Kyle Poyar

About | Kyle Poyar's Growth Unhinged | Kyle Poyar’s Growth Unhinged

Growth Unhinged is a weekly newsletter authored by Kyle Poyar that provides practical, data-driven advice on startup marketing, pricing, and growth strategies. With a subscriber base exceeding 84,000 GTM leaders, founders, and investors, the publication focuses on high-impact areas such as product-led growth (PLG), SaaS benchmarks, and go-to-market (GTM) strategy. The content typically includes tactical deep dives, case studies of fast-growing startups, and applications of AI in GTM plays. Kyle Poyar brings 15 years of experience to the newsletter, having previously served as an Operating Partner at OpenView, where he helped scale revenue for portfolio companies from $1M to over $100M ARR. His background also includes a leadership role at Simon-Kucher & Partners, a premier consulting firm specializing in pricing and packaging. Notably, Poyar and his team at OpenView are credited with coining the term "product-led growth" (PLG), establishing it as a fundamental concept in the tech industry. The newsletter aims to bridge the gap between theoretical growth concepts and actionable execution through data-rich reports and real-world examples.

What's going on in SaaS: The 2025 SaaS Benchmarks Report

The 2025 SaaS Benchmarks report, based on data from over 800 B2B software companies, reveals a period of continued market stabilization characterized by a widening performance gap between AI-native startups and traditional SaaS firms. AI-native companies—defined as those where AI is core to the product or founded post-ChatGPT—are growing three times faster than their peers across all revenue bands. For instance, AI-native startups with $1M-$5M ARR show a median growth of 110% compared to 40% for traditional SaaS. However, this growth comes with a cost: gross margins for AI-centric companies are roughly 5 percentage points lower, and early-stage companies overall have seen gross margins compress by nearly 10 points year-over-year, likely due to AI infrastructure and compute expenses. The research introduces an 'Efficient Growth Matrix' that prioritizes Net Revenue Retention (NRR) and CAC payback periods as the primary predictors of long-term success. Companies in the 'cash cow zone' (high NRR and low CAC payback) average a 71% growth rate and a 47% Rule of 40. Conversely, those in the 'danger zone' (low NRR and high CAC payback) struggle with a mere 10% growth rate. A significant trend in 2025 is the aggressive move toward profitability among later-stage scaleups ($20M-$50M+ ARR). These companies have significantly increased their efficiency, with ARR per full-time employee (FTE) jumping by 42% to 50%, reaching benchmarks of $350k to $400k per employee. While the broader market metrics appear stable, the underlying shift toward AI-driven growth and extreme operational efficiency for mature players marks a fundamental change in the SaaS landscape.

The new UX era

The software industry is entering a new user experience era where the prompt bar has become the primary 'front door' to product value. While this interface offers an illusion of effortlessness, it introduces a new onboarding challenge: the empty prompt bar. Users often struggle to know what to type, making the path to value more fragmented than in traditional UI models. To solve this, activation must shift from a linear 'Setup → Aha → Habit' model to a continuous conversational loop consisting of Prompt, Context, Output, Action, and Habit. In this new framework, the prompt serves as the new signup where users declare intent, and the conversation itself becomes the onboarding process. Analysis of over 40 AI products reveals several critical design patterns for successful prompt-driven UX. First, products must anchor the experience in real use cases rather than generic 'ask me anything' boxes; Canva, for example, uses specific task-oriented suggestions like 'Design a social post.' Second, grouping prompts by intent—such as create, explore, or analyze—helps users build mental models of the tool's capabilities. Third, the distinction between the AI's role as a builder, assistant, or co-pilot must be crystal clear through UI cues. Fourth, the most effective experiences integrate context and data sources directly into the flow, as seen in Notion’s connector highlights. Finally, momentum is maintained by recommending clear next steps immediately after the AI generates an output, bridging the gap between a 'wow' moment and actual utility. Ultimately, differentiation in this era is not about the prompt bar itself, but about how effectively a product turns that initial interaction into repeatable, habit-forming value.

