head of Content Strategy at Copy.ai. I’ve had what I can only describe as silly-dumb good fortune in my career. I’ve worked in bootstrapped startups where we had no budget and only big ideas. I’ve worked in luxury content agencies where budgets were big but timelines were even bigger. I’ve been inside VC-backed rockets and helped build from zero to something. Across all of it, I saw the same thing: cracks in the content process.
In 2024, it was my job at Copy.ai to figure out how to operationalize content marketing from the inside out. Not just talk strategy. Not just write copy. But actually build a content system that worked and that could scale with a fast-moving GTM motion. It had to focus on pipeline without draining the team in the process.
And I jumped at the opportunity.
Slowness. Waste. Duplication. And the worst of the bunch... blandness.
I watched TOFU blog posts take weeks (and thousands of dollars) to publish. I watched Thought Leadership articles get distributed without a single actual thought from the “leader.” Ghostwriters hallucinate. Stakeholders disappear or switch jobs.
And content, more often than not, floated into the ether with no clear connection to revenue or reality. So when I stepped into this role, I had a chip on my shoulder. If everyone said AI couldn’t do X, Y, or Z — fine. I’d show them it could.
That was the mindset I brought into 2024.
And yes, it rubbed some people the wrong way. I caught my fair share of scoffing, both online and off. A few folks even suggested I was a soulless marketer trying to automate the art out of the craft.
But in the ever-inspiring words of Kendrick Lamar: “The industry can hate me, f** them all and they mama.”
Because here’s the gods-honest truth underneath the bravado: for the first time in my career, I had a crystal clear vision. And I believed in it. I believed that content marketing — real, strategic, bottom-line-focused content marketing — could be rethought from ground up.
I believed it could be powered by AI, but led by humans; I believed it could actually help buyers move forward, not just help brands feel visible. And the really cool thing? I still know that it's the right vision, all the way down to my core.
That clarity didn’t come out of nowhere, though. My former manager, Zac Harris, laid the foundation in our B2C days. He shaped Copy.ai's SEO and organic growth, and honestly, there’s no way we’d have the visibility we do today without the authority he built.
That said, our shift as a company to B2B in 2024 meant change. The stakes got higher. The audience got higher up the C-suite chain. And the pressure to connect every piece of content to revenue was unrelenting. In other words, s*** got real, real fast.
Then our company got a shot to the arm of marketing leadership in the form of Kyle Coleman. He joined in late 2023, and working with him changed how I approach content. Kyle brought a concrete vision of how marketing should function inside a modern GTM team. Rather than a bunch of sales enablers, marketing should be a driving force of direct pipeline, a way to improve customer experience, and a mine of insights for the product team.
Collaborating with Kyle helped me see just how possible it was to build that system, if we were willing to challenge the way things had always been done and brave enough to stick by it long enough for others to walk along the path, too.
Ah yes, the golden question: how much of the text you're about to read is the baby, and how much of it is the bath water? It's rarely asked so bluntly, but some version of this always comes up.
And I'm happy to answer... just not yet.
Not because I’m dodging the question. Not because there’s anything to hide. But because I believe the process doesn’t matter as much as the outcome, at least not while you're in it. I’ll break it all down at the end of the book. I’ll even give you a link to a forever-free Substack post that walks you through the full system I used — what I did by hand, what AI supported, what I iterated on, and what I threw away.
But as you’re reading this book, here’s something I’d like you to keep in mind: you do not need to know how the donuts were made to know if they taste good.
None of the thinking, the framing, or the strategy was machine-generated. The hard part, the thinking part, was human. The structure, the story, and the stakes are all mine. And if I’ve done my job, you’ll stop wondering how it was made and start thinking about how to build your own.
So read. Annotate. Disagree.
Then stick around for the post-credits scene. I promise it’s worth it.
– Nathan
Let’s start with a hard truth most content teams don’t want to say out loud: the traditional content creation model was never designed to keep pace with today’s go-to-market (GTM) reality. And when it comes to getting anything off the ground, most of us have been silently trained to defer to something known as the Iron Triangle.
The Iron Triangle is a concept borrowed from project management, but it hits content teams square in the inbox. It’s the unavoidable tension between speed, cost, and quality. You can pick two—but never all three.
Most GTM teams spend their days teetering on this triangle like it’s a tightrope. Prioritize one side, and the others wobble.
It’s not just the triangle’s rigidity that causes friction. The traditional content process is weighed down by complexity. More specifically, there are too many people touching each piece of content.
Hear me out, please.
What begins as a crisp idea from a founder or subject matter expert (SME) gets passed from marketing manager to content strategist to copywriter to designer to legal and back again. It’s like a game of telephone, except instead of a funny misheard punchline, you end up with a blog post that doesn't do anything for anybody except a handful of peoples' portfolios.
Worse, every handoff is an opportunity for delay or derailment. But here’s the thing: the issue isn’t that there are too many humans involved. It’s that we’re involving them in the wrong ways.
Let’s take ghostwriting, a seemingly obvious solution to the SME bottleneck. After all, executives are busy. Founders are focused on product. SMEs don’t have time to wordsmith a 1,200-word piece on GTM alignment.
So we bring in ghostwriters to bridge the gap. Makes sense, right?
In theory, yes. In practice, it usually backfires.
Most ghostwriting engagements start with a game of “find the insight.” Writers chase SMEs through Slack threads, old meeting transcripts, and a graveyard of Google Docs labeled “final_v2_revised_FINAL_FINAL.”
The result is an expensive and long-delayed post with fragmented ideas, padded with generic filler. (I don't sound bitter, do I?)
Even if the ghostwriter somehow wrangles something decent, the draft then begins its tour of the organization. Marketing tweaks tone. Product wants technical accuracy. Legal adds disclaimers. Leadership edits by committee.
