Understory Unfiltered: Octave’s GTM Engine - AI Agents, Contextual Messaging & 10% Reply Rates

TLDR: AI-powered research engines driving high reply rates through contextualization

Zach Vidibor and Julian, co-founders of Octave, built the AI-powered go-to-market coordination platform they wished existed in their prior roles. Zach brings 15 years scaling B2B SaaS at LinkedIn, DocuSign, and Dropbox, while Julian built Gmail's spam protection before founding Smite (sold to Twitter). Together, they help SaaS growth teams achieve 10% reply rates with nearly 100% positive responses by eliminating manual research coordination.

Listen to this episode to learn how Octave centralizes ICP definitions and messaging frameworks, why they just raised $5.5M to build the coordination backbone for SaaS growth operations, and how they're closing the feedback loop between conversations and real-time campaign optimization.

Meet the guests: Zach Vidibor and Julian, Octave's co-founders

Zach Vidibor and Julian are the co-founders of Octave, an AI-powered go-to-market platform that centralizes ICP definitions, messaging frameworks, and campaign execution. Zach spent 15 years scaling B2B SaaS companies including LinkedIn, DocuSign, and Dropbox before meeting Julian. Julian built spam protection systems at Gmail, then founded Smite, which was sold to Twitter.

At Smite, Zach impressed Julian by "shape shifting" their complex technology across vastly different use cases, from credit card fraud detection to disinformation campaigns. Their combined experience with sophisticated technology sales and successful exits drives their thesis that intelligent SaaS growth tools need opinionated workflows, not generic capabilities.

Chapters

[00:21] From Gmail spam protection to Smite's Twitter acquisition: why these founders are building together again

[02:49] The foundational GTM problem: scattered ICP definitions, personas, and messaging across decks and docs

[07:07] Grounded LLMs and avoiding AI bullshit: how Octave's intelligence engine actually works

[11:43] Building research briefs at scale: LinkedIn profiles, company analysis, and prospect intelligence

[15:03] Contextualization beats personalization: why pain-point relevance drives 10% reply rates

[19:38] The origin story: filing away the "better system" idea after Smite's Twitter exit

[25:31] Raising $5.5M and hiring 10 people: building the coordination backbone for SaaS growth

[28:19] Product roadmap: closing the conversation feedback loop for real-time message optimization

Key Insights

Contextualization beats personalization when it comes to prospecting reply rates.

As we discuss in the episode, when Understory used Octave for a cloud cost allocation tool campaign, we achieved 10% reply rates with nearly 100% positive responses, compared to industry average sub-1% reply rates and 40-50% positive ratios for good campaigns. The difference wasn't generic personalization like "I saw you went to this school."

Instead, Octave researched each prospect's role, company challenges, and industry context to deliver messages like "people in FinOps, especially at companies like yours, have been struggling with cloud costs creeping up." "Personalization is not just saying saw you went to this school, this coffee shop near you is pretty good, right? Now do you want to buy my product?" Julian explains. "We're really saying this person has a job and they're trying to be a hero. Like how do we best equip them?" Zach coined the term that resonates with our experience: "What people respond to more is talk to me like other people like me that have a problem. That's what makes people feel comfortable in messaging, you seem to know about others like me."

Your ICP and messaging foundation needs to evolve every few weeks, not annually.

Traditional SaaS growth planning assumes your core message stays stable between quarterly business reviews. Zach points to the new reality: "No longer are we in a world where it's like, hey, sales team, here's the message at kickoff and go see how it works and we'll see you next year, check in at the QBR. There's this velocity of your core message that needs to be picked up."

Product development cycles have accelerated to the point where your solution can "literally turn over every couple of weeks and be something new." As we discuss in the episode, this mirrors what Clay calls finding "GTM alpha," the message that resonates for a specific time period before you need to refine it again. Yet most SaaS companies still store their ICP definitions, persona research, and positioning across scattered decks and docs, making rapid adaptation impossible.

AI-powered research works best when grounded in specific go-to-market context, not generic business knowledge.

Julian emphasizes that large language models "are very good bullshitters" without proper constraints. Octave's approach differs from vanilla ChatGPT because their agents think specifically about "who is this person? What is their career trajectory? What pains do they have in their current role and who is on their team?"

The system runs research on both the individual prospect and their company, then uses classifiers to determine which elements from your message library are relevant for that specific person, persona, and company. As we discuss in the episode, when we tested Octave's outputs, we fact-checked the industry-specific claims it made and consistently found they referenced accurate information from target company websites, demonstrating how grounding prevents generic bullshit and ensures go-to-market nuance.

The most powerful prospecting research happens across data streams your SaaS growth team can't manually synthesize.