AI agents for marketing: inside a 40-agent stack

The document explores the implementation of an AI-agent-first marketing organization, specifically detailing how Jacob Bank, CEO of Relay.app, utilizes a stack of over 40 AI agents to perform the work of a five-person marketing team. This shift reflects a broader trend where major companies like Shopify, Amazon, and Klarna are making AI proficiency a baseline expectation for employees. Unlike traditional automation tools that follow simple 'if this, then that' logic, AI agents are described as software systems capable of complex, multi-step processes involving reasoning, memory, and autonomous decision-making. The marketing organization is structured into six core functions: social media, blog and website management, email marketing, lead qualification, community engagement, and partner programs. The strategy is heavily weighted toward social media and community, where agents handle tasks ranging from content research and creation to tracking and follow-ups. A notable concept introduced is 'spiky intelligence,' which suggests that while AI agents may struggle with simple tasks, they can perform at a PhD level in specific areas like coaching CEOs on sales calls, analyzing massive reports, or critiquing marketing strategies. Specific examples of agents in use include a LinkedIn engagement coach that analyzes post performance to provide strategic advice and a LinkedIn content research agent that monitors high-profile influencers to identify trending themes. These agents often include a 'human-in-the-loop' component, allowing managers to review outputs before they are published. The document emphasizes that building a personal AI agent stack is becoming a portable, foundational skill that individuals can carry throughout their careers.

How to get recommended by ChatGPT

AI search engines like ChatGPT, Perplexity, and Google Gemini are becoming primary gatekeepers in the B2B buying journey, necessitating a shift from traditional Search Engine Optimization (SEO) to Answer Engine Optimization (AEO). With ChatGPT reaching one billion users and generating 2.5 billion daily prompts, being mentioned in AI-generated answers is critical for brand visibility. Unlike traditional search, the goal is often influence rather than direct clicks, as many AI interactions result in "zero-click" searches where the user receives information without visiting a website. Research from Josh Blyskal at Profound indicates that different AI engines behave uniquely; for instance, ChatGPT and Perplexity share only an 11% overlap in domain citations. To succeed, brands must identify specific prompts to "own," categorized into core must-wins, competitive "knife-fights," and experimental opportunities. Content strategy should prioritize hyper-specificity and granularity, focusing on niche differentiators rather than broad topics. A "surround sound" approach is essential, leveraging high-authority domains that AI engines trust. Reddit has emerged as the most-cited domain for both ChatGPT and Perplexity, making presence in relevant subreddits a top priority. Furthermore, AI engines prefer structured data: one-in-three citations come from comparative listicles, and structured tables are significantly more effective than images or charts, which AI models currently struggle to interpret accurately. Measurement must also evolve. Since referral traffic may not capture the full impact of AI recommendations, brands should focus on share-of-voice and manual attribution methods, such as "How did you hear about us?" lead form fields. Finally, marketers must account for "citation drift," where nearly half of cited domains can change within a single month, requiring frequent analytics refreshes and adaptive content strategies to maintain visibility as LLMs evolve.

Your guide to Webflow’s $4B PLG engine

Webflow reached a $4 billion valuation and $100 million in annual recurring revenue (ARR) by effectively combining a product-led growth (PLG) engine with a strategic sales-assist motion. Founded in 2012, the company bootstrapped to $10 million in ARR over seven years before raising significant capital. Its early success relied heavily on organic growth, specifically through community building and SEO. By launching a Discourse forum just before their beta release, the founders fostered a space where freelance designers and developers could support one another. This community-generated content significantly boosted long-tail SEO for technical queries like "how do I build a modal?" The growth strategy evolved with the introduction of the Showcase product in 2014, which allowed users to share cloneable assets. This created a powerful flywheel where creators gained credibility while providing Webflow with over 100,000 assets that further drove organic traffic. As the company matured, they utilized data enrichment tools like Clearbit and Segment to identify high-value users within large organizations. Discovering that employees from major enterprises were signing up daily, Webflow experimented with a sales-assist model. This involved proactive outreach to "hand-raisers"—users in large accounts who needed help with security, procurement, or complex implementations—eventually leading to their first enterprise contracts and a dedicated sales-assist team of over 40 people.