Meanwhile, great writers (those people who could be crafting compelling narratives or driving strategic messaging) are stuck polishing “per my last email” into LinkedIn-ready connection requests. SMEs, instead of focusing on new insights or customer problems, spend cycles editing line-by-line to make sure they don’t sound like a PR bot.
This is a major misallocation of your most valuable resources. But let’s be crystal clear: the goal isn’t fewer humans. It’s better role design.
What content creation needs is not fewer players but clearer positions. Because when roles are aligned, workflows become faster, voices stay authentic, and content actually does what it’s supposed to do: move the needle.
There was a time not too long ago when marketing teams operated like kids in a candy store. Capital was cheap. Budgets were generous. The name of the game? Growth at all costs.
Need a dozen new hires to test five flavors of lead gen? Go for it. Want to launch a microsite, start a podcast, and host a 20-city roadshow? Absolutely. There was room to experiment, to break things, and to keep sprinting even if the finish line was fuzzy.
That era is over.
Today’s marketing teams are facing a very different climate. Economic headwinds have shifted the mood from indulgence to scrutiny. CFOs are questioning every dollar. Every campaign, tool, and headcount now needs to prove its worth. Success is measured by how directly your work contributes to pipeline and revenue. The phrase “return on investment” used to live in quarterly decks. Now it lives in every conversation.
And with fewer resources to go around, marketing teams can't afford to spread themselves thin or operate in isolation. Which brings us to one of the most persistent and problematic issues in go-to-market organizations.
For years, marketing, sales, product, and customer success operated in separate swim lanes. Marketing focused on awareness and lead generation. Sales picked up leads and closed deals. Product built the roadmap. Customer success tried to keep people happy once they signed.
But today’s buyer does not care about your internal departments. Their journey is not linear. It is not a funnel. It is a maze of touchpoints, revisits, side doors, and search queries. Buyers educate themselves across a mix of channels before they ever talk to a sales rep. Sometimes they never talk to one at all.
This shift has made the old handoff model obsolete and in dire need of a makeover.
Think of the buyer journey like the plumbing in an old house. It looks fine on the surface, but underneath, it is full of leaks and outdated pipes. And when someone turns on the faucet, all those weak spots start to show.
Here is what modern go-to-market teams need to fix:
When these connections are missing, the whole system backs up. Sales cycles drag. Win rates drop. Everyone works harder, but results shrink.
Marketing can no longer be the department of blogs and brand voice alone. It needs to act as the connective tissue between customer insights, product capabilities, and revenue outcomes. That means shifting from solo execution to cross-functional collaboration.
Teams that take the time to embrace this operate with more focus, better handoffs, and a much clearer path to results. Teams that don’t will find themselves stuck in old patterns, burning budget while buyers quietly slip away.
And we’re already seeing it happen in practice.
B2B buyers have changed. The way we go to market has not. Today’s buyers expect more than a sequence of emails and a friendly follow-up call. They want experiences that feel relevant, responsive, and personal. Some want text, some want a call. They want content that understands their world, not just sells into it.
But traditional go-to-market models were not built for that. They were built for a time when buyers moved through predictable stages and marketing handed off leads to sales like a baton in a relay race.
Account-based marketing (ABM) was introduced as the solution. A way to treat each target account like a market of one. A strategy where every message, touchpoint, and asset is personalized to that company’s unique context.
In theory, it is the gold standard for targeting high-value customers. In practice, it is a logistical migraine.
To execute ABM properly, teams need to align on a long list of moving parts:
Multiply that process across a dozen accounts and the cracks start to show. Do it across hundreds and you start sprinting into the Iron Triangle.
You can have speed and personalization, but not without blowing up the budget. You can have scale and savings, but not without cutting corners on quality. The good news is that, for fear of using an overly AI-generated phrase, the game has changed.
Here is what it looks like now:
You are no longer choosing between moving fast and doing it well. With the right systems in place, you can do both.
In 2024, I watched a familiar pattern repeat across marketing teams of every size.
Someone would bring up AI, usually on LinkedIn, and a wall would go up. Eyes would narrow, arms would cross, and someone, inevitably, would say something like:
"AI is just going to churn out soul-less, low-quality content."
"AI slop" as it's now called.
It was the same reaction people had to content farms, to clickbait, to the rise of SEO in the early 2010s. And on the surface, I get it. The fear is that AI reduces creativity to a set of formulas. That it turns marketing into a factory — in the coldest, most industrial sense of the word. But that’s exactly the misconception I want to challenge.
Most people, when they think of automation, default to the Henry Ford model: a grim, grease-covered assembly line cranking out identical Model Ts.
Each unit the same. Rigid. Efficient. Soul-less. Sans soul.
That’s what AI seems like at first glance. But that image misses the magic.
If Henry Ford built a factory for repetition, Willy Wonka built a factory for imagination. His systems were automated, yes. But they weren’t dull. They were fantastical. Enchanted. Each room a different universe of creative intent. Everlasting Gobstoppers. Fizzy Lifting Drinks. Rivers of chocolate running through the floor that sucked spoiled kids up for hours of lesson-learning fun.
Wonka didn’t build a factory that replaced humans. He built a system that channeled human creativity into something bigger than anyone could produce on their own.
But here’s the catch (and this is where most teams stumble).
In Wonka’s factory, the magic looks effortless. But look closer, and you’ll see something else. Systems underneath the spectacle. The chocolate doesn’t flow without the pipes.
You can’t have AI that amplifies creativity unless you’ve done the hard, often boring work of structuring your inputs. You need workflows. You need alignment across teams. You need to decide who owns what, how ideas move through the organization, and what good actually looks like.
