The uncomfortable truth about content marketing: most of it does nothing for pipeline. Not views, not engagement—actual revenue. Seven growth leaders who've scaled companies from HubSpot to Nowports to Databox reveal a pattern that separates the content that converts from the 90% that's pure theater.
The counterintuitive claim at the heart of this thesis: the best content isn't actually content marketing at all. It's co-creation with the market itself.
The Niche Paradox: Small Audiences, Big Checks
Robin Conquet runs a podcast that most marketers would call a failure. DataGen targets data professionals exclusively—a tiny vertical slice of the business world. His show doesn't crack mainstream podcast charts. The audience isn't massive by industry standards.
The business does €350,000 in annual revenue. From a niche podcast.
It's counter-intuitive, but I find it more qualitative to work on content. To be hyper selective, to be very careful, to make quality content that fits well, where in fact the community is happy because you're making good content with clients. That's why I often talk more about co-production than advertising.
— Robin Conquet, Founder at DataGen
Conquet's model flips the typical content playbook. Instead of interruption ads and sponsor reads, he produces roughly 7 episodes monthly—only 2 of which are sponsored. Those sponsored episodes aren't ads at all. They're full interviews with leaders from companies willing to pay for the privilege of being featured, because the audience is so precisely targeted that every listener represents a qualified buyer.
The lesson here cuts against every growth-at-all-costs instinct: smaller, more qualified audiences convert at exponentially higher rates than mass reach. Conquet generates approximately €200,000 from podcast sponsorships alone by obsessively serving a narrow vertical rather than chasing vanity metrics.
Peter Caputa, CEO at Databox, operates from the same playbook—but disagrees on the execution.
People tell me don't put links in your posts, do vertical video instead of horizontal video, don't schedule posts. I ignore all of that stuff. I publish when I feel like it. I write about what I want. If I have a link that's relevant to what I'm talking about, I'm putting it in the post.
— Peter Caputa, CEO at Databox
Where Conquet is strategic about content frequency and placement, Caputa treats content as journaling—organic, unfiltered, driven by what he's genuinely thinking about that day. Both approaches work because they share a foundation: they're writing for a specific type of person, not for an algorithm.
Caputa's LinkedIn posts don't follow a pattern because that's the point. He's building relationships with professional services firms and marketing agencies by engaging them in dialogue, not broadcasting at them. His content becomes the top of a funnel that leads to collaboration, contribution, and conversion.
The split here reveals an important truth: there's no single "right" format. The commonality is specificity of audience and authenticity of voice.
Pull Strategy vs. Push: Why Inbound Still Wins (When Done Right)
Carolina Samsing built HubSpot's Latin American expansion from 3 clients to a viable regional business. Her background included watching HubSpot effectively invent inbound marketing as a category. At Capsule (now Descript's parent company), she applied a pull-over-push philosophy.
At Capsule we always said we had more of a pull strategy than push. HubSpot is a company that coined inbound marketing.
— Carolina Samsing, VP of Growth at Nowports
The distinction matters. Push content interrupts. It's outbound sales disguised as thought leadership. Pull content solves a problem the buyer already has, creating gravity that draws prospects in.
Samsing's experience reveals that pull strategy doesn't mean passive. It means understanding what your market is already searching for and creating definitive resources that answer those questions. At Nowports, she leveraged this philosophy to scale a SaaS business across Latin America by identifying the specific pain points of logistics companies and creating content that addressed them directly.
Kieran Flanagan, CMO at HubSpot (formerly VP of Growth), offers a related but distinct perspective on what makes content actually drive pipeline:
Growth teams are primarily focused on how do we make the go-to-market more touchless? How do we help onboard people? How do we get people to use the product? What I think AI does is extrapolate all that away to just how do we make our go-to-market much more AI-centric?
— Kieran Flanagan, CMO at HubSpot
Flanagan's vision suggests that the traditional division between content that educates and content that converts is collapsing. AI multimodal agents can now guide a prospect from awareness through onboarding through expansion—all in a single conversational flow. The implication: content that moves pipeline in 2025 needs to be structured to feed these systems, not just to be read by humans.
This creates tension with the "authentic journaling" approach Caputa champions. Can you be genuinely organic while also optimizing for AI-assisted buyer journeys? The leaders who crack this will dominate.
Content as Data Infrastructure, Not Just Brand Theater
Peter Caputa's benchmark groups at Databox represent a different species of content entirely. Instead of blog posts or podcasts, Databox built what amounts to the world's largest repository of live performance benchmark data across marketing, sales, and finance functions.
Companies opt into it by connecting their data source. They can connect their Google Analytics account or their Facebook Ads account or their HubSpot CRM account, and then they can instantly see how their performance compares to a group of other companies.
— Peter Caputa, CEO at Databox
This is content as product. Users contribute their data and receive immediate comparative value. Databox runs roughly 50 surveys at any given time with about 100 partners, each designed to let participants benchmark their processes or tool usage against peers.
The content isn't written—it's generated by the market itself. Databox curates, anonymizes, and packages it. The partner ecosystem (primarily marketing agencies) helps recruit respondents in exchange for co-branded insights they can use with their own clients.
