Product descriptions:
ServiceNow helps thousands of enterprises streamline digital workflows across IT, employee and customer experiences. But behind their GTM engine was a maze of disconnected systems, slowing the decision-making of their 2,000+ sellers, impacting sales processes and stalling growth. By adopting the Databricks Data Intelligence Platform, they unified data and deployed AI models in real time. Databricks powers their Lead Scoring, Outreach Assist and Demo Assist solutions, cutting processing times from hours to minutes and doubling email reply rates. “Databricks helps us go from idea to production faster,” Mili Merchant, Senior Product Manager at ServiceNow, said.
Overcoming data silos to accelerate GTM performance
To support ServiceNow’s ambitious goal of reaching $15 billion in revenue by 2026, the company’s go-to-market (GTM) teams, which include over 2,000 global sellers, needed a more scalable and intelligent way to operate. However, the path to growth was mired in inefficiencies such as siloed data, disconnected systems and time-consuming manual workflows. Sellers spent hours every day searching for leads, researching prospects and assembling demo materials. This left limited time for what mattered most: building relationships and closing deals.
The GTM AI team identified three key stages in the deal journey — lead scoring, outreach and demo creation — where AI could make a transformative impact. But existing infrastructure created roadblocks. Lead Scoring relied on batch processes that delayed follow-up. Outreach personalization required 20+ minutes of manual research per lead. And demo decks took up to 24 hours to assemble. These inefficiencies were tied to deep-rooted technical challenges, including that data was siloed across platforms like Snowflake, Adobe Experience Platform and Dynamics — and there was no unified platform to consolidate and act on this information in real time.
Latency wasn’t just a technical inconvenience. It was a barrier to seller effectiveness. “Historically, our sellers spent too much time navigating fragmented systems and slow processes instead of doing what they do best — selling,” Amulya Gupta, NLP scientist at ServiceNow, said. The challenge grew with the rise of generative AI, which demanded rapid orchestration, real-time data and scalable infrastructure. The GTM AI team needed a central execution layer to support model deployment, observability and iteration across a fast-growing portfolio of AI tools.
Bringing AI to life with end-to-end automation
ServiceNow adopted the Databricks Data Intelligence Platform as the core execution layer for their GTM AI strategy, enabling scalable, real-time intelligence across the deal journey. All three flagship solutions — Lead Scoring, Outreach Assist and Demo Assist — now run on Databricks. For Lead Scoring, ServiceNow built a production pipeline that processes over a million leads per year using more than 1,000 behavioral and firmographic signals. MLflow manages experimentation and version control, enabling the team to retrain models quickly without interrupting production. Previously, lead scoring relied on 15-minute batch cycles and hours-long data prep. Now, it runs in real time, reducing latency from four hours to just 30 minutes and helping sellers respond faster.
Outreach Assist uses large language models (LLMs) to generate highly personalized prospecting emails in under two minutes. Integrated with sellers’ email tools via API, it draws from public data, internal assets and case studies to ensure relevance. The AI engine is orchestrated with Lakeflow Jobs and includes multistep LLM chains with fallback systems and monitoring to ensure quality and control. “We built the orchestration framework in Databricks so we could modularize the stages of each message, from intro to tailored customer example to CTA, and maintain reliability at scale,” Amulya said.
Demo Assist rounds out the GTM AI stack. Using a propensity model trained in Databricks, it predicts which products a prospect is most likely to purchase and auto-generates customized pitch decks complete with messaging and curated customer success stories. What used to take a full day of manual assembly now happens in minutes. This has unlocked faster, more relevant conversations at the final stage of the deal cycle.
“Databricks is our execution layer,” Mili said. “It powers the engines behind Lead Scoring, Outreach Assist and Demo Assist — helping us go from idea to production faster and with more confidence.”
Boosting seller performance and accelerating deal cycles
Databricks now serves as a key foundational piece supporting every stage of ServiceNow’s GTM funnel, from initial lead identification to personalized outreach to high-impact demo delivery. With Databricks as the execution layer, these AI solutions work in concert to increase seller efficiency, accelerate deal cycles and drive measurable business outcomes.
At the top of the funnel, Lead Scoring helps sellers focus their energy on high-conversion prospects. With real-time scoring powered by MLflow, sellers can now act within 24 hours of engagement. The system scores leads with 91% accuracy, and those prioritized by the model are 3x more likely to convert into the pipeline. “When sellers focus on MQLs [marketing quality leads], they’re seeing 3.2x more meetings, 3.3x more opportunities and 3x more pipeline,” Amulya said.
In the middle of the funnel, Outreach Assist transforms how sellers engage prospects, replacing time-consuming manual research with hyper-personalized generative AI emails. In 2024 alone, it generated 65,000+ emails, reducing creation time from 20 minutes to under two and driving a 3.3x lift in meetings. “We expected a better reply rate, but we didn’t expect such a lift in actual meetings,” Amulya explained. “Generative AI helped establish relevance between our solutions and the customer’s needs.”
At the bottom of the funnel, Demo Assist accelerates late-stage conversations by generating customized presentations in minutes, reducing prep time by 24 hours and increasing deal velocity by 1.5x. The result is a more responsive sales process that equips sellers with tailored content exactly when they need it. This leads to faster cycles, more qualified conversations and higher close rates.
Looking ahead, ServiceNow is building on this momentum by developing a GTM knowledge graph and conversational agent framework to further enhance seller decision-making and collaboration. “Databricks has served as the execution layer for our entire GTM engine, saving days of work and delivering the intelligence our sellers need when they need it,” Mili concluded.