Conversation
Highlights

 

How Coactive AI Discovered Their Killer App by Turning Customer Problems Into Product Strategy

Josh Heller joined Coactive AI as Chief Revenue Officer facing a challenge familiar to many infrastructure companies: powerful platform technology with no repeatable path to revenue.

Coactive’s core technology searches and analyzes visual content at scale using natural language—think vector embeddings for image and video data. The applications seemed limitless. Retail companies could use it for product discovery. Social media platforms for content moderation. Media companies for content monetization. They could sell to product teams, data scientists, marketers, content producers.

But serving everyone means serving no one effectively.

Despite backing from Andreessen Horowitz, Bessemer Venture Partners, Greycroft, and Cherry Rock Capital, and despite having co-founder Will—a PhD in neuromorphic computing from Northwestern—building sophisticated technology, revenue remained modest.

Josh’s first 90 days became an exercise in extracting signal from early-stage noise.

Building Opportunity Matrices When Your CRM Tells You Nothing

Most CROs inherit years of pipeline data. Josh had early-stage chaos where the CRM “was good, but it wasn’t really capturing the things that we needed to make decisions.”

His solution: manual deal forensics across every customer interaction. “I made a long kind of opportunity matrix. Every deal that I could think about, a relationship of what was the conversation like, what happened? Why did we win, why did we lose? And started to create some matrix to get a little bit of initial data.”

This wasn’t tracking stage progression or close rates. Josh mapped qualitative patterns: Which buyer personas understood the value proposition immediately? What technical objections surfaced repeatedly? Which industries had budget authority for this category? Where did conversations stall?

The matrix revealed that product and data teams in media and entertainment grasped Coactive’s ROI immediately. As Josh explains: “For them content wasn’t a means to sell your product, it actually is the product. So they have a lot of it, they understand the ROI on it and they really get the concepts that we’re working through.”

This wasn’t about company size or deal value. It was about finding buyers where visual content directly impacts their P&L, not a supporting function.

They abandoned horizontal ambitions to focus exclusively on media and entertainment.

The Platform Company’s Dilemma: Infinite Possibilities, Zero Urgency

Focus on one vertical didn’t solve the fundamental problem platform companies face.

Coactive positioned as infrastructure—”the Databricks for unstructured content,” giving customers “the building blocks for product and data people to build intelligent applications on top of.” Technically accurate. Strategically difficult.

Josh describes the challenge: “The platform sale is long and you have to start to tie it to revenue in a way that is impactful.”

Platform buyers evaluate architectural decisions. They consider technical debt, integration complexity, migration risk. Sales cycles stretch to 9-12 months while they build POCs, run security reviews, negotiate MSAs.

Meanwhile, your startup burns cash.

Coactive needed what Josh calls “a killer app, akin to the Databricks notebook”—the specific application that compresses evaluation time and generates immediate revenue while proving platform capabilities.

The answer emerged from their largest customer.

How a Single Customer Use Case Became Product Strategy

Ryan McConville, NBC Universal’s EVP of Product and Ads, shared a story during a customer conversation that would redirect Coactive’s entire roadmap.

NBC had pursued a major advertiser for years. The advertiser demanded environmentally conscious content placements across Peacock’s video library—scenes with nature, sustainability themes, eco-friendly messaging.

The manual process was prohibitive. “Normally that process would take like four to six weeks,” Josh explains. “It would be back and forth, bunch of data scientists, you’d ultimately do it at the video level and you come back with answer that didn’t map to what was the reality of that content.”

Data scientists would write queries, sample videos, manually review clips, tag content. The output was video-level metadata that couldn’t capture scene-specific context. An hour-long show might contain 30 seconds of relevant content, but the entire video got tagged.

Using Coactive, NBC took a different approach. “They were able to use Coactive and importantly they pulled up our UI, showed them how they, with natural language they could find that content and then iterate on it in real time and they were able to land that deal.”

Natural language search against visual embeddings meant the sales team could demonstrate capability live during the pitch. “Show me outdoor scenes with visible greenery.” “Find shots featuring renewable energy.” The advertiser watched the system iterate in real-time, refining results until they matched brand guidelines exactly.

