SiftDocs
Getting Started

Your First Conversation

Test your AI agent in the Playground and understand lead qualification

Your First Conversation

The Playground is where you test your agent before deploying it to real leads. It simulates a full conversation with the same AI engine that powers the live widget.

Open the Playground

  1. Go to your dashboard
  2. Click Playground in the sidebar

You'll see two panels:

  • Left: Agent configuration (name, personality, greetings)
  • Right: Chat simulator where you play the role of a lead
Screenshot of the Playground showing the config panel on the left and chat interface on the right

Try a Sample Conversation

Try a message that a real lead would send:

  • "Hi, I'm looking for a 3-bed semi in south Manchester, budget around £350k"
  • "We're first-time buyers, got our DIP last week, looking for something with a garden"
  • "I need to move by September — what 2-beds do you have under £250,000 in M20?"
  • "We're chain-free, sold STC — looking for a 4-bed detached, flexible on location"

What to Watch For

As you chat, notice how the agent:

  1. Greets you with the message you configured during onboarding
  2. Asks qualifying questions — budget, timeline, location, property type, number of bedrooms
  3. Asks about UK-specific signals — chain status, DIP/mortgage, buyer scheme eligibility
  4. Searches your properties and suggests matching listings
  5. Scores the lead — you'll see scoring in the lead details once the conversation is saved

Understanding Lead Scores

After a conversation, Sift assigns a score:

ScoreMeaningVisual
🔥 HotHigh intent — budget confirmed, timeline clear, ready to viewRed badge
🟡 WarmInterested but missing signals — maybe no budget or vague timelineYellow badge
🔵 ColdLow intent — browsing, no specifics, or disengagedBlue badge

The score is based on what the lead reveals during conversation. More details at Lead Scoring.

Troubleshooting

Next Steps

On this page