Cloudflare Leverages AI Agents to Double GTM Team Efficiency

Cloudflare is proposing a three-pillar approach to enhance the efficiency of its Go-To-Market (GTM) teams. This framework aims to reduce the time spent on data-related tasks from hours to mere minutes, thereby enabling human teams to concentrate on higher-level strategic activities.

Justin Joyce stated, "the framework that I'm proposing here, which I think is is it's going to really be effective in as we flesh this out in the future is a three-pillar approach." The strategy focuses on scaling analytical capabilities, delivering timely insights, and providing self-service options to empower GTM teams.

The second pillar of Cloudflare's GTM strategy involves scaling insights by transforming raw data into actionable stories for the team. This moves beyond reactive data requests to proactively delivering essential information.

Justin Joyce explained, "The second pillar is how can is to scale insight. There's story in the data, and how can we provide that to the team?" This approach tailors information to different management tiers and provides relevant context for sales interactions.

The third and arguably most significant pillar is to provide self-service capabilities to the Go-To-Market team. This allows sales teams to immediately access expert-level information without needing to consult other departments.

Justin Joyce stated, "And the third one, which is arguably the biggest one, is uh to provide self-service capabilities to the go-to-market team." This empowers representatives to independently assess situations during customer calls, covering aspects like handling rejections, upselling opportunities, and customer satisfaction.

Cloudflare has developed role-specific skill files that embed business context directly into raw data, making it accessible for both technical and non-technical users. This innovation allows individuals without SQL knowledge to query data independently.

Justin Joyce noted, "what we have done as we have built role-specific skill files, which have the context of the business information tying it to the data." This approach significantly reduces bottlenecks that typically arise from relying on dedicated data engineers, democratizing data access across the organization.

The skill files are pre-loaded with common business inquiries to optimize query coverage for complex business tasks. This design allows the system to handle approximately 80% of routine questions, reserving more intricate strategic challenges for human intervention.

Justin Joyce explained, "we've included the types of questions that the business would ask of the data... so that we can answer essentially 80% or or more of the questions." This strategy reduces dependency on centralized data teams, allowing them to focus on unique and high-value analytical work.

Cloudflare is automating data-driven storytelling by proactively pushing insights to users rather than making them search for information. This approach is designed to deliver relevant data directly to the Go-To-Market teams.

Justin Joyce stated, "there's a story in the data and they really shouldn't have to search for it. ... we bring the story to them." Weekly summaries highlight business performance and goals, while trends and watch items are delivered directly to sales teams, modeling the user experience after mobile digital assistants.

Cloudflare acknowledges that dashboard adoption varies significantly among employees within GTM teams, with some loving dashboards and others rarely using them. This necessitates automated analysis to bridge data understanding gaps across the business.

Justin Joyce noted, "You're going to have people who are love dashboards, people are never going to look at them. So, I think you really need to have a way to scaffold that across the business." The goal is to simplify data structures to ensure consistent and effective processing by AI agents, making insights accessible to everyone regardless of their preferred data consumption method.

Cloudflare is actively transforming raw business data into AI-ready formats through a meticulous data engineering and preprocessing phase. This involves structuring data by specific dimensions such as time, manager, and theater, as well as by metrics.

Justin Joyce explained, "Here what we do is we transform the data by the dimension of time, also slice of the logical part of the business, which is manager, theater, and finally the metric." The process includes converting wide data into a long format to facilitate trend analysis and embedding business logic and aggregation rules upfront, ensuring the data is optimized for AI processing.

Cloudflare employs a multi-agent workflow architecture to ensure both accuracy and appropriate tone in its AI-generated outputs. This system involves a sequence of specialized agents collaborating to process and present data.

Justin Joyce detailed, "we have a multi-agent workflow where we first get the data, and then we do a first pass draft on the data calling our MCPs, um and then this we have a second reviewer agent who checks the veracity of the data, and then we have a third um agent, which is a tone agent." The first agent retrieves data and creates an initial draft, the second agent reviews for data veracity, and a third tone agent uses multi-shot prompts to balance risk and opportunity, all with full observability into every LLM call.

Cloudflare has introduced an internal tool called Cloudflare OS, which functions as an agentic workspace designed to empower GTM teams with self-service capabilities. This platform runs on Cloudflare's infrastructure, leveraging Cloudflare Workers and Durable Objects for storage.

