Lead Management APIs: Build a Scoring Process Your Sales Team Can Explain
A score should help a person decide what to do next, with the evidence and uncertainty visible.
Read the guideThe Selling API app suite
A little lab for
your next big idea.
Four focused tools to explore sales workflows. Test the inputs, inspect the output, and take a useful next step. Everything here runs in your browser.
Start with the example inputs. These tools calculate, format, and simulate locally; they do not call an AI model or connect to a CRM.
No account · no API keyMake a lead score explainable.
Explore how fit, intent, and relationship context contribute to one transparent score. This is a weighted rule, not a conversion prediction.
Less evidenceStronger evidence
How closely does this example match your intended customer profile?Less evidenceStronger evidence
How clearly has the buyer described a relevant need or next step?Less evidenceStronger evidence
How much useful, recent context does your team have?Formula: (fit × 0.4) + (intent × 0.4) + (relationship × 0.2), rounded to the nearest whole number. Suggested review bands: 80–100, 55–79, and 0–54. These demonstration thresholds are not validated business benchmarks.
Illustrative lead score
66out of 100
Nurture review
There is useful context here. Confirm the buyer’s need and choose a helpful follow-up.
30Fit points
26Intent points
10Relationship points
Give the conversation a useful brief.
Build a reusable prompt from a goal, an audience, and approved facts. This tool assembles instructions; it does not generate an AI answer.
Your reusable prompt
Review the prompt before using it with your chosen AI provider. Provider accounts and usage are separate.
Give a sales event a clear contract.
Build and inspect a small illustrative JSON event. This sandbox creates a local file; it does not send a webhook or call an endpoint.
Checks JSON syntax, an allowed event_type, a payload object, and a Boolean contact_allowed value. This is an illustrative contract, not a provider schema.
Your event payload
{
"schema_version": "1.0",
"event_id": "evt_example_001",
"event_type": "lead.created",
"mode": "local_example",
"payload": {
"lead_id": "lead_example_001",
"company": "Example Company",
"owner": "Sales team",
"contact_allowed": false
}
}Use the developer guide to plan identity, authentication, retries, and review for a production integration.
See where a conversation should go.
Try three predefined questions and inspect the response path. This is a scripted demonstration of grounding, uncertainty, and handoff.
Example sales assistantScripted demo
01 · Answer from known facts
Keep the product answer tied to the example’s limited scope. A useful assistant should be able to point to the information behind its answer.
02 · Recognize missing information
No price is provided in this example. The right next step is to acknowledge the gap and ask for an approved source.
03 · Handoff with context
Summarize the request and explain what a human would need. This local demonstration does not create a ticket or contact anyone.
Read the AI chatbot sales guide to explore retrieval, memory, account context, and evaluation.
Tool inputs and outputs stay in this page unless you copy or download them. Guide bookmarks and your animation preference may be saved on this device. Privacy details.
Add some context
Take the idea a little further.
Conversational Sales Prompts That Keep the Customer in Focus
Design prompts that help an assistant listen, explain, and suggest an appropriate next step.
Read the guide
AI Chatbot Sales APIs: From Product Questions to Useful Handoffs
Build a chatbot workflow that answers clearly, asks purposeful questions, and carries the right context to the sales team.
Read the guide