NBC closed the deal in one meeting instead of losing it to a six-week delay.

The downstream impact validated the approach. NBC launched Contextual Live Targeting as a new product offering. Josh shares the metrics: “They’ve seen 56% improvement in search relative to other streaming platforms. They’ve seen 7x website traffic relative to kind of baselined” from measurement partners.

This became Coactive’s product strategy. “We’re now building an application that sits on top of our core platform that allows direct sales teams to sell contextual bundles in a significantly easier way. All inspired by our biggest customer.”

Not a top-down vision. Customer co-development that revealed the killer app hidden inside platform capabilities.

Hiring Technical Translators Instead of Traditional Sellers

The shift from platform positioning to application focus demanded different seller profiles.

Josh needed people who could whiteboard vector embeddings with data scientists, understand why existing content tagging approaches failed technically, and co-develop solutions during discovery calls. Not traditional enterprise sellers optimized for established categories with clear ROI calculators.

His team composition reflects this. “One guy on the team was a former software engineer and then he got into kind of solution sales and then he became a seller.” Another is “a chemical engineer” from Yale who, during a company hackathon, “submitted an AI agent that customers could come to and get questions answered on our website.”

Others came from management consulting within the industry—people who understood media workflows, content monetization models, and advertising technology stacks.

Josh’s hiring criteria: “At this early stage you have to bring that level of product mindset to be curious about your customer, to not accept said answer, to be willing to solution with them.”

This extends to Coactive’s sales methodology. Josh follows Todd Capone’s “Transparency Sale” philosophy: “Being transparent with your customers and being honest about what’s in the product, setting the right expectations builds an incredible amount of trust.”

The ultimate expression of this approach: “You have to be able to walk away from an opportunity if you don’t think it’s the best thing for your customer.”

Walking away from deals requires either strong pipeline coverage or sellers confident enough to qualify out bad fits. Early-stage companies rarely have the former, so you need the latter.

Why Hiring a CRO Should Increase Founder Sales Time

Josh offers counterintuitive advice about founder involvement after hiring revenue leadership.

When he joined, CEO Cody Coleman anticipated relief from sales responsibilities. Josh’s response: “I kind of joked with our CEO after he hired me. He was saying that now that Josh is here, I don’t have to be on a plane just as much. I have to go to as many meetings. And I said if I’m doing my job, you’re in more meetings, which is true.”

The logic is straightforward. Founders bring advantages that scale sales effectiveness:

• They can make product commitments that close deals
• They have technical depth to solution with senior technical buyers
• Their presence signals strategic importance to enterprise accounts
• They maintain the product-market feedback loop

The CRO’s job isn’t replacing the founder—it’s creating leverage. Systematizing what works. Building repeatable process. Hiring the right profiles. Enabling the founder to focus on the highest-value sales activities instead of everything.

Josh emphasizes this for founders building revenue organizations: “Continue to stay in sales. I think some founders want to step away from it.”

Stepping away breaks the feedback loop between customer conversations and product decisions. The very feedback loop that helped Coactive discover their killer app through NBC’s use case.

The Systematic Path to Product-Market Fit

Coactive’s evolution from horizontal platform to vertical application offers a repeatable framework:

Build forensics into early customer learning. Don’t rely on CRM stage tracking when sample sizes are small. Create qualitative matrices mapping buyer personas, conversation patterns, technical objections, and win/loss reasons across every interaction.

Find customers where your product solves their core P&L problem. Media companies where “content is the product” had budget authority and technical sophistication that retail companies—where visual content supports e-commerce—lacked.

Let your best customer define the killer app. NBC’s specific use case—compressing a 4-6 week manual process into real-time demonstration—became the product roadmap that unlocked an entire category.

Hire sellers who think like product managers. People who can whiteboard technical concepts, understand why current approaches fail, and co-develop solutions in discovery calls.

Keep founders in the sales motion. The CRO creates leverage and process, but founders maintain the customer feedback loop that drives product strategy.

As Josh reflects on the transformation: “Our alignment between go to market and product has come together in a way that’s been magical.”

That alignment came from systematic customer learning, not visionary product strategy. From letting market feedback—specifically, watching how your best customer uses your product—guide what you build next.