In other words, magic requires plumbing.
Too many companies treat AI like a bolt-on tool. Something you add to move faster or cut corners. But that’s missing the opportunity. Used correctly, AI becomes the foundation for consistency, creativity, and cross-functional alignment.
One team might use AI to personalize outbound sales messages at scale. Another might use it to transform transcripts into newsletters, case studies, and LinkedIn posts that actually reflect what was said — not just what fits in a template.
There’s no single right way to build it. Because your factory is different.
This book isn’t about AI as a gimmick. It’s about building a system that works; one that connects marketing, sales, product, and customer success through shared insights and streamlined processes.
And the system you build will depend on the kind of chocolate you want to make.
You might be focused on content velocity. Or account-based marketing. Or turning support calls into revenue-generating assets. Your blueprint will look different than someone else’s. But the foundation will be the same.
Lay the pipes. Turn on the flow. Then let the real magic begin.
Let’s talk about traffic.
For years, top-of-the-funnel (TOFU) content lived and died by one metric: how many people you could get to show up on your site. If you could drive traffic, you were doing it right. More visitors meant more leads. More leads meant more revenue. At least, that was the assumption.
This belief led to a tidal wave of content across the internet — much of it forgettable, a lot of it unhelpful, and some of it… let’s just say surprising.
Take SEMrush, for example.
I’m a fan of the platform. I’ve used it to learn SEO, run audits, and benchmark content strategies. It’s an excellent tool. And when I check their traffic stats, I’m always impressed. As of mid-2025, SEMrush pulls in around 10.5 million monthly visitors. That’s elite territory.
But when you dig a little deeper, things get strange.
A huge portion of that traffic — millions of monthly visits, according to Ahrefs — comes from programmatically generated review pages. And not just any reviews. These are rankings of adult entertainment websites, with titles like “porndude” and “eronity.”
To be clear, I don’t believe SEMrush is targeting that audience intentionally. I’d bet good money that their ideal customer profile doesn’t include someone Googling “incestflix.” But that’s the point.
This content ranks. It drives clicks. It juices domain authority. But it has almost nothing to do with their core business.
So what happened here?
How did we get to a place where one of the most respected SEO tools on the planet has millions of monthly visitors coming from what amounts to traffic candy?
Back in the mid-2010s, programmatic SEO exploded. Companies learned how to spin up thousands of pages based on search volume and simple templates. You didn’t need deep insight, just the right structure and a content database.
Many websites “won” the internet with this strategy. They dominated search, gamed algorithms, and racked up traffic like it was arcade points.
But most companies didn’t follow suit. And not because they were above it, but because they couldn’t afford to.
If you wanted to flood the web with content in 2016, you had to make a choice:
In short, the same Iron Triangle that breaks content today broke it then. The speed-cost-quality tradeoff meant most startups couldn’t scale content without sacrifice. And for TOFU content, that sacrifice usually showed up in one of two ways: bland content that no one read or brilliant content that took three months to ship.
But now, that equation has changed.
Today, we use AI to rewrite the rules of TOFU content. Not by removing humans, but by rethinking their roles.
I like to follow a four-stage model:
In our view, the only red flag with using AI is when you outsource your thinking. But the above system is what I truly consider to be the perfect blend of speed, scale, and soul.
By letting AI handle the first draft of our TOFU content, we eliminate about 80 percent of the slowdown in our TOFU workflow. That doesn’t mean we’re removing writers; it means we’re freeing them up to do their best work: refining the message, improving the strategy, and building alignment across the funnel.
That said, I still do all the keyword research by hand, and humans still handle the content refreshes with tools like Clearscope. Not much has changed on the strategy side of things, except we all execute a lot faster.
There’s no denying that traffic matters. You can’t generate pipeline if no one finds you. But volume without relevance is just noise.
That shift from manual grind to intelligent collaboration is what makes today’s content creation fundamentally different from 2016. And nowhere is that shift more obvious than in how people discover content in the first place. Traditional SEO used to be the gate you had to unlock. Now, AI-powered search engines are rewriting the map entirely.
If you want your top-of-funnel content to actually get seen, it's not enough to publish faster. You also have to structure it for how modern search engines think.
Search behavior is changing. Fast.
We are no longer in a world where Google is the sole gatekeeper of information discovery. New AI-powered search engines like ChatGPT, Perplexity, and DeepSeek are changing how people ask questions and how they expect answers.
This change has big implications for content marketers, especially those focused on top-of-the-funnel strategy. Traditional SEO rewarded a very specific formula: optimize your keywords, build backlinks, and win the domain authority game. If you could check all the boxes, your page climbed the ranks.
But AI search engines play by different rules. They do not care how many backlinks you have if your content does not deliver value. They are not looking for keyword stuffing. They are looking for clarity.
Here is what these tools prioritize instead:
At Copy.ai, we have already started optimizing for these AI-driven search engines. And the results speak for themselves.
We’re continuously seeing more and more qualified demos coming from ChatSearch.
And this came with a lovely note from our SEO consultant, JH from Growth Plays: March was another record breaker for traffic and demos from Gen AI. Demos from ChatGPT were up 42.8% MoM!!
If you want your content to rank in AI search, then I believe that you need to think differently. Here is how we approach it:
If your content is hard for a machine to extract meaning, it won't be shown to a human.
As we have refined our top-of-funnel strategy with automation and AI, our success metrics have shifted. Traffic still matters, but it is no longer the star of the show. What matters more is what traffic leads to.
Here is what we track now:
AI rewards what we should have been doing all along — creating useful, well-structured, human-sounding content that answers real questions. If your content is clear, trustworthy, and easy to parse, you will show up where it counts.