Octave's research engine pulls from LinkedIn profiles, company websites, CRM interaction history, email exchanges, and call transcripts to build comprehensive prospect profiles. Julian notes that "most companies are not taking things that their reps are finding in the field and applying those in the actual emails."

When a use case resonates in sales conversations, that insight rarely makes it back to marketing campaigns or website positioning. The platform identifies patterns like "when you talk to hospitality companies, specifically restaurant chains, and you talk about this type of problem with this type of person, that seems to be getting a good response" then immediately applies those learnings to the next email going out.

Strategic LLM selection beats building proprietary models for most SaaS growth tools.

The platform runs on OpenAI, Anthropic, and Gemini models with an LLM-agnostic architecture. "We know we're not going to out-research or outspend any of the main providers," Julian explains. "I really see these LLMs as engines and you need to build a vehicle around it to get you to the workflow." This approach lets them drop in new models as they release and gives customers choice in sophistication levels based on budget constraints.

As we discuss in the episode, we've found similar patterns with custom GPTs and Claude projects, the models themselves are incredibly capable, but the real value comes from fine-tuning workflows and opinions around them. Their competitive advantage lies in go-to-market-specific opinions baked into the workflow and guidance given to these models, combined with access to contextual data, rather than foundational model capabilities.

Most SaaS growth stacks lack the coordination backbone to make intelligent operations possible.

Zach points out that "Salesforce as a system of record doesn't know if you're Gucci shoes or State Farm insurance. It's got a bunch of names and phone numbers and addresses and POs." This creates a fundamental problem as SaaS growth teams adopt more intelligent tools: each system operates in isolation without understanding what you actually sell or who you serve.

Octave positions itself as the coordination backbone that makes every other tool smarter by providing ICP context, persona definitions, and messaging frameworks. "All your tools need a spinal column to make them smarter and know what the hell you do," Zach explains. This vision drove their recent $5.5M funding round and reflects a broader market need for centralized go-to-market intelligence.

Effective prospecting messaging requires understanding career trajectories and promotion motivations, not surface-level personalization.

Traditional personalization focuses on demographic details or recent activities that feel superficial to prospects. Octave's research agents analyze career progression, role responsibilities, and team dynamics to understand what success looks like for each prospect.

"This person has a job and they're trying to get promoted and they're trying to be a hero," Julian notes. "How do we best equip them?" The platform identifies prospects who match your ICP based on actual pain points and solution fit, then crafts messages that position your product as a tool for career success rather than just another vendor pitch.

Real-time feedback loops between sales conversations and campaign optimization will define next-generation SaaS growth platforms.

Octave's product roadmap focuses on closing the loop between field intelligence and campaign execution. Currently, insights from sales calls rarely influence marketing campaigns running simultaneously. Their vision connects conversation intelligence with campaign optimization so learnings from discovery calls immediately improve messaging for prospects in similar segments.

Zach describes the opportunity: "When you talk to hospitality companies, specifically restaurant chains, and you talk about this type of problem with this type of person, that seems to be getting a good response. How do we immediately feed that into the next email that's going out 30 minutes later?" This represents a fundamental shift from quarterly campaign planning to continuous message evolution.

Sophisticated technology products struggle with go-to-market positioning when they serve multiple use cases.

Julian's experience building Smite, a Trust and Safety platform sold to Twitter, taught a crucial lesson about product complexity. "We built this really powerful tool that could do a million different things, but it also made it hard to sell because it could do a million different things." They faced customers who needed solutions for credit card fraud or disinformation campaigns, problems that "might as well exist on two different ends of the earth."

Without a system to capture which use cases resonated with which customer segments, knowledge lived in founders' heads. "There wasn't a place for me to systematize that, or every time we had to hire someone, I had to teach them how to look for those signals," Julian explains. This pain point became the founding thesis for Octave.

Clay integrations transform how agencies deliver client work by enabling scalable, intelligent research workflows.

As we discuss in the episode, Understory connected Octave to Clay via HTTP API integration to analyze prospects at scale for a cloud cost allocation tool client. "We plugged it into Clay via the HTTP API integration, and it analyzed every single row, every individual contact," Alex explains. The workflow researched all prospects, identified relevant messaging for each persona (CTOs, heads of data engineering, CloudOps, FinOps roles), then went multiple layers deeper with career history and company-specific context.

This created a very unique copy for each person that achieved exceptional results. The integration demonstrates how modern agencies can deliver hyper-personalized campaigns at enterprise scale without proportionally increasing manual effort, a coordination advantage that separates sophisticated agencies from traditional service providers.

Want more insight on AI-powered research and contextualized messaging? Listen to the full episode on YouTube and subscribe to Understory's podcast for more insights on SaaS growth strategy.

Looking to eliminate the coordination overhead consuming your growth team's time? Book a call with Understory to explore how coordinated paid media, outbound, and creative can streamline your SaaS growth without specialist management complexity.