Vanessa Schneider, Head of Marketing at Descript, extends this thinking into how AI-native teams should approach content creation:
In terms of expectation setting or incentives, I think it's really important to clarify that the nearest-term objective is not to deliver efficiency. That won't happen right away. Instead, the nearest-term objective is to rewire your brain. It's to change how you see things. It's to be able to abstract workflows into processes that could be automated.
— Vanessa Schneider, Head of Marketing at Descript
Schneider's point: the real shift isn't using AI to create more content faster. It's reconceptualizing what content is and how it's produced. Teams that treat AI as a way to pump out more blog posts miss the point. Teams that use AI to fundamentally rethink content workflows—to turn editorial processes into data infrastructure—win.
Nicolas Rojas, founder of DAPTA, built his company's go-to-market on this exact principle. DAPTA enables non-technical users to build AI agents for sales and go-to-market use cases. Rojas himself is a content machine—streaming on Twitch, producing across TikTok, Instagram, YouTube.
A big part of my strategy and competitive advantage has been that I can generate a lot of interest on social media and channel it into our software products.
— Nicolas Rojas, Founder at DAPTA
Rojas treats content not as marketing expense but as distribution infrastructure. Every piece of content is designed to create inbound interest that flows directly into product trial. The content is the funnel.
Where the Experts Disagree: Format, Frequency, and the Algorithm Question
Peter Caputa and Robin Conquet both run successful content businesses targeting B2B buyers. Their approaches couldn't be more different.
Conquet is strategic to the point of rigidity: exactly 7 episodes per month, only 2 sponsored, strict rules about brand fit and editorial quality. He's building a media property with clear boundaries.
Caputa rejects templates entirely. He posts when inspired. He ignores algorithmic best practices. He treats LinkedIn as a personal journal that happens to be public.
Both work. The difference is in business model. Conquet monetizes the content itself—sponsors pay for placement. Caputa monetizes relationships that begin with content—agencies and service firms become partners and customers.
Justin Kistner, founder of CopyClub.ai and CopySub, stakes out a third position. He started by trying to use AI to create content at scale, quickly realized the quality was terrible, then built a hybrid model:
The real differentiator was that we were leveraging a heavy human-in-the-loop process to solve for the quality problems of AI. But leveraging AI to solve for the scale problems of just human only.
— Justin Kistner, Founder at CopyClub.ai
Kistner's CopySub service grew to $18K MRR in months by using AI to augment human editors, not replace them. The content is AI-assisted but human-directed. It's faster than pure human creation, higher quality than pure AI generation.
When clients asked him to teach them his process, he launched CopyClub.ai—a community teaching people how to use AI in content workflows through "quests" where members build AI agents together.
The disagreement here is fundamental: Is content a product you sell (Conquet), a relationship-building tool (Caputa), or a service you deliver with AI assistance (Kistner)? The answer depends on your business model and who's paying.
The Co-Creation Model: Content With Your Market, Not For It
Every leader interviewed here converges on one insight: the content that moves pipeline is created with the audience, not for them.
Conquet co-produces episodes with sponsors, ensuring the content serves both the audience's need for expertise and the sponsor's need for credibility.
Caputa's benchmark groups require users to contribute their own data to get value—classic co-creation.
Kistner's CopyClub runs collaborative quests where members build together.
Samsing's pull strategy at Nowports meant understanding what logistics companies were already searching for and creating resources they wanted to find.
Even Kieran Flanagan's vision of AI-centric go-to-market is fundamentally about conversation—multimodal agents that guide prospects through their own journey rather than broadcasting at them.
Vanessa Schneider frames the principle clearly:
What all these places have in common is this idea of augmentation or amplification or speed to help people do the thing they already know they want to do that makes them feel more self-actualized.
— Vanessa Schneider, Head of Marketing at Descript
The content that converts doesn't persuade. It enables. It doesn't tell prospects what to think—it helps them accomplish what they're already trying to do.
This is why most B2B content fails. It's designed to generate leads, not to solve problems. It optimizes for MQLs, not for genuine utility. The moment a piece of content feels like it exists to capture an email address, trust evaporates.
What the 10% Actually Have in Common
The content that moves pipeline shares three characteristics, regardless of format or channel:
First, it's specific. Not niche for niche's sake, but built for a precise audience with a defined problem. Conquet's data professionals. Caputa's agency partners. Samsing's logistics companies. Generalist content gets generalist results.
Second, it creates value before asking for value. Databox's benchmarks give immediate utility. Kistner's CopyClub teaches before it sells. Conquet's episodes deliver expertise whether or not the listener becomes a customer. The transaction happens after trust is established, not before.
Third, it treats the audience as collaborators, not targets. This is the throughline. Co-production, co-creation, contribution, conversation. The content that converts is built with the market, not pushed at it.
AI changes the execution but not the principle. Flanagan's multimodal agents, Schneider's workflow automation, Kistner's human-in-the-loop process—all tools to enable collaboration at scale.
The 90% that fails does so because it's still playing the broadcast game. Create content, distribute content, hope someone converts. It's push disguised as pull.
The 10% that works invites the market to build with you. The content becomes the beginning of a relationship, not a campaign asset. Pipeline moves when prospects see themselves in your content because they helped create it.
The shift from content marketing to content co-creation isn't semantic. It's the difference between theater and infrastructure. One performs. The other compounds.