Justin Joyce stated, "What I'm showing you here is our internal tool called Cloudflare OS, which is an agentic workspace that is running on Cloudflare where the go-to-market team can come in here." This system allows sales teams to retrieve specific data on-demand, streamlining their access to crucial business information.

AI agents are being actively utilized across various Go-To-Market (GTM) use cases, significantly enhancing sales operations and account management at Cloudflare. These applications span from generating critical reports to preparing for customer interactions.

Justin Joyce listed, "Some use cases that these teams are using it for is doing a forecast brief, building QBR decks, building a purchase deck on what the customer that they're onboarding has purchased, doing account planning, general queries of the data, as well as renewal preparation." Specific applications include creating forecast briefs, building QBR (Quarterly Business Review) and purchase decks for customer onboarding, facilitating account planning, conducting general data queries, and preparing for renewals to maximize upsell and adoption opportunities.

Cloudflare maintains a centralized alias for presenting skills, which are then curated collaboratively by the GTM and operations teams. This structured process ensures the repository maintains expert-level accuracy and prevents the proliferation of redundant or ineffective skills.

Justin Joyce noted, "We have a central alias where skills are presented uh to uh the central team, curated by the uh go-to-market team, as well as by operations team, and they're reviewed, so we can make sure that we're not having a proliferation of skills." This careful curation provides a robust foundation of expert knowledge to address various customer situations effectively.

Effective Go-To-Market (GTM) support relies on three core pillars: self-service data access, pushing proactive insights to the business, and scaling the analytics team to meet demand. Without all three, organizations face significant challenges in optimizing agentic systems.

Justin Joyce emphasized, "And so, bring it together with the three pillars that I talked about. If you really don't have all these, I think you have issues with serving the go-to-market needs in terms of optimizing the use of agentic systems." These pillars ensure that GTM teams have timely, relevant data and analytical support to drive strategic initiatives and respond effectively to market needs.

Skill curation forms the fundamental basis for all agentic workforce initiatives, as it involves embedding comprehensive business knowledge into skill files. This ensures that AI agents can operate predictably and deterministically.

Justin Joyce stated, "The first thing is skill curation is the basis for all of this agentic workforce. If you're able to embed the knowledge of the business into the skill files as well as the skills for an analyst to be able to build answer questions or build applications... you're really able to give them the ability to use the agentic systems in a more predictable and deterministic way." By deeply integrating business context, AI systems can answer questions and build applications with greater accuracy and reliability.

Cloudflare stresses the critical importance of feedback loops throughout its AI development and implementation process for internal teams. Treating internal teams similarly to external product customers allows for continuous, iterative improvement.

Through the implementation of its three-pillar framework and agentic systems, Cloudflare has successfully doubled the efficiency of its Go-To-Market teams. This significant improvement has enhanced service capabilities and provided teams with essential job-related information.

Justin Joyce confirmed, "Through all these different pillars that I talked about, we've really been able to 2x our efficiency and be able to serve the teams as well as allowing them to be able to get the information that they need to do their job." This measured success underscores the effectiveness of their AI integration strategy.

Cloudflare plans to deepen the integration of its AI agentic systems with existing workflows and systems, including CRMs like Salesforce. The goal is to automate more complex tasks and embed necessary artifacts directly into meetings.

Justin Joyce outlined, "Number one is a deeper integration with our systems that we work in... how can we set up meetings for the go-to-market team and embed those artifacts in those meetings so they don't have to actually pull them." This includes automatic artifact embedding in QBRs, seamless CRM updates, and managing data security and access, all to further streamline operations and reduce manual effort.

Cloudflare recognizes that the use of agentic systems has reached a 'Cambrian stage,' characterized by an explosion of excitement and diverse AI use cases. However, the company is now focused on moving beyond experimentation towards standardization.

Justin Joyce observed, "We're sort of reached the Cambrian stage of using agentic systems, which means there's an explosion of excitement and skills and finding out ways to solve anything with AI. But, I see as we get to this fuller integration and standardization, we're going to want to come back and and not really limit, but just figure out a really strategic approach." This transition involves developing a strategic approach for fuller integration and standardization, ensuring that AI solutions are aligned with overarching business objectives rather than merely solving isolated problems.