Middle of the funnel content has always had one job. It helps guide prospects from general awareness to serious consideration. That role has not changed. What has changed is the context it lives in.
Buyers today are stepping into the consideration phase with more information than ever. Thanks to AI-powered search tools, conversational chatbots, and a sea of top-of-funnel content, they are already educated by the time they land in the middle of the journey. They do not need another beginner’s guide to content marketing. They need a reason to trust you. They need proof that you understand what is really happening in their world. They need help separating trend from truth.
This is where middle of the funnel content earns its keep. And it is where thought leadership becomes the sharpest tool in the kit.
Thought leadership doesn’t necessarily mean you have all the answers. It means you have a real perspective that’s rooted in experience. It is the moment a brand or an individual says:
"I understand what you are going through. I have been there. Here is what I have learned. And here is how you can move forward with confidence."
By the time prospects reach this part of the funnel, they already know they have a problem. What they want now is direction. That is why they turn to podcasts, panels, webinars, LinkedIn posts, and industry voices they trust. They are looking for experience, to supplement your explanation. They want insight that feels earned, not borrowed.
The best thought leadership content does three things:
When done right, thought leadership builds trust before a salesperson ever enters the chat. It lays the foundation for future conversations and becomes the reference point buyers come back to when they are weighing their options.
Top of the funnel content is often broad and educational. It casts a wide net. Middle of the funnel content narrows the focus. It zooms in on specific changes, new challenges, and emerging best practices.
It requires more than research. It demands lived experience.
This is why generic content falls flat at this stage. Buyers can tell when something was written by someone outside the room. Middle of the funnel content has to come from the inside — from those who have done the work, seen the patterns, and can speak with conviction. That does not mean every piece has to be written by your CEO. But it does mean your content needs to reflect the thinking of the people who actually understand the problem.
Thought leadership should feel personal, specific, and real.
AI can generate drafts. It can summarize transcripts. It can even surface key themes from a messy conversation. But what it cannot do (and should not do) is decide what matters.
That is the job of strategy.
Middle of the funnel content is only as good as the thinking behind it. And the best content strategists know their role is not to write everything themselves. It is to guide the process. To ask better questions. To dig out the insights that make a piece feel sharp, timely, and real.
Here is where strategy still belongs firmly in human hands:
At Copy.ai, we use AI before the content is even created. When preparing for a podcast, webinar, or interview, we run a simple workflow:
And then our human in the loop can look at those questions, our speaker, and determine what is most likely to lead to the best conversation for our customers to hear.
This upfront work ensures that our conversations do not stay surface level. And that means the content we produce afterward — whether it is a blog post, a newsletter, or a social clip — is grounded in something real.
But creating a brilliant piece of thought leadership is only half the job. The other half is making sure the right people actually see it. This is where most companies fall short. They spend time uncovering insights, crafting strong narratives, polishing the voice, and then let that content sit in a shared folder, waiting for someone to promote it. Or worse, they do a single post about it and move on.
Distribution is an ongoing system, and, when done well, it turns one piece of content into many, each tailored for a specific format, audience, or moment in the buyer journey.
In the past, distributing middle of the funnel content was a slog. For every thought leadership post, you had to:
Even if the original idea was strong, the lift required to promote it across all those channels was massive. Which meant most content only got partial distribution at best, and a lot of great insights never reached their full potential.
But with the right workflows in place, a single transcript from a webinar, interview, or fireside chat can be instantly transformed into a full content suite without starting from scratch each time. This also means that companies can codify their brand voice across channels, regardless of the ever revolving door of writers coming and going.
At Copy.ai, we have built this into our operating rhythm. We take one thought leadership conversation and automatically generate:
This approach makes sure that our best thinking does not stay locked in one place. It becomes modular. It travels. It meets people where they are, whether they're scrolling LinkedIn, reading our blog, or watching a video recap. To make middle of the funnel content effective, you have to think like a media company. That means distributing content with the same energy and intentionality that you put into creating it.
AI gives you the leverage to do that without overwhelming your team. It turns one conversation into a campaign. One moment of clarity into many touchpoints. You do not necessarily need to create more content. You may just need to get more out of the content you already have.
Unlike top of the funnel content, which is often measured by page views or impressions, middle of the funnel content should be judged by its ability to influence real decisions. That means tracking what happens after someone reads it.
Did they talk about it?
Share it?
Did it help move a deal forward?
Here are the metrics we pay attention to:
If top of the funnel is about reach, middle of the funnel is about resonance. Being seen isn’t enough; you also need to be remembered, trusted, and shared.
The truth is that generic content will not cut it anymore. The people you want to reach are drowning in noise. What they are looking for is clear content that understands their challenges and points them toward smarter decisions. AI takes the friction out of production. It turns spoken insight into written impact. It removes the bottlenecks that have kept thought leadership locked in calendars and closed-door meetings. But it does not replace the thinking.
And that is the real shift.
Marketing teams are now curators of expertise. They are bridge builders between what buyers are struggling with and what the business can solve.
At the bottom of the funnel, the buyer’s question is no longer "What is this?" — it’s "Is this right for me?"
They already understand their problem. They are actively comparing solutions. They might be weighing budgets, feature sets, integrations, timelines, or internal politics. What they need now is clarity. Something specific and confidence-building.
Bottom of the funnel content should deliver that clarity.
But most companies miss this opportunity. Instead of giving buyers the answers they are actually looking for, they deliver the same old checklists, pitch decks, and product pages.
Informational, maybe. Persuasive, not so much.
The most valuable insights do not come from brainstorms or competitor research. They come from the actual words your prospects say in real conversations.
Every sales call includes:
This is the stuff great content is made of. And sales reps are capturing it daily.