Inside Replit's path to $100M ARR

Replit, led by founder Amjad Masad, achieved a massive financial milestone by reaching $100M ARR in June 2025, a tenfold increase from $10M at the end of 2024. This explosive growth is attributed to the launch of its AI agent in late 2024, which has facilitated the creation of over two million applications. The platform has successfully democratized software development, allowing non-technical users—such as designers, product managers, and executives—to build and deploy production-ready apps through natural language prompts, a trend often referred to as "vibecoding." Replit is currently the largest consumer of Anthropic models by tokens on Google Cloud. The document details Replit's UX philosophy of "progressive disclosure," which aims to provide a simple natural language interface for beginners while keeping advanced tools like code editors and OS access just clicks away for power users. This approach avoids the common pitfall of overwhelming users with complexity or oversimplifying to the point of constraint. Masad emphasizes that rapid user activation is critical; the platform aims for a "wow" moment within minutes of a first visit. This is achieved through guided experiences, such as one-click "habit tracker" templates, followed by higher-level activation when users input their own unique prompts. Despite its success, Replit's journey included a difficult three-year bootstrapping period and multiple rejections from Y Combinator before finding its current product-market fit in the AI era.

Europe's fastest growing startup?

Lovable, a Swedish AI coding startup, has achieved unprecedented growth, reaching $75M in Annual Recurring Revenue (ARR) just eight months after its launch. The company's trajectory began with a rapid ascent from $0 to $10M ARR in 60 days, hitting $30M within four months. This growth was achieved with extreme capital efficiency, burning only $2M to reach the $30M milestone while maintaining a lean team of 18 people, which later expanded to 45. This translates to over $1M ARR per employee, a metric practically unheard of in the startup ecosystem. In July 2025, Lovable announced a $200M Series A funding round at a $1.8 billion valuation, officially reaching unicorn status. The startup's success is rooted in its predecessor, an open-source project called gpt-engineer created by co-founder and CEO Anton Osika in June 2023. While gpt-engineer gained significant traction on GitHub with over 50,000 stars and hundreds of thousands of users, the commercial transition required several iterations. The team spent 18 months building and refining the product before achieving breakout success. Initially, a dedicated GPT Engineer app launched in December 2023 without much fanfare, followed by a better version in August 2024 that saw modest success before growth flatlined. The turning point occurred in late November 2024 with the rebrand to Lovable. The team focused on solving critical technical hurdles, such as preventing the AI from getting "stuck" and ensuring high performance on large codebases, which unlocked "real" use cases and greater reliability. The platform aims to democratize app building, allowing individuals without coding backgrounds to turn ideas into functional software. This shift from a developer-centric tool to a broad-market app builder was central to their scale. The company's ability to maintain a small headcount while scaling revenue highlights a new era of AI-native business models where human capital is leveraged through high-efficiency automation and viral word-of-mouth growth.

The evolution of Miro's user onboarding

Miro’s journey from a 50-person startup to a global leader with 50 million users highlights a strategic shift from high-concept "big bets" to a culture of "smart iterations." In the early startup phase, the team focused on a visually polished, role-segmented onboarding flow. Despite positive qualitative feedback, the data showed negative results: the interactive "beautifications" distracted users from core tasks like sending invites or using templates. This taught the team that while data reveals what is happening, only user interviews explain why. During the pandemic-driven hyper-growth phase, Miro faced a less tech-savvy audience needing immediate value. They invested heavily in a "robo-collaboration" experience—a human-led video walkthrough using Synthesia AI. While it increased content creation, it failed to improve the "Aha moment" because many users skipped the tutorial or found it irrelevant. This period led to a more rigorous definition of activation through three stages: the Setup moment (initial contact), the Aha moment (first experience of core value), and the Habit moment (repeated core actions over time). In the current growth-at-scale phase, Miro pivoted to "smart iterations" focusing on specific user segments like "Joiners." By identifying that "Miro reactions" provided a low-friction way to "break the ice," the team introduced a "Say Hi" prompt. The first version failed, but through post-analysis, they iterated on discoverability and cognitive load—darkening the tooltip background and triggering reactions directly from the button. These small, evidence-based adjustments finally moved the needle on the Aha moment. The core takeaway for product teams is to decompose big investments into smaller tests, always run a second iteration, and use "quiz workshops" to build collective product sense and unpack the behavioral "why" behind every experiment result.

