OpenAI, the pioneering AI research and deployment company, changed the world as we know it with the launch of ChatGPT in 2022. When they launched ChatGPT Enterprise, just nine months later, they unleashed a firehose of inbound interest.
“We were under-staffed for the demand we were facing,” recalls Keith Jones, OpenAI's GTM Systems Lead. “Our inbound lead enrichment process, which relied on a single provider, left huge gaps.”
In his search for a multi-source enrichment provider that could grow with OpenAI's flexible needs, Keith discovered Clay. "Clay was a way to leapfrog us into a multi-source model with the lowest amount of overhead and labor," he explains.
Today, Clay remains a core part of the tech stack of OpenAI’s growing team. Both RevOps team members and the sales organization use it for lead enrichment, account research, and targeting. In the words of Scotty Huhn from OpenAI’s RevOps team, “Clay enables our team to rapidly experiment with trigger driven workflows, and 3rd party enrichment data. We’re able to move fast and drive outsized impact on GTM execution – all while using a tool that’s fun, creative, and cutting edge.”
Impact Highlights:
- More than doubled enrichment coverage from low 40% to high 80%
- Maintained consistent weekly usage across sales team members using Clay’s on-demand enrichment Salesforce package, with 8500+ total enrichments run
- Rapidly iterated on enrichment strategies without disrupting seller workflows
- Successfully scaled from single-source to multi-source data model with minimal overhead
Use Case 1: Doubling inbound lead enrichment rates
OpenAI needed to improve their enrichment coverage for inbound leads. Keith felt that transitioning to a multi-source enrichment model was inevitable: “It’s an inflection point that every B2B enterprise faces at some point. If you're not multi-source, you're going to be one day. No company can get by on one data enrichment provider, not even the best provider in the world.”
The traditional playbook for shifting to a multi-source enrichment model often requires contracts with multiple vendors, a customer data platform, a reverse ETL, and other traditional marketing tools. But OpenAI didn’t have time to implement a complex multi-tool workflow.
Clay's multi-provider waterfall enrichment augmented by OpenAI’s models provided an immediate solution. "With Clay, we more than doubled our enrichment coverage from low 40% to high 80%," Keith notes. The stronger data foundation helped leads get properly scored, routed, and responded to.
Harsha Chilakamarri, also from the RevOps team, says, “Clay has allowed us to easily leverage multiple providers to scale our inbound lead enrichment process. I've been able to build complex systems that used to require days of data science work now within hours.”
OpenAI continues to iterate on Clay’s flexible waterfall configuration to achieve the best mix of data providers for them. "We weren't even two weeks in before we refined our combination and ordering of providers," Keith explains.
Use Case 2: Custom GTM research at scale using AI
OpenAI used Clay’s AI research agent to mimic and augment the research process of their best sales reps—at scale.
“We started by automating what our best sellers already did,” Keith said. “They visit company websites, check LinkedIn pages and profiles, and look for key information. They search for a company’s latest developments, revenue figures, any significant changes in the last 90 days—maybe skim quarterly earnings reports to find insights to use in their outbound.”
The RevOps team used Clay's AI agent with GPT models to automate this workflow and summarize information from earnings reports, company websites, and other key sources for every relevant account. The AI agents inspired trust by showing their reasoning process, explaining, for example, how they determined revenue from an earnings report.
"Having an approach that replaces repetitive manual work at scale on sellers' behalf is a boon," Keith said. Best of all, the AI agent is infinitely flexible: any research process a sales rep would do manually can be automated. OpenAI is able to do a lot of basic data enrichment, such as finding LinkedIn profiles or enriching company industry/size, purely agentically as well. "We are doing enrichment with OpenAI models," Keith explains. Our strategy sees us employing a combined application of dedicated data sources, and native enrichment provided by Clay which is further augmented by the use of our models.
Use Case 3: Helping sellers research on-demand—in Salesforce
Though RevOps teams can operationalize and scale data enrichment behind the scenes, sales and success teams will always need access to enriched data on an ad hoc, on-demand fashion.
The conventional approach to on-demand enrichment requires a dedicated third-party data enrichment tool, ideally integrated with your CRM. Sales and success team members must log into this tool to either extract data directly or trigger updates to corresponding CRM records. Although this approach works, it creates an additional burden: GTM team members must constantly context switch between applications, and RevOps/Systems teams must provide training and maintain the setup over time.
Instead, OpenAI implemented bespoke workflows, dubbed “Enrichment Actions”, as a Clay package within Salesforce itself “We unlocked a way to make our advanced multi-source on-demand data enrichment available to our GTM staff natively within our CRM," Keith explains. “It didn’t require painful training and a new context switch for our staff.”