Here is what makes sales transcripts so powerful:
Bottom of the funnel content needs to be personal. It needs to reflect the real-world concerns of the people making buying decisions.
In the past, this level of customization was time-consuming and hard to scale. Marketing teams could create one or two strong assets, but after that, it became a production bottleneck. AI changes that.
With the right workflow, you can now:
At Copy.ai, we use this process to create entire content suites from a single call. For example:
This is not content for content’s sake. It helps the right person say yes faster because they see their own story reflected in what you share.
Before AI entered the picture, personalizing bottom of the funnel content at scale felt nearly impossible. The intent was there, of course. Every marketer wants to create materials that feel tailored to the specific prospect, industry, or deal in play. But the execution required time, resources, and manual effort that most teams simply couldn’t afford.
Creating one-off assets for every account meant chasing notes or rewriting similar content over and over. And it meant constantly pulling bandwidth from other parts of the funnel. It worked in theory (and in theory only).
AI tears out this particular thread of the Iron Curtain. With the right inputs, primarily sales call transcripts, the right AI workflows can generate personalized content that feels handcrafted, without requiring a team of full-time writers to pull it off.
Here is how that plays out in real workflows:
This kind of personalization removes friction and shows that you were actively listening to your audience. Plus, it gives prospects the exact information they need to move forward with confidence.
Many companies still evaluate marketing performance based on top-of-funnel metrics: website traffic, page views, and social shares. But at the bottom of the funnel, those numbers matter much less.
What matters here is movement.
BOFU content should not just be seen. It should help people decide. To evaluate whether your BOFU content is doing its job, ask:
Here are some indicators that let you know your BOFU content is working:
Bottom of the funnel content is not flashy. It does not usually go viral. But it might be the most important content you create. This is the final stretch. The moment when a buyer is trying to connect the dots. When they are weighing your product against competitors, their internal politics, and the risk of change. This is where content should feel like a trusted advisor, not a billboard.
AI allows marketers to deliver that kind of experience to each of their customers, regardless of the audience size. It removes the guesswork and pulls directly from the voice of the customer. Instead of building content in isolation, you’re building it from the actual conversations that move deals forward. And that is the real power. Not just creating content, but capturing what matters most to your buyers and delivering it right when they need it.
There’s a point in every deal where the pitch stops working. The buyer is nodding, but not because they’re convinced, but because they want proof. They’ve heard the features. They have seen the landing page. What they want now is to know if it actually worked for someone like them.
This is where case studies matter.
In a modern go to market funnel, case studies serve as the connective tissue between awareness and decision. They are the content that sales teams send when a buyer is close but still uncertain, and they’re what marketing teams link to when they need to show that a message has teeth. A good case study creates clarity at the moment when someone is ready to buy, but not ready to commit.
It is bottom of the funnel content, but built with the same insight and storytelling that drives the middle of the funnel. The structure is familiar. The stakes are higher. And the impact is real.
We treat case studies like a mix of thought leadership and sales enablement. They’re story-driven, but results-focused. And they follow the same workflow structure we use across most of our deeper content.
Here is how we build ours:
From there, the AI gives us a clean draft. Not something we would publish as-is, but something that gives us 85 - 90 percent of what we need within minutes. But there are two things to note:
One of the most overlooked steps in any AI-assisted workflow is defining what “good” looks like before you ask for more of it.
In our case study process, we always provide the model with a human-written sample we consider excellent. More specifically, our Head of Lifecycle, Jacalyn Beales, wrote the first case study by hand, and it was absolutely stellar. So we had an example that we could refer to, built in-house, and it was something that we all considered excellent. Not just technically correct, but structured well, voice-aligned, and focused on outcomes.
AI can replicate patterns. It can mirror tone. It can apply structure. But it cannot invent standards on its own. If the input it sees is vague, bloated, or generic, your output will be too. We treat that sample as a style guide with a heartbeat. It helps the model learn what matters to us: narrative flow, conversational voice, outcome-first framing, and clear credibility.
This upfront step takes a little more time, but it pays off with cleaner drafts and less back and forth later.
Ok, so the prompting and input matters on the frontend. But what about the backend? The moment case studies start bending the truth, they lose their power. That’s why no case study should ever move forward without a human review. Every needs to be verified against the transcript using a simple search. Every stat should be checked for accuracy and internal consistency. We look for subtle things too, like whether the tone actually reflects how the customer speaks, or whether the story feels artificially neat.
Since moving to GPT-4o and other tuned models, we haven’t had issues with hallucinations, either. But we still check. That step matters more here than anywhere else. Accuracy builds trust. Skipping that step breaks it.
Once the content is structured, tagged, and reviewed, we unlock a toolkit that can be used across the funnel:
What used to be a one-and-done PDF is now a modular asset that supports awareness, consideration, and conversion. And the best part is that this system compounds. With every case study added, the library grows. The stories multiply. And the customer voice gets louder inside the company and outside it.
Webinars, digital events, and podcasts have become essential channels for B2B marketing. These are content engines that can power every stage of the buyer journey.
Live and recorded sessions produce rich transcripts filled with unfiltered insight. With AI in the mix, those transcripts are no longer passive documentation. They become raw material for high-impact content.
A single webinar can drive:
Teams doing this well are already using one event to fuel a month’s worth of content, aligned with real conversations buyers care about. Plus, there has been a noticeable shift in how audiences respond to digital content. Highly produced, over-rehearsed webinars once signaled professionalism. Now, they feel sterile. I believe that people are looking for more raw energy, something they know isn’t an AI avatar.
This shift has opened the door for faster, more frequent content creation. It lowers the barrier to entry and allows teams to focus on substance. With the right systems, every conversation becomes a source of insight that can be repurposed, reshaped, and reused across channels.