Surprising lessons from 20,000 experiments

Steal these brilliant growth tactics

Your guide to product-led onboarding

Your guide to SaaS packaging 201

SaaS packaging involves assembling features and services into specific editions or SKUs that align with customer needs and willingness to pay. The foundational framework for most successful SaaS companies, including Salesforce, Figma, and Slack, is the 'Good-Better-Best' (GBB) model. This tiered approach balances the simplicity of a one-size-fits-all model with the flexibility of custom builds, effectively avoiding decision paralysis while providing a clear path for account expansion and upselling. GBB allows companies to differentiate pricing based on customer sophistication and helps fund ongoing R&D by creating a mechanism to monetize new developments. Beyond the core tiers, selective add-ons can be used to increase deal size and provide sales flexibility. Features are best suited as add-ons if they are 'polarizing'—highly valued by a subset of users but unnecessary for the average customer—or if they appeal to a different buyer's budget. Add-ons are also effective for features that compete with third-party vendors or those that involve significant operational costs. However, to maintain a smooth sales process, the average customer should not need more than one or two add-ons during their initial purchase. Successful packaging requires each tier to have a defined 'job' or North Star. Stakeholders must agree on who each package is for and what role it plays in the overall revenue strategy. This alignment is measured through KPIs such as adoption rates, customer success metrics, and upgrade velocity. Finally, packaging should be leveraged as a competitive weapon. Rather than mimicking the market, companies should structure their tiers to highlight unique value propositions and monetize capabilities that competitors cannot match. A key pitfall to avoid is gating features that are essential for driving user engagement and collaboration, as these are the primary engines for long-term product growth.

Reverse trials: The PLG strategy that combines freemium and free trial

The reverse trial model represents a strategic synthesis of freemium and free trial approaches, designed to optimize both user acquisition and revenue conversion. In this model, new users are automatically granted a time-limited trial of premium features—typically 14 days—at the start of their journey. Upon expiration, instead of losing access to the product entirely as in a standard free trial, users are downgraded to a permanent free tier. This approach leverages behavioral psychology, specifically loss aversion, where the threat of losing access to advanced features acts as a more powerful motivator for conversion than the promise of gaining them. Airtable, valued at $11 billion, serves as a primary case study for this strategy, using it to balance their dual 'North Star' metrics of user growth and self-serve revenue. Lauryn Isford, Airtable's Head of Growth, emphasizes that while standard free trials create urgency and higher conversion rates (often 2-3x higher than freemium), they risk losing users who haven't reached 'time-to-value' within the trial window. Conversely, freemium models excel at top-of-funnel acquisition and viral loops but suffer from slower cash payback and lower conversion urgency. The reverse trial mitigates these trade-offs by keeping the door open for long-term nurturing. Successful implementation relies on robust packaging, where features are categorized into 'moats' (high switching costs), 'sticky' features (usage-based value), and 'sophistication' (complex needs). A healthy plan typically follows an 80:20 rule, where 80% of users remain on the free tier while 20% find the premium features compelling enough to upgrade. For emerging startups, the reverse trial offers a way to build trust over months or years, ensuring that a single conversion conversation doesn't end the customer relationship prematurely.

