Second-Party Data: AI Unlocking Ecosystems

Bob Moore's entrepreneurial journey is deeply rooted in his background as a "data nerd," having studied computer science and operations research. His career began at a venture capital firm, Insight Partners, where he quickly recognized a significant gap: many highly successful businesses lacked sophistication in data analytics, struggling with concepts like customer lifetime value and cohort analysis. This observation led him to manually assist these companies in analyzing their data using SQL and Excel, eventually sparking the idea for his first company, RJ Metrics. Launched in 2008 with co-founder Jake Stein, RJ Metrics aimed to productize this analytical work, bringing "venture capital grade analytics" via SaaS to a broader audience. This early experience laid the groundwork for understanding how businesses could leverage their data more effectively.

The second company, Stitch Data, emerged directly from challenges observed at RJ Metrics. Businesses often had their data scattered across various systems, such as Shopify, inventory platforms, payment processors, and marketing automation tools like HubSpot or MailChimp. This fragmentation made it incredibly difficult for companies to consolidate data to answer complex business questions, like whether a marketing campaign led to refunds. We designed Stitch as an intuitive platform for extracting data from these disparate API endpoints and depositing it into a centralized data warehouse, coinciding with the rise of cloud data warehouses like Amazon Redshift and Snowflake. This low-friction product achieved rapid success, and Talend acquired it in 2018.

The experiences at Stitch, particularly the realization that Stitch was often purchased as part of a broader data stack (e.g., alongside Snowflake or Looker), directly led to the genesis of Crossbeam. This insight highlighted that the most qualified prospects for Stitch were those who had just invested in complementary technologies. The problem of effectively collaborating with these partners through traditional, flawed account mapping methods—involving manual spreadsheet exchanges with issues of oversharing and data quality—became evident. Furthermore, the ephemeral nature of valuable account mapping information meant that annual or quarterly updates were insufficient, leading to the vision for Crossbeam as a real-time, secure data escrow service.

Crossbeam fundamentally addresses the challenge of securely comparing CRM data across company lines to identify overlaps in sales pipelines, target markets, and common customers. It functions as the world's leading platform for account mapping, designed to preserve data security and privacy by ensuring that only the right partner sees the correct data under specific conditions. This unique capability stemmed from Bob Moore's obsession with a "crossbeam problem" experienced at Stitch, where understanding partner overlaps was crucial for revenue but hindered by manual, inefficient processes. The platform acts as an "escrow service for data," pulling and transforming data from various systems of record, allowing companies to control what specific slices of data are shared with partners when certain conditions are met.

The adoption of Crossbeam within an organization often follows a distinct pattern, starting with partnership ecosystems teams. Its viral nature means that seven out of ten companies sign up after being invited by an existing user, making partnership teams instrumental in initiating its use due to their role as "social butterflies" and relationship owners. These teams are responsible for inviting partners and setting up data-sharing rules. However, as the platform becomes more integrated, it unlocks a valuable data layer that serves as a critical intelligence asset for the entire revenue team. This leads to a secondary persona: Chief Revenue Officers (CROs), sales managers, and individual sales representatives, who leverage this ecosystem data as a continuous source of signals about their accounts.

Crossbeam supports multiple "legs" or personas within an organization, providing a comprehensive view of partner ecosystems. For classic tech partnerships, the primary use case is prioritizing which partners are worth investing time in by assessing the overlap of customer bases and identifying if these overlapping customers are inherently better (e.g., higher ACV, lower churn). Within direct sales, Crossbeam helps accelerate deals and multi-thread opportunities and improve account prioritization by providing insights into complementary product usage and key contacts known by partners. For channel teams, it is invaluable in avoiding channel conflicts and facilitating better decisions earlier in the sales funnel by providing selective account mapping visibility.

Bob Moore firmly believes that network effects and network-effect-related businesses will be the crucial differentiator in the AI era, primarily because they generate a unique source of knowledge that large language models (LLMs) cannot replicate from publicly available internet data. He differentiates between three classes of data: third-party, first-party, and second-party. Third-party data, which includes publicly available information like email addresses or job titles, is rapidly being commoditized by AI tools, driving a "race to the bottom" in that industry. Though proprietary to an organization, first-party data is limited to existing customers and contacts, offering restricted "greenfield" growth opportunities.