Now, OpenAI GTM staff can:
- Enrich individual account records with the latest multi-source waterfall
- Request batches of curated leads from any given account record
- Enrich individual leads or multiple lead records at a time
Clay’s Salesforce package sees heavy use across the sales team. On busy days, individual sellers may run up to 150 lead enrichments, with especially heavy usage at the start of quarters as teams build pipelines.
The RevOps team can update and improve the backend enrichment logic without disrupting the simple click-through experience for sellers. “We were able to launch a version of it, brand it the way we wanted to,, and then refine the strategy based on the input from the field,” Keith explains. The team has made numerous improvements to their data sources and waterfall prioritization since first launching the tool.
"In the near future, we will likely develop new iterations that will bring additional benefits to our GTM staff, using Clay's advanced data enrichment in unique ways—augmented by OpenAI's technology and deeper integration points to Salesforce."
How Clay continues to accelerate OpenAI's GTM motion
Clay helped OpenAI build a strong data foundation with the flexibility to implement any growth use case.
Since implementing Clay, usage remains strong despite a much larger team. “One of the main reasons GTM technology fails is inconsistent adoption," Keith explains. “But for us, Clay has had consistent adoption week over week,”—across the Revenue Operations, Sales, and GTM team.
What's next for OpenAI and Clay? Other teams at OpenAI like data science, marketing ops, and recruiting are now inspired by what's possible for them in Clay. Data science is starting to use Clay to estimate the number of knowledge workers at each company across their user base, and recruiting is exploring using Clay to find people who were at high growth companies during peak years of hypergrowth.
Clay’s flexibility and ability to support rapid experimentation has been key. "I think it's worth overstating how flexible Clay is and how many different needs it can meet," Keith says. Their RevOps team continuously improves their data enrichment strategy by tweaking sources and AI agent workflows, using both their existing providers and Clay integrations.
Most importantly, Clay has enabled OpenAI to stand up automated, recurring GTM infrastructure while maintaining the flexibility to handle emergent requests from GTM teams. The platform supports OpenAI's data sourcing and activation layer, with Keith noting, "We see outsized growth potential with Clay."
OpenAI, the pioneering AI research and deployment company, changed the world as we know it with the launch of ChatGPT in 2022. When they launched ChatGPT Enterprise, just nine months later, they unleashed a firehose of inbound interest.
“We were under-staffed for the demand we were facing,” recalls Keith Jones, OpenAI's GTM Systems Lead. “Our inbound lead enrichment process, which relied on a single provider, left huge gaps.”
In his search for a multi-source enrichment provider that could grow with OpenAI's flexible needs, Keith discovered Clay. "Clay was a way to leapfrog us into a multi-source model with the lowest amount of overhead and labor," he explains.
Today, Clay remains a core part of the tech stack of OpenAI’s growing team. Both RevOps team members and the sales organization use it for lead enrichment, account research, and targeting. In the words of Scotty Huhn from OpenAI’s RevOps team, “Clay enables our team to rapidly experiment with trigger driven workflows, and 3rd party enrichment data. We’re able to move fast and drive outsized impact on GTM execution – all while using a tool that’s fun, creative, and cutting edge.”
Impact Highlights:
- More than doubled enrichment coverage from low 40% to high 80%
- Maintained consistent weekly usage across sales team members using Clay’s on-demand enrichment Salesforce package, with 8500+ total enrichments run
- Rapidly iterated on enrichment strategies without disrupting seller workflows
- Successfully scaled from single-source to multi-source data model with minimal overhead
Use Case 1: Doubling inbound lead enrichment rates
OpenAI needed to improve their enrichment coverage for inbound leads. Keith felt that transitioning to a multi-source enrichment model was inevitable: “It’s an inflection point that every B2B enterprise faces at some point. If you're not multi-source, you're going to be one day. No company can get by on one data enrichment provider, not even the best provider in the world.”
The traditional playbook for shifting to a multi-source enrichment model often requires contracts with multiple vendors, a customer data platform, a reverse ETL, and other traditional marketing tools. But OpenAI didn’t have time to implement a complex multi-tool workflow.
Clay's multi-provider waterfall enrichment augmented by OpenAI’s models provided an immediate solution. "With Clay, we more than doubled our enrichment coverage from low 40% to high 80%," Keith notes. The stronger data foundation helped leads get properly scored, routed, and responded to.