Why everyone’s switching to AI credits

The state of B2B monetization in 2025

The B2B software landscape is undergoing a fundamental shift in monetization strategy, moving away from traditional seat-based and flat-rate subscriptions toward hybrid models that combine subscriptions with usage-based components. Data from 240 software and AI companies reveals that flat-fee subscriptions have dropped from 29% to 22% in the last year, while hybrid pricing has surged from 27% to 41%. This transition is largely driven by the integration of AI, which decouples value from headcount. As AI agents perform more work, the traditional 'per-seat' metric becomes obsolete because customers may actually require fewer human users to achieve higher output. Companies like Klarna and Microsoft are already seeing massive increases in revenue per employee due to AI efficiency, forcing vendors to reconsider how they capture value. Hybrid pricing, exemplified by companies like Clay, uses a combination of subscription tiers for features and credits for usage. This approach offers a natural upsell path, protects vendor margins against the high compute costs of AI, and provides buyers with a level of predictability. Beyond hybrid models, outcome-based pricing is viewed as the ultimate goal, with 25% of companies expecting to adopt it by 2028. However, successful implementation requires the CAMP framework: Consistency in value, Attribution of results to the product, Measurability of outcomes in real-time, and Predictability of performance across different customers. Currently, only 5% of the market has achieved this level of maturity. Despite the trend toward digital self-service, pricing transparency is actually stalling. As models become more complex with credits, overages, and platform fees, buyers increasingly prefer speaking with sales representatives to navigate the nuances. Furthermore, many organizations face a 'pricing no-man's land' between $5M and $20M ARR, where ownership of monetization strategy often falls through the cracks between sales, product, and finance. To remain competitive, companies must treat pricing as a strategic, resourced function rather than a one-time decision.

Your pricing is (probably) broken: Here's how to fix it

Software founders often face anxiety regarding unoptimized pricing models, particularly as AI shifts market expectations. This analysis categorizes common pricing structures—flat-fee, feature-based, and seat-based—identifying their inherent flaws and providing tactical remedies for each. Flat-fee subscriptions, while predictable for customers, often lead to missed revenue from large enterprises and negative margins from power users. To counter this, companies should introduce premium editions (typically 50-100% higher in price), charge for niche add-ons, implement annual price escalator clauses of 5-8%, and establish fair usage policies for the top 10% of users who often drive the majority of resource consumption. Feature-based pricing, or 'Good-Better-Best' packaging, frequently suffers from 'bundle hell' where users find plans arbitrary. Solutions include distilling each package's value into a single sentence, unbundling new features before re-bundling them every 18 months, and utilizing progressive feature gating to allow users to sample premium tools. Seat-based pricing is increasingly challenged by AI's ability to reduce headcount, making human-count metrics less reflective of value. Instead of a total pivot, companies can redefine the model by introducing 'lite user' seats at 10-40% of the standard price, setting minimum user counts for specific plans, or layering in usage paywalls. Ultimately, no pricing model is perfect, but iterative improvements and the addition of expansion levers are essential for sustainable growth in a landscape where AI agents are decoupling value from human labor.

From selling access to selling work (and what it means for you)

The software industry is undergoing a fundamental shift in its economic foundation, moving away from selling access via seat-based subscriptions toward selling the actual work delivered by AI agents. Traditional Annual Recurring Revenue (ARR) metrics, which rely on predictable, high-margin seat licenses, are being challenged by AI-driven models that charge based on specific outputs. These include charging per successful support ticket resolution, per photo edited, or per task completed by an autonomous agent. This transition is necessary because the value provided by AI and automation is often disconnected from the number of human users logging into a platform. Prominent examples of this disruptive pricing include Salesforce’s Agentforce, which charges $2 per conversation, and Intercom’s Fin AI agent, which utilizes a $0.99 per resolution model. Zendesk has followed suit with similar outcome-based pricing, offering both pay-as-you-go and upfront commitment options. This trend extends across categories: legal AI might charge for generated demand packages, while creator tools charge for video generation volume. Many of these systems utilize AI credits to differentiate between basic and advanced workflows. This selling work model is a sophisticated evolution of usage-based pricing. By bundling software with the service it performs, vendors can lower the total cost of ownership (TCO) for customers while simultaneously increasing their own pricing power. Historically, software pricing evolved from on-premise models to subscriptions. While traditional SaaS products typically capture only 10-15% of the economic value they deliver, success-based or output-based models can capture 20-30% because the correlation between the product's cost and the customer's result is much tighter. This shift represents a move toward capturing a larger share of the value created by replacing or augmenting human labor with AI-driven outputs.

