Reactive chat waits for customers to ask for help. Proactive chat shows up at exactly the moment a customer is most likely to need it - and most likely to leave if no one reaches out. The difference in conversion rate between the two approaches is not marginal. It's transformational.
But proactive chat done badly is worse than no proactive chat at all. An intrusive pop-up three seconds after page load isn't helpful - it's harassment. The scripts below are designed around three specific signals that indicate a customer is genuinely hesitating, and the timing rules that make intervention feel helpful rather than pushy.
The 3 Signals That Tell You to Fire a Proactive Chat
Not every visitor needs a proactive message. Firing at everyone degrades the experience. The three signals that reliably predict hesitation:
- Time on page exceeding threshold. A customer who has been on a product or pricing page for 90+ seconds without scrolling to the bottom or clicking a CTA is reading, re-reading, and hesitating. This is not casual browsing - it's active deliberation.
- Cart abandonment signal. Cart value above a threshold combined with inactivity (mouse not moving, no scroll, no clicks) for 45+ seconds suggests the customer has stalled. They have intent - something is stopping them.
- Exit intent. Mouse cursor moving toward the browser's address bar or close button. This is the last moment to intervene before the session ends. High urgency, low time budget.
The 5 Scripts
Cart Abandonment - Stalled Checkout
Trigger: cart value > $50, 45s inactivity📈 A/B tested over 2,400 sessions: 31% response rate, 22% conversion lift vs. no intervention
Exit Intent - Before They Leave
Trigger: exit intent detected📈 27% response rate on exit intent. Top responses: pricing questions (38%), shipping cost (24%), sizing/compatibility (19%). All answerable in under 2 minutes.
Pricing Page - Comparison Hesitation
Trigger: 90s on pricing page, no plan clicked📈 34% response rate. Most common outcome: direct recommendation leads to plan selection without further deliberation. Avg session value 2.3x higher when chat is engaged.
VIP Return Visitor - Warm Welcome Back
Trigger: returning visitor, 2+ prior sessions, no purchase📈 41% response rate from return visitors. High-intent segment: these visitors are 3.2x more likely to convert than first-time visitors when engaged by chat.
Post-Purchase - Upsell & Delight
Trigger: order confirmation page, 20s time on page📈 Post-purchase proactive chat drives 18% average order value increase when the upsell is genuinely relevant. Key: the suggestion must be pre-configured with actual product logic, not generic.
Velaro's proactive chat triggers let you set these rules in minutes - no code required.
Set up proactive chatWhat NOT to Say: The 3 Openings That Drive People Away
Avoid These Openings
- "Hi! Can I help you?" - The weakest opener in proactive chat. It's a yes/no question with an obvious no as the default answer. 78% of people dismiss this immediately.
- "I noticed you've been on our site for a while!" - This feels surveillance-y. Customers don't want to be told they've been watched. State what you can do for them, not what you know about their behavior.
- "Don't leave! We have a special offer just for you!" - On exit intent, this reads as desperate. It signals your product isn't worth full price and trains customers to always exit-intent to get a discount.
How to Set Up These Rules in Velaro
- 1 In Velaro, go to Engagement Rules → Proactive Chat Rules → New Rule.
- 2 Set the URL condition (e.g., URL contains "/cart" or "/pricing") and the trigger event (time on page, exit intent, inactivity, return visit).
- 3 Paste the message copy for that rule. You can set different messages per rule - or A/B test two variants.
- 4 Set the agent assignment: which team or agent skill receives these chats. Proactive chats should go to your highest-performing sales-oriented agents.
- 5 Set a session cap: a visitor should see at most one proactive prompt per session. Multiple prompts train visitors to ignore them.
"The best proactive chat message is one that makes the customer think 'that's exactly what I needed' rather than 'why is this popping up on me.'"
Measuring What's Working
Once your rules are live, Velaro's analytics break down each proactive rule by:
- Trigger rate: How often the rule fires (per 1,000 sessions)
- Response rate: % of customers who respond to the proactive message
- Conversion rate: % of proactive chats that lead to a purchase or desired action
- Revenue influenced: Total order value from sessions where proactive chat was engaged
Run each rule for at least 200 triggers before making changes. The data variance below 200 is too noisy to draw conclusions. After 200, compare response rate and conversion rate across your variants and keep the winner.
The Bottom Line
The best proactive chat scripts are short, specific, and triggered at the right moment. They acknowledge where the visitor is in their journey, offer something concrete, and leave the door open without pressure. Generic scripts underperform because they apply the wrong message to the wrong visitor at the wrong time. Write page-specific scripts, test them in volume, and let the data tell you what's working.
Frequently Asked Questions
What should I say in a proactive chat message?
Effective proactive chat messages are short (under 20 words), page-specific, and offer something concrete. Reference what the visitor is likely evaluating - "Comparing plans? I can walk you through the differences." - rather than a generic greeting. Avoid questions that require long answers. A simple, open-ended offer works better than a complex question.
What are good proactive chat examples?
High-performing proactive chat examples by page type: Pricing page - "Questions about which plan is right for you? Happy to help." Cart page - "Need help completing your order? I'm here." Product page - "Want to see how this works for teams your size?" Exit intent - "Before you go - can I answer any questions?" Each message is under 15 words.
How do you write a proactive chat script?
To write a proactive chat script: identify the page and what a visitor there is likely thinking, write a one-sentence message that acknowledges that context, include a soft offer ("happy to help" or "I can walk you through it"), keep it under 20 words, and avoid yes/no questions. Write 2-3 variants and A/B test them after 200+ triggers.
What makes a good proactive chat greeting?
A good proactive chat greeting is specific to context, not generic. It avoids "Hi! How can I help?" - which has the lowest response rates. It references the visitor's likely intent ("Comparing options?", "Ready to get started?"), keeps the ask low-pressure, and arrives at the right moment (after sufficient time-on-page, not immediately on load).
What proactive chat messages have the highest conversion?
The highest-converting proactive chat messages combine context-awareness with a concrete offer. On pricing pages, messages that reference plan comparison outperform generic greetings by 3-5x. On cart pages, messages that reference the specific product in cart convert significantly better than generic "need help?" prompts. Page-specific scripts consistently outperform site-wide default scripts.