Ramp's Arman Vaziri on Building Blocks for Go-to-Market Orchestration

Arman Vaziri is the leader of the product and sales-led growth engineering teams at Ramp. His presentation focuses on developing the foundational building blocks necessary for effective Go-to-Market (GTM) orchestration.

Once described, the system is designed to automate the distribution of these motions across multiple channels, including outbound communications, targeted advertisements, and web platforms.

This automation aims to resolve the significant bottlenecks that typically arise from manual coordination, ensuring a more efficient and scalable GTM strategy.

By automating these processes, companies can eliminate the need for extensive human intervention in the repetitive aspects of campaign execution.

To illustrate the concept, Vaziri uses an example of offering Pro V1 golf balls as an incentive to golfers working at East Coast construction companies to encourage them to try Ramp or engage with sales. The core idea is to move beyond managing individual channels and instead focus on expressing a clear intent.

The desired outcome is a seamless orchestration that automatically generates outbound sequences, ad copy, and in-app nudges based on this stated intent, simplifying complex multi-channel campaigns.

Historically, Go-to-Market (GTM) operations have faced several fundamental challenges, primarily stemming from messy and inconsistent data across disparate systems. This inconsistency means that various departments often operate using different sources of truth, leading to confusion and inefficiency.

Another significant challenge is the high operational burden placed on sales representatives, who frequently find themselves in back-to-back meetings with little time for preparatory or follow-up tasks. Furthermore, the processes of coordination and distribution have traditionally been costly and time-consuming, often extending over several months to implement a GTM motion.

Ramp's strategic approach to resolving GTM bottlenecks begins with a critical focus on data ingestion and ensuring data consistency. This addresses a foundational problem that is regularly prioritized on their roadmap each quarter.

The strategy then progresses to building vertical efficiency, which aims to save time and enhance conversion rates by streamlining specific workflows. Finally, Ramp seeks horizontal scaling by replicating common and successful workflows across different teams within the organization, fostering widespread efficiency.

To be effective, agents within the GTM orchestration system must possess a comprehensive understanding of the company's entire go-to-market landscape. This includes knowing why specific products are useful and how to segment buyers, prospects, and existing customers.

The strategy needs to accommodate a wide spectrum of customer engagement, ranging from cold prospects with no prior data to active users who have specific, articulated needs.

This platform integrates traditional data sources such as CRM data, product usage data, external enrichment data, web analytics, and buying signals.

The system also incorporates external signals, like funding announcements, and leverages internal behavioral models to provide a holistic view of customer interactions. Postgres databases are used to maintain transactional integrity across various CRM and product entities, ensuring reliable data synchronization.

This allows for immediate action and analysis based on current interactions.

Additionally, unstructured data, including call transcripts and notes, is indexed to make it searchable and retrievable, providing valuable context that might otherwise be overlooked.

This ensures that the data used for GTM orchestration is always up-to-date and thoroughly supplemented.

For offline processing, the system utilizes DBT and Snowflake to transform and analyze large datasets before piping the refined information back into operational systems via reverse ETL. This dual approach allows for both real-time enrichment and robust batch processing.

This ensures that initial implementations are robust and highly effective.

Common goals shared across different teams that benefit from this approach include the automation of outbound communications and the efficient preparation for sales meetings, highlighting areas where focused solutions can yield significant impact.

To enhance efficiency and avoid redundant computations, the system incorporates persistent fuzzy matching capabilities for meeting data. This feature allows for the matching of various data points across different business contexts, saving resources by not recomputing matches every time.

The system is specifically designed to support multiple businesses concurrently, ensuring that its matching algorithms can operate effectively in diverse environments.

Vaziri highlights that unstructured information, stored in warehouses, notes, or other internal repositories, is often the most valuable asset a company possesses. Leveraging real-time data such as call transcripts and emails provides rich context for agents.

To efficiently retrieve this information, the system combines vector search, attribute search, and keyword search, enabling precise and relevant results. The approach avoids ingesting the full corpus of data to maintain efficiency and prevent information overload for agents.

Ramp has developed a customizable skill library, allowing users to personalize their agents to align with their specific preferences and workflows. This feature is particularly useful for tasks like generating meeting briefs, where different individuals may require different formats.

The system combines pre-defined, system-level skills with user-provided instructions, offering a flexible and adaptable framework for agent customization.

Ramp has engineered its GTM orchestration platform in a way that provides all employees with access to the same tools and skills utilized by the background agents. This democratization of tools empowers teams to develop their own agents and automations.

This approach facilitates the automatic federation of capabilities across the entire organization, ensuring that successful automations can be easily shared and adopted company-wide.

The platform's role is to take these individual solutions and productionize them, effectively turning them into company-wide tools. This collaborative process allows staff to contribute prompts and 'vibe-coded' applications that can benefit many others facing similar challenges.

The vertical builds developed within Ramp's GTM orchestration serve as the foundational elements for multi-team and multi-channel distribution. These specialized tools enable the creation of standardized playbooks tailored to specific needs, such as selling to procurement departments or construction sectors.

This structured approach simplifies the testing of experimental ideas, like targeting golfers, by providing a robust framework for execution and measurement.

Go-to-Market (GTM) orchestration, when driven by centralized intent, transforms a single strategic goal into a comprehensive, multi-channel campaign. For example, a sales development representative (SDR) can specify the desire to create an audience, and the system will then generate personalized copy and sequences for outbound communication.

The orchestration extends to creating web landing pages, spinning up relevant images, and developing creative assets, all while maintaining human oversight for quality control. This ensures that every element of the campaign is aligned with the initial intent and executed effectively across all channels.

Arman Vaziri advises startups to avoid the trap of spending too much time building an overly complicated, perfect system architecture from the outset. Instead, he suggests focusing on identifying and automating very specific outbound sequences and solving immediate, existing problems.

The pragmatic approach involves iteratively piecing together vertical solutions. This allows startups to address pressing needs first and build out their GTM orchestration capabilities incrementally, rather than attempting a large-scale, all-encompassing build that could delay critical progress.