Ask Cal: Push to Salesforce

CaliberMind creates a unified dataset and gives customers the ability to create different data models to answer difficult questions. With the new Ask Cal Push to Salesforce feature, teams can deliver AI-generated buyer journey summaries directly to Salesforce. Some common use cases for marketing and sales teams include:

  • Pushing buyer journeys for company that reaches an MQL stage
  • Pushing buyer journeys for each company that reaches a certain Engagement index, or is considered to be surging

By giving users the ability to generate a List containing an object and data related to that object, CaliberMind can create batches of calls to Google’s Vertex API and push those outputs to those specific IDs in Salesforce. The User Interface allows you to select an object, select a List and create a prompt.

This first iteration opens up the possibility to push any data to any object in Salesforce. As we work with customers to identify commonly requested Flows, we will work to make those standard Flows with a simplified User Experience. In the meantime, please work with your Customer Success Representative to push the data that is most important to your organization.

Prerequisite: Create a rich text and date timestamp fields in Salesforce to support the AI Summaries. Example:

  • Field 1: CM Buyer Journey – Long Rich Text
  • Field 2: CM Buyer Journey Last Update – Date Timestamp

Below you will find a walkthrough of this functionality.

  1. Navigate to the admin section of CaliberMind by clicking the cog icon in the left navigation menu
  2. Select the Flows > Create Flow section from the admin section
  3. Select the ‘Push Ask Cal to Salesforce’ flow
  4. Name your flow and click ‘OK’
  5. Complete the configuration page to create a new flow:
    • Select the Context List from the dropdown
      a. You can view a list preview with the eye icon, or link to the list configuration page with the link icon
      b. there are requirements listed for the list: it must have 2 columns – context_id (must be unique) and data_to_summarize
    • Select the Salesforce Object you would like to push this data to
    • Select the Context Column that will be updated
    • Enter the prompt for the AI summary to use
    • View the Estimated Credits section to make sure you have enough credits available to activate this flow
    • Choose if you want the flow to run on a schedule or just one time

SAMPLE LIST

SAMPLE PROMPT

You are a sophisticated marketing professional employed at a B2B company named, skilled at explaining complex marketing concepts using easy to understand language

You are tasked to explain the buyer journey of a B2B company using marketing event data

Use the following format for the summary, make sure it is in HTML. Do not output as JSON or Markdown.

A header that says "Buyer Group for" and then the company name
Just underneath in italics put "Last Run on:" and the value from query_run_time formatted as Year-Month-Day only

Buyer Group: a bulleted list of individuals most engaged along with their key interactions. Use Real names followed by email
Key Marketing Activities: a bulleted list summarizing each event by marketing activity type
Buyer Journey: a bulleted list of recent activity by person and event. Summarize at the month or quarter level
Do not include additional notes, recommendations or a conclusion

The marketing event data below, representing buyer journey events on a timeline for a Prospect company , contains the following columns and is in CSV format:
event_date: date of the event
company_name: The name of the company to summarize
person_name: The name of the person who performed the marketing event
email: The email of the person who performed the marketing event (when available)
job_level: The job level like VP or Director of the person who performed the marketing event
department: The department of the person who performed the marketing event
event_name: The name of the marketing touchpoint like sales email, paid social, etc..
event_class: Categorization of the class of event
event_type: Categorization of event type
campaign_name: The name of the marketing campaign
campaign_type: The type of Marketing Campaign
channel: The sourcing channel for the event of the event
event_count: The total number of this event that occurred on this date by this person
query_run_time: The timedatestamp of when the query was run in UTC

The output must be in Standard HTML format using tags for Headers, bold and italics. Do not start or end with ```html.

Related Articles

Table of Contents