The "holy grail" is second-party data – the knowledge partners possess and are willing to share, which is proprietary and unlocks new growth channels. This data exists within a "specialized walled garden" of interconnected partner ecosystems, accessible only through these collaborations. Partners know things that a company does not, creating new "vectors of potential action" that would otherwise be unavailable. This inherent proprietariness ensures that LLMs cannot simply scrape this data, making it a sustainable source of competitive advantage.

While powerful, the context we provide ultimately limits AI agents and co-pilots. The true differentiation in the AI era will come from the quality and relevance of this contextual data. Second-party data is the "holy grail" that pulls this all together, informing AI models on who to contact, how to message, how to prioritize, and how to execute sales and marketing strategies more effectively. This intersection of partnerships and AI is a "beautiful match," unlocking immense flexibility and value creation without requiring extensive custom programming of every use case. AI models are proving adept at ingesting unstructured or incorrectly structured data and reorganizing it on the fly, making second-party data more accessible and flexible for driving insights.

When considering the application of AI models to second-party data over the next 12-24 months, the strategic direction is likely to be driven more by end-user needs than by technological capabilities alone. Bob Moore suggests a "bottoms-up approach" is more effective, where we apply AI to "atomic units" of action before building up to broader strategies and campaigns. For instance, Crossbeam's Deal Navigator product dynamically prioritizes individual accounts for sales representatives by applying deterministic and AI-powered scoring algorithms at an atomic level. This involves analyzing thousands of accounts and rapidly contextualizing them with localized, walled-garden insights from partners.

The LLMs in this context receive finite, discrete inputs for each account and make assessments and scores. While higher-order insights, such as best-performing partners or ideal deal characteristics, serve as inputs, the core AI analysis occurs at the granular account level. We then use classic software engineering methods to roll these AI-driven insights into a singular, consumable view for human users. This allows for a broad strategic impact—like every sales manager leveraging ecosystem data in weekly pipeline reviews—by applying AI effectively at the individual account level.

This approach balances the need for accuracy, narrow-range predictability, and scalability. Focusing AI on discrete problems helps mitigate hallucinations and improves accuracy. The challenge with traditional second-party data for software engineering is its nested and complex organization, making it harder to retrieve and parse through different lenses (e.g., by partner or by account). However, LLMs excel at reorganizing unstructured data on the fly, offering immense flexibility when dealing with hundreds of accounts across dozens of partners. This capability makes second-party data more accessible and valuable, driving its growing role in future applications.

The landscape of technology tools supporting channel, partnership, and sales teams is evolving from isolated "islands of technology" to a cohesive "stack," much like the modern data stack that emerged years ago. Bob Moore views platforms like Crossbeam (for data orchestration) and ZINFI (for workflow orchestration and PRM) as complementary pieces that create a "one plus one equals three" scenario. While data can coexist in primary record systems like Salesforce, the actual value emerges when we implement these solutions to enable end-to-end playbooks. Crossbeam's insights on prioritizing partners and deals and customizing approaches act as a force multiplier when layered on platforms that manage partner onboarding, enablement, co-marketing, and co-selling workflows.

This convergence leads to a significantly more strategic and comprehensive approach to partner ecosystems. For example, Crossbeam's data orchestration can identify "who to go after, why to go after, and where the opportunities are," providing crucial signals. Workflow orchestration tools can then automate actions based on these insights, such as sending notifications for partner collaboration on a deal, launching joint promotions, or managing incentive payouts. The combined goal is to scale and simplify complex workflows and data insights. This tight integration ensures that the data drives direct action, and the success of those actions can feed back into improving future recommendations, creating a powerful closed-loop system for growth.

Looking ahead, the AI era continues to drive rapid change, emphasizing the importance of strong collaborations and adaptability. Bob Moore reflects on his entrepreneurial journey, advising aspiring founders to surround themselves with ambitious, low-ego collaborators who can provide grounded thinking and accelerate good decision-making. For the future, he anticipates paradigm shifts in human existence and societal values due to AI's rapid advancement. This underscores a need for continuous attention and strategic positioning to stay ahead of these transformative changes, highlighting that the intersection of partnerships and AI is still in its early stages of being fully leveraged for its immense potential.