Planning webinars used to require weeks of work. You had to coordinate schedules, prepare outlines, draft talking points, and build promotional content. Even the internal meetings to plan the webinars created a pile of follow-ups and manual tasks.
At Copy.ai, we start with single conversation. Our team meets to discuss the topic, speakers, and goals for the month. We do not take notes. We record the meeting and run it through an AI workflow. Within minutes, we have a structured campaign brief that includes:
This turns casual brainstorming into an actionable plan. There is no need to manually transcribe ideas or chase down next steps. The system captures what matters and builds a foundation the team can move on quickly.
Then, everything gets codified on the production side, but more personalized on the audience side. And that makes a difference. Because it’s no secret that webinar performance depends not only on the content but on how well you personalize the experience.
AI allows us to segment audiences based on behavior, role, and engagement level. We can send targeted invitations, personalize reminder emails, and adjust follow-up messaging. Instead of one-size-fits-all follow-up, we can create different messages for:
Every message is aligned with their experience and every touchpoint adds value. This kind of personalization used to require hours of coordination and list management. In other words, it’s what everyone knew they should be doing, but only a handful of companies could actually afford it (and even fewer had people who could manage the number of people it required efficiently enough to execute).
Success does not start or end with registrations. It begins with engagement. Attendance is the clearest signal that your content resonated enough to earn someone’s time. I don’t know about you, but I don’t know many folks in B2B SaaS swimming in free time, so that’s a strong indicator they’re interested in your message. From there, growth is measured not just by volume, but by engagement among the right audience segments.
Here is what we track consistently:
These numbers provide awesome feedback.
If attendance is low, the message did not land. If engagement is flat, the format may need adjusting. Unfortunately, many companies give up too early. They launch a series, get discouraged by early numbers, and scrap the entire program. But those first webinars are where you gather the insights that make the next ones work.
Success comes from sticking with it. Keep publishing. Keep optimizing. Keep showing up.
Webinars don’t have to be standalone efforts that drain time and energy. When paired with smart workflows and structured repurposing, webinars become a reliable engine for full-funnel content creation.
One conversation becomes ten assets. One event fuels outreach, organic traffic, thought leadership, and sales follow-up. And instead of starting from zero every time, you build momentum with every round.
If webinars are planned events, podcasts are pressure valves. They are where real conversations happen. They’re less scripted, more relaxed, and often more insightful. Guests speak freely. Hosts ask what they are actually curious about. The energy is different, and that difference creates opportunities that are hard to fake.
In a podcast, there is room to explore nuance, to dig deeper, to let an idea breathe. That space is where the good stuff comes out. And when captured and repackaged with intention, that good stuff can power every corner of your content strategy.
A single episode can become:
At Copy.ai, we built a workflow that starts long before the guest hits "record" and keeps delivering long after the episode is published.
Here is how it works:
These assets are structured, consistent, and designed to move people through the funnel: from awareness to interest to intent.
This gives us more time having meaningful conversations and less time figuring out how to promote them. We can be creative with our questions, not stuck rewriting the same social caption six ways. And because the process is repeatable, we can scale without burning out.
Podcasting has grown into a channel that builds trust fast. Guests lend credibility. The format invites honesty. And when content feels conversational, audiences stay engaged longer. It’s also one of the few formats where attention spans are expanding, not shrinking. Someone might scroll past a five-second ad. But they will listen to a 35-minute interview on their commute, on a run, or while working through their inbox.
By treating podcasts not as one-off shows but as strategic content inputs, you unlock their full value. Every conversation should be seen as a source of influence that drives brand affinity and pipeline movement.
Whether you are running a webinar series or a podcast, the secret isn't a bigger budget or a flashier format. The secret is consistency, structure, and systems that free your team to focus on what really matters: the message.
Yes, AI will give you the foundation. And yes, it handles the busywork like a champ, standardizes the output masterfully, and makes sure no insight gets lost. But the real value comes from how you show up.
Create a repeatable rhythm. Build workflows that let your team move faster. Design processes that amplify the voices of your guests and leaders. When you do that, your events and interviews become cornerstones of your go to market motion. And just like that, one conversation turns into ten assets. One episode becomes a campaign. And your marketing engine keeps running, not just because you are producing more, but because you are building smarter.
Let me start with a confession: I am not a social media expert. You will not find me writing threads on growth hacks or posting 37-slide carousels on brand archetypes. What I can tell you is how we approach social media as a growing B2B company that cares about being intentional with time, content, and attention. We are not trying to be everywhere. We are trying to show up in the places that matter.
For us, that means LinkedIn, X (formerly Twitter), and YouTube.
We used to be active on Instagram and TikTok during our B2C days. Those platforms brought attention and views, but they weren’t connected to the kinds of conversations that led to pipeline. So we let them go. We’ll revisit them when it makes sense for our stage and strategy, but not a second sooner.
We work with a dedicated social media scheduler. Her challenge is not lack of time. If anything, it's too much efficiency. Once our systems are in place, scheduling becomes the easiest part. The real work is making sure we have enough quality material to fill the calendar with purpose.
That’s why we've built our AI workflows to act as a team of ghostwriting first-drafters. They help us create and maintain a steady drumbeat of content without having to start from a blank page.
Here is how we built our system:
This system could easily scale to include platforms like Reddit or Quora. But just because we can post there does not mean we should.
We are focused on depth and discipline, not reach for its own sake.
One of the most effective parts of our strategy is how we enable our internal team to become part of the social story. We built a workflow that helps everyone on the team share content confidently and consistently.
