Salesforce has long been a leader in CRM, helping businesses manage customer relationships and streamline operations. But the real magic happens when Salesforce becomes a hub for enriched, actionable data. By connecting your data warehouse to Salesforce with Fivetran Activations, you can supercharge your CRM workflows with AI-powered insights. With AI Columns, you can go a step further, using AI to create dynamic, context-aware insights directly within Salesforce.

Below are some use cases and AI column prompts designed to make Salesforce an even more powerful tool for your teams. The field names are represented in a format for ease of use, and the prompts should be customized to fit your specific data model. Fivetran Activations allows you to easily model your customer data as Datasets, ready for activation.

[CTA_MODULE]

Use case 1: Automating fit scoring in Salesforce 

Building an Ideal Customer Profile (ICP) analysis and account scoring system can be a daunting task for RevOps teams, often involving extensive data analysis and consensus-building among sales leaders. However, with AI Columns, this process can be streamlined significantly.

Step 1: Set up your ICP analysis

Begin by creating an ICP analysis the AI Column prompt that consolidates customer traits, success metrics, and account data. This analysis will help identify attributes that predict high-value outcomes for your business.

• Define key attributes: Determine characteristics meaningful for your ICP, such as high lifetime value (LTV), short sales cycles, industry fit, or specific investor backing.
• Create a JSON array: Aggregate this data into a JSON format, capturing all relevant information in a single field per customer.

This ICP analysis serves as the benchmark against which new accounts will be scored.

Step 2: Write your fit score or ICP analysis GPT prompt

With your ICP analysis ready, craft a GPT prompt to evaluate each account's fit score based on your criteria.

• Customize success indicators: Include attributes that align with your ICP, like revenue thresholds, employee count, or specific technologies, and assign weights based on their correlation with success.
• Generate a fit score: Establish a scoring system (e.g., 0-10) with explanations for each score to ensure transparency for the sales team.

Using `{ {record['CUSTOMER_TRAITS'] } }`, perform an Ideal Customer Profile (ICP) analysis using regression. List positive and negative traits with their correlation strengths (strong, moderate, weak), noting data gaps without assuming negatives. Identify valuable traits exclusively based on regression analysis, with category weights summing to 100% for LLM scoring.

Use these metrics to assess success, loyalty, and conversion:
ARR: Annual recurring revenue, indicating customer value
ever_customer: If the account has been a customer
is_customer: Current active customer status
LTV: Expected revenue per customer
months_as_customer: Duration as an active customer
weeks_to_convert: Time from first interaction to conversion

Categories (with weights)
1. Technology (30%): Identify high/medium confidence tech with success correlations.
2. Industry (15%): Highlight industries with positive/negative success correlations.
3. Annual Revenue Range (10%): List revenue ranges correlating with success.
4. Employee Size Range (15%): Specify employee count ranges with success alignment.
5. Geographic Fit (10%): Identify country or regional correlations.
6. Investors (10%): Highlight investor partnerships with positive or negative impacts.
7. Hiring Trends (10%): Analyze relevant job titles and hiring volume; exclude negative indicators.

Final Output: Structure findings with headings, correlation strengths, and category weights. Note data gaps and additional traits from regression analysis for scoring and lead comparison.

Utilizing GPT Columns, you'll receive an ICP analysis output that ranks each account against these success indicators, creating a tailored ideal customer profile.

Step 3: Test the account fit score with familiar accounts

Testing validates the accuracy of your scoring.

• Select a sample set: Choose a handful of known accounts to evaluate, allowing you to quickly assess if the scores align with expectations.
• Fine-tune as needed: Adjust the prompt based on test results until the fit scores consistently match your ICP benchmarks.

Step 4: Score new accounts and customize the prompt for Sales

After testing, you're ready to score new accounts.

• Run your prompt: The prompt will evaluate each new account, comparing it to your ICP benchmark.
• Provide clear explanations: Use a "stoplight" system (e.g., green for good fit, yellow for moderate, red for low) to help the Sales team interpret scores at a glance.
• Add sample summaries: Include example summaries to demonstrate how the fit score output should look, enabling the sales team to quickly grasp key insights.

By following these steps, you can build an automated ICP analysis and account scoring system in a single afternoon, transforming weeks of work into actionable sales insights.

Use case 2: Building an AI product usage summary with AI Columns

Understanding and utilizing product usage trends is critical for empowering GTM teams with actionable insights. Fivetran Activations allows you to integrate product event data and build comprehensive summaries to enhance decision-making in Salesforce.

Step 1: Build an Activity Schema Dataset

To get started, consolidate your product event data into a single schema for analysis.

• Define the scope: Identify key product usage events to track, such as login frequency, feature adoption, or concerning inactivity signals.
• Create a JSON array: Create a dataset to build an activity schema where events for each account are aggregated into a JSON array. This allows you to store all activity data compactly in Salesforce for seamless access.

Example:

SELECT   

account_id,   

JSON_AGG(    

JSON_BUILD_OBJECT(      

'event_type', event_type,       

'event_time', event_time,       

'event_value', event_value    

)  

) AS activity_summary 

FROM product_events 

GROUP BY account_id;

Step 2: Write a GPT prompt to summarize usage trends

Once your activity data is organized, use GPT to analyze and summarize product usage trends for GTM teams.

Key Insights to Include:

• Compare current 30-day trends to the previous 30 days.
• Highlight positive, negative, and exploratory behaviors.
• Suggest actionable next steps based on event patterns.

Sample GPT Prompt:

Analyze the following product event data for account :  

- Summarize product usage trends over the past 30 days compared to the prior 30 days.  
- Identify key patterns: positive trends (e.g., increased feature adoption), negative trends (e.g., reduced usage), and exploratory behaviors (e.g., testing new features).  
- Suggest 3 actionable next steps for the sales team.  
- Present findings in this structure:  

1. **Key Usage Trends:** Summary of major trends.  
2. **Concerning Events:** Highlights of potential risks.  
3. **Actionable Next Steps:** Tailored recommendations for GTM teams.  

Step 3: Structure data for Salesforce

With your analysis complete, ensure that the insights are formatted for easy consumption by the sales team in Salesforce.

• Organize outputs: Align the output structure with your Salesforce fields, such as “Key Events,” “Concerning Trends,” and “Next Steps.”
• Include suggested sales talking points: Provide GTM teams with ready-made talking points based on product usage insights to personalize their outreach.

Step 4: Validate and Iterate

• Test outputs: Review sample accounts to ensure insights align with expectations.
• Fine-tune prompts and data: Adjust the activity schema and prompt logic as needed to maximize relevance and usability.

By centralizing and summarizing product usage data, you can empower your GTM teams with actionable insights directly in Salesforce. This approach ensures that key product events are not just tracked but actively leveraged to drive impactful sales conversations and customer engagement strategies.

Leveraging AI Columns to empower Salesforce teams

Incorporating AI Columns into your go-to-market strategy empowers sales and marketing teams with actionable, AI-driven insights tailored to their unique needs. By enriching Salesforce with data from your warehouse, you can guide account executives toward high-value opportunities with personalized recommendations, prevent churn with proactive customer retention alerts, and craft hyper-personalized marketing campaigns that drive engagement. 

Whether it's helping sales reps identify the next best action, providing account managers with re-engagement strategies, or delivering marketing teams tailored content suggestions, these AI-powered prompts unlock new levels of efficiency and effectiveness. Start leveraging these strategies today to drive revenue, retain customers, and create impactful campaigns that resonate with your audience.

[CTA_MODULE]