The definitive SaaS homepage framework

Your guide to GTM metrics 2.0: What replaces your broken MQL funnel

You should be vibe coding for GTM - by Alex Shartsis

Vibecoding represents a shift in Go-To-Market (GTM) strategy where non-technical leaders use AI-powered coding agents to build custom software tools from scratch. Unlike standard GTM automation, which relies on 'if-this-then-that' logic through platforms like Clay, vibecoding utilizes tools such as Bolt.new, Lovable, Cursor, and Replit to generate functional web products, Python scrapers, and React applications. These tools handle the entire development lifecycle, including environment setup and cloud hosting, allowing users to move from a concept to a live URL in a matter of hours without prior coding knowledge. A primary application of this approach is the creation of high-fidelity sales tools, illustrated by a case study involving Pico MES, a manufacturing SaaS company. By converting a complex, static Excel ROI spreadsheet into an interactive React application using Bolt.new, the company transformed a friction-filled sales step into a strategic conversation starter. The resulting tool reframed the product's value around executive-level metrics like EBITDA multiples and exit valuations rather than just operational savings. This interactive format not only captured lead data and sent Slack alerts but also encouraged prospects to be more forthcoming with their data during live sales calls. To successfully execute a vibecoding project, users should provide AI agents with visual context, such as screenshots of existing spreadsheets or brand guidelines, and offer explicit instructions regarding the target audience and desired user flow. Starting with a focused Minimum Viable Product (MVP) is recommended over attempting to build complex features like CSV exports in the first prompt. This methodology provides a significant speed and cost advantage, enabling GTM teams to bypass engineering backlogs or expensive agency fees to deploy professional-grade lead magnets and sales enablement assets.

Kyle Poyar’s Growth Unhinged | Substack

Growth Unhinged is a specialized publication authored by Kyle Poyar that focuses on the tactical and strategic frameworks utilized by high-growth startups to achieve market dominance. With a substantial subscriber base of over 85,000 readers, the newsletter positions itself as a primary source for "real-life growth advice," moving beyond high-level theory to provide actionable playbooks and in-depth case studies. The core curriculum of the publication centers on three critical pillars of modern software business: product-led growth (PLG), pricing strategy, and go-to-market (GTM) tactics. These elements are explored through the lens of successful companies, revealing the "hidden tactics" that often go unnoticed by casual market observers. The content is highly regarded by industry experts, such as Paweł Huryn of The Product Compass, for its ability to deconstruct the complex strategies of rapidly expanding ventures. By focusing on PLG, the newsletter addresses how companies can leverage their own software to drive acquisition and retention, a trend that is particularly relevant in competitive sectors like Educational Technology and SaaS. Furthermore, the emphasis on pricing strategies provides readers with the tools to evaluate value-based versus cost-based models, which is essential for sustainable scaling. The go-to-market insights offer a roadmap for navigating the initial stages of product launch and subsequent market expansion. For researchers and professionals, Growth Unhinged serves as a repository of competitive intelligence, offering a clear view of the mechanics behind the success of today’s most prominent startups. The publication’s commitment to providing these insights for free makes it an accessible yet high-value resource for anyone looking to understand the current landscape of business growth and digital transformation.