Here’s how it runs:
This has helped us turn our team into champions for the brand. It removes the hesitation and time pressure that often keep people from posting. More importantly, it builds visibility and momentum without burning out any one person.
Social media does not need to be a time sink or a vanity metric machine. It can be one of the most reliable ways to build presence, reinforce your message, and stay top of mind with your audience.
Our playbook is simple:
When you think about it in this way, AI isn’t so much a crutch for lazy marketers as it is a pair of high-end running shoes for creatives. Stop trying to beat the algorithm and start trying to win attention in the right places, from the right people, and at the right moment.
Account-based marketing used to be like a group project in high school: too many people with varying levels of motivation clunk their way forward until it gets so tiring that someone just throws something together. The idea of ABM seems simple, too. Treat your most important accounts like a market of one. Speak directly to their problems. Show them a solution crafted for their business, not just their industry.
But most teams failed to make it work at scale. Too many moving parts and too much manual effort. Frankly, lots of teams struggle to get the right systems from sales speaking to the right systems in marketing (as a marketer, when’s the last time you logged into Salesforce or Gong…and be honest).
We took a different approach.
Instead of starting with content, we started with workflow. We asked, "What if personalization was not the bottleneck? What if we could build a system where ABM was the default rather than the exception?"
Now, we've built the system and we have our answers. Here's how our ABM process works.
Every ABM sequence starts with account research. But not in the traditional sense of pulling up their website and checking the latest headlines.
We scrape publicly available data about the company with custom built workflows. We look for signals that show what they are focused on, where the pain points are, and how our solution could align with their strategy. This includes:
This information feeds directly into the workflow. The system builds a full profile, and that profile powers everything that follows.
Our product marketing team has already defined core value propositions. These live in our platform, tagged by use case, pain point, and persona.
Once the account research is complete, the system automatically selects the value props that are most relevant to that company. Then, it builds sales angles that connect those value props to the account’s specific priorities.
This does two things:
Now, every message is grounded in what we solve and why it matters without sounding like a generic template. Each account can truly have its unique problems connected to our suite of solutions.
Next, a personalized landing page is created in Webflow. Automatically.
The page includes:
This page is not built by a designer on a tight deadline. It is created through a system that understands the structure of great messaging and can apply it consistently across accounts. It feels personal because it is personal. And it is fast because the heavy lifting is automated. It's worth noting, however, that a human in the loop always verifies the information before anything gets sent to the prospect.
With the account research, messaging, and landing page ready, the system generates outbound emails to the target contact. Each email reflects:
If the first message does not get a response, we’ve got plenty of copy generated for follow-up emails.
If a meeting gets booked, we generate a custom battlecard for the sales rep. This includes:
This makes it easy for sales to walk into the call prepared, confident, and aligned with everything the buyer has already read, seen, and clicked. As one of our reps, Jennifer Junkin said, “I get to read the book instead of writing it every time.”
And this is how ABM becomes scalable. The system connects research, messaging, content creation, and sales enablement into one continuous flow.
It is not magic; it’s just infrastructure. And the more accounts you run through it, the stronger it gets.
Each time we build a page, send an email, or win a deal, the system learns what works. Over time, our messaging sharpens. Our templates improve. And our sales reps walk into every call with more context than most teams get by the end of the deal.
There is a moment when your content starts working harder than your team does. Not because you wrote more of it. But because you built it inside a system that actually knows where everything lives, what it says, and when it should be used.
At Copy.ai, we figured this out by accident. We started creating content in workflows. It was efficient. We could draft fast, stay consistent, and ship more without burning out.
But then we realized something more important. When content lives inside a structured workflow — not just as PDFs or blog posts floating in folders — you unlock a different kind of value. Content stops being static, and it becomes searchable. Dare I say, usable to sales teams?
That’s when you know the system is getting smart.
We use Tables and Infobase to structure the content we create. Every asset — case studies, social posts, one pagers, product narratives — is generated inside a workflow and stored inside a system that tags, organizes, and remembers it. This even includes quotes or research, so we can add or reference that information for future workflows. This does two things right away:
Reusable content means you can grab a quote from a case study, pull a value prop from a product doc, and insert it into an email or a new asset without rewriting it from scratch. Discoverable content means your team can actually find what exists when they need it most, especially in sales conversations.
That’s where the Content Finder workflow came in.
We built a search layer on top of our structured content library, and gave it to the sales team. Now, reps can type in a topic, a persona, or a use case, and immediately pull content that fits. No more Slack threads asking, “Do we have anything on X?”
Now we do. And it shows up when they need it.
Once we had the Content Finder, we realized we could do more.
We attached that workflow to another one. This one took sales call transcripts and wrote personalized follow-up emails (complete with content pulled from the library that aligned with what the buyer actually said).
So now, instead of a generic “Great talking to you!” email, our reps can send:
That connection alone changed how we think about follow-ups. They’re faster to send, easier to personalize, and way more relevant. And the content works harder because it’s showing up where it’s needed, not just sitting on a blog.
Then we took it even further.
Using Zapier and Google Slides, we built a system that automatically generates personalized one pagers (I’m being very liberal with the word “we” on this, we have a killer cofounder and in-house designer that deserve all the credit here). These are based on the same inputs, like sales conversations, CRM fields, and our structured content library.
Each one includes:
Every new page gets sent to a dedicated Slack channel, so our reps always have the latest version ready to go. They don’t have to request it, and they don’t have to write it. They don’t even have to search for the right slide to copy and paste.
Here’s the big takeaway.
None of this worked because we got clever with automation. It worked because we started by laying the pipes. We built our content inside systems that were structured, stored, and searchable. That foundation made every other creative idea possible.
The more your workflows talk to each other, the more context they carry. The more context they carry, the smarter your outputs become. And the smarter your outputs, the more useful your content is for every conversation across your go to market motion.
