The RevvyAI Research Accounts node is in Beta. The product and this information may change. For questions contact your CSM.
Intelligent Workflows include nodes, called RevvyAI Research nodes, that use RevvyAI to create and use custom fields that assist with researching accounts and people.
For more information refer to RevvyAI Research Nodes Overview and Set Up RevvyAI Research Nodes.
RevvyAI Research nodes provide information that may not already exist in your 6sense or CRM data and requires research across publicly available sources. This information provides additional input to your GTM automation.
RevvyAI Research nodes use one or more RevvyAI fields, which specify the public data. You can create a RevvyAI field by asking a research question about an account or person. RevvyAI researches publicly available information and returns the result in a structured format that can then be used elsewhere in your workflow.
For example, you can research:
Whether a company was mentioned in a podcast or video
Whether a company provides professional services for its products
You can use the resulting information to enrich records, make decisions, or drive downstream workflow actions.
Example use case workflows for a RevvyAI Research node
Competitive intelligence
Example prompts:
"Which competitors does [Name] company publicly reference?"
Example workflow:
segment → RevvyAI Research node → decision node → action node
Business intelligence
Example prompts:
What strategic initiatives is [Name] company currently focused on?
What podcasts mention [Name]?
Does [Name] provide professional services for its product?
Example workflow:
segment → RevvyAI Research node → decision node → enrichment node
Personalize account strategy
Example prompts:
"What are [Name] company’s publicly known strategic priorities?"
Example workflow:
segment → RevvyAI Research node → decision node → enrichment node
Best practices for RevvyAI field prompts
Criteria for good research questions
Good research questions are:
Specific and clearly defined
Based on information that can reasonably be found publicly
Relevant to the account or person
Clear about the timeframe when one is important
Clear about what the expected answer should contain
Example of a good prompt: "Has this [Name] company publicly announced a major technology platform migration in the last 18 months?"
Example of a less effective prompt: "Tell me everything about this [Name] company's technology."
The more specific the question, the easier it is for RevvyAI to return a useful and actionable result.
Avoid subjective questions
You can create research questions that involve interpretation, but you should make the criteria as explicit as possible.
Instead of:
"Is [Name] a good company?"
Use:
"Based on publicly available information, does [Name] meet all three of the following criteria: operates in North America, sells B2B software, and has more than 500 employees? Return true only if all three conditions are met."
Clear criteria generally produce more consistent results.
Avoid asking for speculation
Avoid prompts that require Revvy AI to present speculation as fact. For example, instead of:
"Will this company buy marketing software next year?"
Ask:
"Identify publicly available signals that indicate this company may be evaluating or investing in marketing technology."
This keeps the research grounded in observable evidence rather than requiring the model to predict the future.
Do not ask RevvyAI to research personal or sensitive information
RevvyAI Research nodes should be used for business-relevant research based on appropriate publicly available information.
Do not use prompts to request sensitive personal information, personal contact information, or other information that should not be used for business research.
When appropriate, use information already available in 6sense instead of in a RevvyAI Research node
If the information already exists as a supported 6sense field, generally you should use the existing 6sense data rather than consuming RevvyAI Research node credits to recreate it.
The RevvyAI Research node is most valuable for net-new intelligence that requires external research.
Document changelog
Sep 1 2026: First Beta release