Who's actually hiring in GTM right now? AI companies are doubling SDR hiring

The H1 2026 State of GTM Hiring report, based on real-time data from Sumble, reveals a complex landscape where overall Go-To-Market (GTM) job posts have declined by 15% year-over-year. Despite this broader market pullback, AI-native companies like OpenAI, Cursor, and Decagon are bucking the trend by doubling their Sales Development Representative (SDR) headcount. This paradox suggests that while these companies sell automation, they still rely heavily on human-led outbound efforts to scale. Conversely, customer support roles have seen a significant 37% decline in job posts, with AI-native firms maintaining support teams that are 67% smaller than traditional B2B digital natives, indicating successful automation of transactional service tasks. Sales and solution engineering (SE) roles remain the most resilient, accounting for three out of every five open GTM positions. This stability highlights a persistent buyer preference for human interaction in complex enterprise deals. Within marketing, growth marketing is the only sub-category showing growth, up 8% year-over-year, while roles like product marketing and events marketing have seen double-digit declines. A significant emerging trend is the rise of GTM engineering, a role that has doubled in headcount over the past year. These specialists, often using tools like Clay and Claude Code, focus on data hygiene and automated prospecting systems. The data shows that marketing departments are the most aggressive adopters of AI tools like Claude, second only to GTM engineering. Ultimately, the report suggests a shift toward smaller, more specialized, and systems-oriented GTM teams rather than a total collapse of the hiring market.

How to build your GTM strategy from scratch

2026 State of AI for B2B GTM report

Go-to-market (GTM) leaders are increasingly divided into two camps: those seeing little impact from AI and a small group achieving outsized returns in pipeline generation, conversion rates, and operational efficiency. This 2026 report identifies 40 high-impact AI use cases across four primary categories: content creation, growth and product marketing, prospecting, and sales engagement. Most high-value plays leverage general-purpose LLMs like ChatGPT, Claude, and Gemini alongside affordable, off-the-shelf tools like Clay, Webflow, HubSpot, and Apollo. In content creation, the focus has shifted from simple generation to context-heavy assistants that utilize internal messaging, brand guidelines, and knowledge bases to produce higher-quality LinkedIn posts. Advanced workflows now include the automated creation of hundreds of competitor comparison pages using Clay and Claude to scrape data and generate SEO-optimized headlines. Product marketing teams are deploying "BattleBots" to analyze Gong transcripts and provide real-time competitive strategy to sales reps. Others use "digital twins" trained on real customer data—such as sales call transcripts and G2 reviews—to simulate customer reactions and objections to marketing campaigns before they launch. For prospecting, AI is being used to build binary scoring models for Serviceable Addressable Market (SAM) identification, combining firmographic data with "vibes-based" AI research. Teams are also automating hyper-personalized ABM campaigns that scan prospect websites to generate custom "week in review" emails, significantly increasing meeting bookings without additional headcount. Sales engagement workflows include using custom GPTs for rapid meeting preparation by synthesizing transcripts, email threads, and LinkedIn data in under two minutes. Additionally, automated re-engagement for "closed-lost" deals uses AI call recorders like Sybill to fill CRM fields and trigger personalized content sequences via Tofu. The report emphasizes that 90% of AI's value can be captured without specialized engineering skills by connecting internal context directly to AI assistants.

How to use Deep Research for GTM

This guide explores the application of 'Deep Research' AI modes—specifically within tools like ChatGPT, Perplexity, and Gemini—to Go-To-Market (GTM) projects. While often perceived as academic tools, these agents can condense 10+ hours of manual research into minutes by executing complex, multi-step tasks end-to-end, from planning to final delivery. The effectiveness of these tools depends heavily on active 'handholding' through specific prompting techniques. Key strategies include directing the agent toward high-quality primary sources (such as government data) over social media, requiring in-text citations, and mandating source tables to ensure transparency and accuracy. Furthermore, the quality of the research is tied to the context provided; users should detail their company's operations, specific goals (e.g., budget allocation visibility), and constraints (e.g., legal or budgetary limits). To avoid wasted time or research credits, users are encouraged to ask the AI for a 'research plan' before it begins the full report, allowing for methodology adjustments. The article also suggests using advanced models like GPT-5 or Claude Opus to pre-generate source lists or brainstorm the necessary context before initiating a Deep Research run. Finally, formatting instructions—such as requesting summaries for every section and using tables instead of text blocks—are recommended to make the final deliverables more digestible for stakeholders.