This is where content becomes infrastructure. And when you build it that way, creativity gets easier, not to mention more fun.
There’s a reason I didn’t start this book by talking about AI-driven chatbots that conduct interviews or workflows that analyze survey responses in real time to generate personalized content. We’re doing those things now (and they’re exciting), but they would have been meaningless and wasted efforts without the foundation underneath them.
Before you can automate creativity, you have to build the systems that creativity can move through. Before you try to turn your marketing engine into something magical, you have to lay the pipes.
And that’s what this book is about.
If you’ve made it through these chapters, you’ve built something. Maybe it's a simple outline with pen and paper. Maybe it’s a workflow that's still duct-taped together. Maybe you're miles ahead with an AI system that's already humming. But now you have something, the early blueprint of a content system that is structured, connected, and able to support your entire go-to-market motion.
You’ve seen how:
Once the pipes are in place, you’ll start to find new ways to use what you’ve built. That’s exactly what happened for us.
We didn’t set out to build a chatbot that could help us collect daily team insights for marketing interviews. But once we had workflows built around transcripts and structured content, it made perfect sense to layer in a chat interface. Now we can get raw insight from our team every single day, no calendar invites needed. Plus, if the thought leader is a higher level executive, a chatbot can be trained to push deeper when a freelancer or junior employee could be too intimidated to do so.
Nor did we start with AI-powered surveys that turn buyer responses into personalized assets. But once we had our value props codified and our messaging mapped, we connected it to a Typeform survey. Now, when someone fills that form out, we can generate content tailored to their goals and stage, automatically.
Those ideas were only possible because we laid the groundwork first. We built systems that could support our thinking. And then we pushed the system further.
AI will not tell you what your brand voice should be. It won’t tell you what your product actually solves. It won’t know why your customers love you, or why others left.
You still need to do that work. You still need to decide what you stand for, how you talk, and who you serve. Strategy still lives in the messy, human part of the job, and thank god for that. It means we still have jobs. It means your work still matters.
The good news is that when your strategy is clear, AI can execute on it faster, better, and more consistently than you could on your own. You might not be building your dream factory yet. That’s okay. You’re not supposed to start there. Start with a system. Break it. Pressure test it. Connect the pipes. Then connect the people. Keep going until the system starts to support you back.
We’re not done either. We’re still building. Still testing. Still figuring out what else we can do with the foundation we laid. But every time we make it easier for the right content to show up at the right time — without burning out the people behind it — the work feels a little more like magic.
So here’s your next steps: Start laying the pipes. Test it with water. When it’s ready, upgrade to chocolate.
Let’s talk about the process: not because it changes the value of the ideas, but because transparency matters. And I know some folks will still wonder: Did AI write this? How much? Can I trust it?
Here’s my answer: Yes, AI helped me write this. But only because I told it what to say.
This book wasn’t written to impress my LinkedIn network with prose. It's not my artistic magnum opus. It’s a practical guide for modern marketers — marketers who see content as a function inside a larger system, and who care about or are responsible for driving growth (not applause).
So here’s how I did it:
I grabbed a pen and wrote down the chapter structure, including what I wanted to say, in what order, and why. I recorded myself talking through each section, out loud, to clarify the point of view and capture the tone. I also loaded in my own past writing — Substack posts, memos, strategy docs — anything that captured how I actually think, speak, and structure ideas. I wasn’t starting from scratch. I was starting from me.
I used a custom workflow to generate the first draft of each chapter. Think of it like a blog-writing process: rough outlines that I could then mold and expand. These were not final. They were scaffolding.
Good enough to work with, not good enough to publish.
Then I built a dedicated chat tool, powered by ChatGPT, and used it as an expander and editor. I read through each section, flagged what needed clarity or depth, and guided the tool on where to go deeper or tighter. It was like having a writing partner that always had energy — but none of the ideas.
There was one rule: I had to guide every expansion. The tool could not lead. It could only follow my prompts, my structure, and my perspective.
Once the expanded draft felt complete, I sat with it. Every word. Every sentence. Line by line, I made manual edits to sharpen voice, rework structure, and fix phrasing. The writing got tighter. The thinking got clearer. But the core stayed the same (i.e. mine).
AI didn’t do the thinking. But AI helped structure my thinking. It pushed back on half-baked logic. It gave me drafts to challenge. It helped me go deeper. If you don’t like the way this book is written, that’s fair. But don’t blame the tool — blame the writer.
The AI didn’t decide what was said.
It didn’t set the tone. It didn’t make the arguments. I did.
Every word here has my fingerprints on it. The point is not that AI wrote this. The point is that I didn’t write it alone. And neither should you. Marketing is changing. Content is changing. The way we work is changing. And we are all learning how to think in systems, not silos. This book is just a reflection of that. It’s not perfect. But it’s useful.
And that was the goal all along.
Pipes Before the Chocolate is a tactical guide for modern marketers who are tired of watching content get buried in Slack threads, forgotten in shared folders, or stuck in endless review cycles. This book won’t teach you how to write the perfect blog post, but it will teach you how to build a content engine that actually supports your go-to-market motion. Inside, you’ll learn how to use AI to scale content without sacrificing strategy, turn sales calls, podcasts, and webinars into full-funnel assets, build a search-ready content library your team actually uses, and more. This isn’t a playbook for chasing traffic; it’s a blueprint for laying the pipes — the foundational systems that make your content smarter, faster, and easier to repurpose across every part of the funnel. Whether you’re a CMO, a Head of Content, or a marketer trying to get through the day without burning out, this book will help you rethink how content fits into your business. And once the pipes are in place? You’ll finally be ready for the chocolate.
