AI Chat FAQ: what can I ask?

Wondering what to ask AI Chat — or what other teams are asking it? This page covers the most common ways marketers use the chat every day, with real prompt patterns you can adapt. Every example below is drawn from how teams actually use AI Chat in production.

🟦

Core AI · Included with your Simon platform.

New to the chat? Start with the Chats overview for the basics of the interface, then come back here for inspiration.

Can it build audiences and segments for me?

Yes — this is the single most common thing teams ask for. Describe any audience in plain English and the chat will size it, build it, refine it, or explain it back to you.

Try asking:

• "Create a segment of my top 10% of customers by lifetime value for an exclusive promotion"
• "Build a re-engagement segment of customers who purchased in the last year but have been inactive for 90+ days"
• "How large would a segment of customers who ordered in the last 2 years be?"
• "Add exclusion criteria to suppress anyone who hasn't opened or clicked recently"
• "Set up a 5% random holdout suppression segment across all email subscribers"
• "Explain what this existing segment does in simple, non-technical terms"

Popular patterns include winback and lapsed-buyer audiences, top-spender tiers, geo-radius targeting, engagement-based suppressions, and holdout groups for testing.

Can it report on campaign and journey performance?

Ask for performance analysis on demand — no dashboard building required. The chat pulls opens, clicks, conversions, and revenue, compares periods, and explains what changed and why.

Try asking:

• "Give me a 30-day performance analysis of this journey and rate it red, yellow, or green"
• "What was the open rate for all campaigns in Q1 this year versus Q1 last year?"
• "Show me my top five campaigns last month by engagement"
• "Which content block in this email gets the fewest clicks — what should I swap out?"
• "What's the best day and time to send, based on the last 30 days?"
• "How much revenue came from abandoned cart messages?"

📘

Make it a routine

Many teams run a standing weekly check-in — "rate my live journeys red, yellow, or green" or "look at my last 7 days and flag anything that looks off" — and let the chat surface what needs attention.

Can it answer quick questions about my data and account?

Yes. Counts, lookups, and searches that used to mean writing SQL or filing a support ticket are one question away.

Try asking:

• "How many contacts are opted into email and have opened in the last year?"
• "Which of my email templates mention this phrase or link?"
• "Has this email address received any emails from us? Which ones, and when?"
• "List all my segments with contact counts and whether each is used in a flow or journey"
• "Look up this contact and show their engagement history"

Any table the chat produces includes a Download CSV button, so results go straight to a spreadsheet.

Can it help me figure out how to do something in Simon?

Treat the chat as an always-on guide to the platform. It knows your account's setup, so answers are specific to you — not generic help-center articles.

Try asking:

• "How do I export this segment as a CSV to S3, and keep it refreshing?"
• "What's the difference between flows and journeys — which should I use here?"
• "Write the Jinja to format this date field in my email template"
• "How does the integration with my ESP work — what attributes are pushed?"
• "How do I upload a CSV and use it in segmentation?"

Can it investigate when a number looks wrong?

Yes — bring it anomalies. The chat digs into unexpected counts, delivery gaps, and mismatches between systems, and explains what it finds.

Try asking:

• "Why did we get an unexpected spike of contacts entering this journey on this date?"
• "My segment is much larger than the number of people receiving emails — why the gap?"
• "Why didn't this customer receive their birthday email?"
• "Are we double-mailing people? How often has a contact received the same email in 30 days?"
• "Identify bot clicks and show performance with them excluded"

Can it analyze my customers and find opportunities?

Ask analyst-style questions about customer behavior — lifetime value, retention, personas, purchase patterns — and get real analysis of your data, not canned reports.

Try asking:

• "Analyze my customer data and tell me where the biggest untapped growth opportunities are"
• "How does lifetime value compare across customers opted into email only, SMS only, and both?"
• "What's the average time between a customer's first and second order?"
• "Group my customers into personas based on spend and purchase behavior"
• "What are our retention, churn, and dormancy rates for last month?"

Because context carries across the conversation, you can go straight from an insight to action: find the opportunity, build the audience, then create the campaign — all in one chat.

Can it build campaigns, flows, and journeys?

Yes — beyond audiences, the chat builds the automation itself.

Try asking:

• "Build me a campaign to re-engage customers who purchased in the last year but have been inactive for 90+ days"
• "Create a weather-triggered campaign for customers with cold-weather affinity when cold weather is forecast locally"
• "Build a welcome series for new customers"
• "Convert this live flow into a draft journey"

Can it help with deliverability?

Ask about send volumes, bounces, and inbox placement.

Try asking:

• "How many emails bounced, were blocked, or went to spam on the last send?"
• "Build a segment of addresses with repeated 'mailbox full' bounces to suppress"
• "What's the optimal send cadence and time for my database?"

What should I ask first?

If you're staring at an empty chat, these starter prompts are the most popular first questions across Simon customers:

1. "Analyze my customer data and tell me where the biggest untapped growth opportunities are"
2. "Create a segment of my top 10% of customers by lifetime value"
3. "Build me a campaign to re-engage customers who purchased in the last year but have been inactive for 90+ days"
4. "Give me a 30-day performance analysis of [journey] and rate it red, yellow, or green"
5. "How many contacts are opted into email and engaged in the last 90 days?"

👍

Don't overthink the phrasing

You don't need perfect prompts. Describe what you want the way you'd explain it to a teammate — the chat asks clarifying questions when it needs more detail, and you can refine as you go.

Related features

• Chats — Overview of the chat interface
• AI Context — Give the chat persistent knowledge about your business
• Projects — Organize conversations around a campaign goal
• AI Personas — Explore and chat with behavioral customer clusters

Availability

AI Chat is available to all AI-enabled organizations. Capabilities are being expanded on an ongoing basis.