What’s working in GTM right now

This report, based on a survey of 195 software GTM leaders conducted by Kyle Poyar and Maja Voje, outlines the current landscape and future trajectory of B2B go-to-market strategies. The data reveals six primary GTM motions, with Inbound (23%), Outbound (19%), and Account-Based GTM (18%) leading the way. Product-Led Growth (PLG) remains dominant for lower-priced products (under $5,000 ACV), while Account-Based motions are preferred for high-value deals exceeding $25,000. Interestingly, there is no direct correlation between a specific motion and growth rate, suggesting that execution within a chosen channel is the primary driver of success. In 2025, companies engaged in significant channel exploration, averaging five core channels and 5.5 active experiments. LinkedIn (66%), SEO (53%), and warm outbound (48%) are the most common core channels. However, investment is shifting toward emerging areas like AI Search (AEO), where 51% of respondents plan to increase spending, and intent-based outbound (45%). Conversely, traditional SEO is seeing a decline in relative investment interest. The report identifies 2026 as the 'year of ruthless scaling,' where leaders must transition from broad experimentation to doubling down on high-intent pipeline generators. Regarding artificial intelligence, adoption is nearly universal, with 91% of leaders using general-purpose tools like ChatGPT. Despite this, 53% of respondents report limited or no real impact on GTM outcomes. Disillusionment is particularly high for AI SDRs, which many found failed to generate meaningful opportunities. Success with AI is currently concentrated in three areas: intent-driven outbound, market intelligence (competitor analysis and data enrichment), and content marketing. The emerging GTM tech stack is consolidating around ChatGPT as the primary tool, followed by HubSpot and Clay. New players like Lovable and n8n are gaining traction as GTM teams increasingly seek technical flexibility to build custom agentic workflows and automated plays.

Kyle Poyar's Growth Unhinged | Weekly Newsletter | Kyle Poyar’s Growth Unhinged

Tactical B2B growth strategies for startups and scaleups center on the transition toward AI-native operations, where artificial intelligence serves as a core component of the Go-To-Market (GTM) stack. Insights from the 2025 State of B2B GTM and 2026 State of AI GTM reports reveal how industry leaders are adapting to market shifts and identifying new revenue drivers. Building a modern Account-Based Marketing (ABM) engine now involves integrating HubSpot with automation platforms like Clay and large language models like Claude to conduct deep research and personalized outreach. B2B monetization is also evolving, with a significant number of software companies shifting toward credit-based pricing and specialized AI packaging to capture the value of generative features. Data-driven benchmarks help organizations address the 'slow decay of growth' by providing frameworks to bend the growth curve back upward. Case studies of rapid scaling, including Replit’s journey to $100M ARR and Bolt.new’s $40M ARR in five months, illustrate the effectiveness of unconventional plays like leveraging personal emails for pipeline and hiring 'GTM Engineers' to bridge the gap between technical automation and sales. Furthermore, the emergence of AI agents as potential buyers suggests a future where traditional sales funnels and SaaS metrics must be entirely re-evaluated to account for non-human customers.

Sign Up | LinkedIn

Kyle Poyar is a prominent figure in the software-as-a-service (SaaS) and growth strategy space, best known as the creator and writer of the "Growth Unhinged" newsletter. Based in Greater Boston, he has amassed a significant following of over 100,000 on LinkedIn. His professional background is deeply rooted in market research, pricing strategy, and venture-backed technology. Poyar's educational foundation was established at Brown University, where he studied from 2006 to 2010. His early academic work included a publication in the Journal of Weather, Climate, and Society regarding climate change adaptation planning, demonstrating a long-standing interest in complex systems and strategic planning. Poyar's expertise is extensively documented through a diverse range of publications across major business and tech platforms. He has written specifically for OpenView on mastering SaaS pricing from seed stage to IPO and for Crunchbase News on the requirements for raising venture capital rounds. His insights extend to broader market dynamics, including enterprise tech trends in Venture Beat and competitive analysis of Amazon's tactics in international publications like The Guardian and Sydney Morning Herald. Additionally, he has explored monetization models for traditional media, advocating for premium models in newspapers and analyzing the future of video and pay TV. His work consistently bridges the gap between high-level market trends and actionable business strategies, particularly within the tech and media sectors, making him a key resource for understanding the mechanics of modern software growth.