Harsha Chilakamarri, also from the RevOps team, says, “Clay has allowed us to easily leverage multiple providers to scale our inbound lead enrichment process. I've been able to build complex systems that used to require days of data science work now within hours.”
OpenAI continues to iterate on Clay’s flexible waterfall configuration to achieve the best mix of data providers for them. "We weren't even two weeks in before we refined our combination and ordering of providers," Keith explains.
Use Case 2: Custom GTM research at scale using AI
OpenAI used Clay’s AI research agent to mimic and augment the research process of their best sales reps—at scale.
“We started by automating what our best sellers already did,” Keith said. “They visit company websites, check LinkedIn pages and profiles, and look for key information. They search for a company’s latest developments, revenue figures, any significant changes in the last 90 days—maybe skim quarterly earnings reports to find insights to use in their outbound.”
The RevOps team used Clay's AI agent with GPT models to automate this workflow and summarize information from earnings reports, company websites, and other key sources for every relevant account. The AI agents inspired trust by showing their reasoning process, explaining, for example, how they determined revenue from an earnings report.
"Having an approach that replaces repetitive manual work at scale on sellers' behalf is a boon," Keith said. Best of all, the AI agent is infinitely flexible: any research process a sales rep would do manually can be automated. OpenAI is able to do a lot of basic data enrichment, such as finding LinkedIn profiles or enriching company industry/size, purely agentically as well. "We are doing enrichment with OpenAI models," Keith explains. Our strategy sees us employing a combined application of dedicated data sources, and native enrichment provided by Clay which is further augmented by the use of our models.
Use Case 3: Helping sellers research on-demand—in Salesforce
Though RevOps teams can operationalize and scale data enrichment behind the scenes, sales and success teams will always need access to enriched data on an ad hoc, on-demand fashion.
The conventional approach to on-demand enrichment requires a dedicated third-party data enrichment tool, ideally integrated with your CRM. Sales and success team members must log into this tool to either extract data directly or trigger updates to corresponding CRM records. Although this approach works, it creates an additional burden: GTM team members must constantly context switch between applications, and RevOps/Systems teams must provide training and maintain the setup over time.
Instead, OpenAI implemented bespoke workflows, dubbed “Enrichment Actions”, as a Clay package within Salesforce itself “We unlocked a way to make our advanced multi-source on-demand data enrichment available to our GTM staff natively within our CRM," Keith explains. “It didn’t require painful training and a new context switch for our staff.”
Now, OpenAI GTM staff can:
- Enrich individual account records with the latest multi-source waterfall
- Request batches of curated leads from any given account record
- Enrich individual leads or multiple lead records at a time
Clay’s Salesforce package sees heavy use across the sales team. On busy days, individual sellers may run up to 150 lead enrichments, with especially heavy usage at the start of quarters as teams build pipelines.
The RevOps team can update and improve the backend enrichment logic without disrupting the simple click-through experience for sellers. “We were able to launch a version of it, brand it the way we wanted to,, and then refine the strategy based on the input from the field,” Keith explains. The team has made numerous improvements to their data sources and waterfall prioritization since first launching the tool.
"In the near future, we will likely develop new iterations that will bring additional benefits to our GTM staff, using Clay's advanced data enrichment in unique ways—augmented by OpenAI's technology and deeper integration points to Salesforce."
How Clay continues to accelerate OpenAI's GTM motion
Clay helped OpenAI build a strong data foundation with the flexibility to implement any growth use case.
Since implementing Clay, usage remains strong despite a much larger team. “One of the main reasons GTM technology fails is inconsistent adoption," Keith explains. “But for us, Clay has had consistent adoption week over week,”—across the Revenue Operations, Sales, and GTM team.
What's next for OpenAI and Clay? Other teams at OpenAI like data science, marketing ops, and recruiting are now inspired by what's possible for them in Clay. Data science is starting to use Clay to estimate the number of knowledge workers at each company across their user base, and recruiting is exploring using Clay to find people who were at high growth companies during peak years of hypergrowth.
Clay’s flexibility and ability to support rapid experimentation has been key. "I think it's worth overstating how flexible Clay is and how many different needs it can meet," Keith says. Their RevOps team continuously improves their data enrichment strategy by tweaking sources and AI agent workflows, using both their existing providers and Clay integrations.
Most importantly, Clay has enabled OpenAI to stand up automated, recurring GTM infrastructure while maintaining the flexibility to handle emergent requests from GTM teams. The platform supports OpenAI's data sourcing and activation layer, with Keith noting, "We see outsized growth potential with Clay."
.jpg)
.jpg)
.avif)
