Guide

What nobody tells you about conversion lift from chat

Small business owners often see chat traffic rise but sales stay flat—here's the gap and how to fix it

Key takeaways
  • Most small businesses see a 10–20% lift in conversion when chat is staffed during peak hours, not just 'installed'.
  • Response time under 5 minutes turns browsers into buyers more often than any slick widget feature.
  • Human handoff matters: leads captured by chat but never followed up on drop off by 60% or more.
  • Proactive chat triggers outperform reactive 'contact us' forms by roughly 2-to-1 in our data.
  • Integrating chat with your CRM is where the real lift happens—without it, every conversation is a loose end.
VoxyChat · Key takeaways1
Most small businesses see a 10–20% lift in conversion when chat is staffed during peak hours, not just 'installed'.
2
Response time under 5 minutes turns browsers into buyers more often than any slick widget feature.
3
Human handoff matters: leads captured by chat but never followed up on drop off by 60% or more.
4
Proactive chat triggers outperform reactive 'contact us' forms by roughly 2-to-1 in our data.

Opening the conversation with a number-backed claim feels right. Most teams we see install chat, wait for the phone to ring, and end up disappointed. The widget alone doesn’t move the needle—what happens after the first greeting does. I’ve watched this pattern repeat across two dozen sites we’ve outfitted with AI chat, and the difference between a vanity metric and a real lift usually comes down to three things: who’s answering, how fast they answer, and what happens to the conversation after it ends. If you’ve tried chat before and felt it underdelivered, you’re not alone. The good news is that the fix is usually cheaper and simpler than starting over.

The installation myth nobody warns you about

We’ve all been there. You pay a developer, drop a snippet of JavaScript into the footer, and suddenly there’s a chat bubble sitting in the corner. It looks the part. But a live conversation is not a static asset like a banner ad or a contact form. It’s a live channel, and without a plan for who fielding those messages and how, it becomes another unmonitored inbox. I’ve seen businesses get excited about a 15% bump in chat volume, only to realize half the messages sit unread for 24 hours. That’s not a conversion tool at that point—it’s a brand risk. The installation is the easy part. The real work starts when the first prospect types “hello” and nobody’s there to reply.

I remember a neighborhood appliance repair shop we onboarded last spring. They had a fresh website, the chat bubble glowing bright green in the corner, and optimism radiating from the owner. He’d invested in a premium chat platform, told me he “set it and forgot it,” and was puzzled when the monthly reports showed dozens of conversations logged but zero bookings. The bubble was there. The messages were coming in. But nobody was answering them during the hours when people were actually home from work and thinking about a broken dishwasher. The chat became a digital ghost town. Prospective customers would type their issue, get no response, and leave the site feeling like the business had already closed for the day. That’s the danger of treating installation as the finish line. The bubble is just a mouth with no voice behind it.

What complicates things is the mistaken belief that simply having a chat presence signals approachability. In reality, an ignored chat bubble can be worse for conversion than having no chat at all. Visitors notice the discrepancy. They expect a response. When they don’t get one, the silence reads as neglect. I’ve had clients tell me that after a bad experience with a dead chat widget, they were more skeptical of the business’s legitimacy. The takeaway here is brutal but necessary: a chat widget without a staffing plan is a liability, not an asset. The technology is inert until someone decides to pick it up.

Staffing models that actually convert

Here’s what our data keeps showing: conversion lift correlates more strongly with staffing coverage than with any fancy AI prompt engineering. Most small businesses we work with don’t need a full-time chat operator. They need predictable coverage during the hours when visitors are actually browsing. For a local retail client, that meant two hours in the evening after work. For a B2B service site, it was three blocks of time across the workweek. The pattern is consistent: if someone initiates a chat and gets a response within five minutes, they’re roughly twice as likely to move to a lead stage than someone who waits fifteen or twenty. That five-minute window is the single biggest predictor we’ve found. It doesn’t require a round-the-clock team—it requires a schedule that matches your traffic peaks.

Let me break down how we approach this with the businesses we manage. First, we look at the analytics. Not just total visits, but where the traffic spikes happen. For a coffee shop we work with, the data showed a distinct surge between 4:00 PM and 6:00 PM—people finishing work, browsing on their commute home, deciding where to stop on the way home. That two-hour window became our staffing block. For an accounting firm, the peak was Tuesday and Thursday mornings, 9:00 AM to 11:00 AM, when small business owners were reviewing weekly reports and had questions about quarterly estimates. The key is aligning human availability with visitor intent.

Once the window is identified, the next step is setting the expectation. We advise teams to treat that block like a scheduled appointment. It goes on the calendar. If the designated person is unavailable, there’s a backup plan—whether that’s a colleague stepping in or the chat shifting to an AI-only mode with clear limitations communicated upfront. The goal is never to let a conversation languish. We’ve found that the “five-minute rule” holds up across industries. It’s the tipping point where a browser becomes a buyer. Beyond five minutes, intent begins to erode. At fifteen minutes, many visitors have already moved on to a competitor who’s faster to respond.

What about off-hours? This is where many small businesses trip up. They want chat available 24/7, but they don’t have the staff for it. Our recommendation is to be honest about capacity. If you can’t staff 24/7, don’t display the chat widget outside your operating hours, or configure it to show a “leave a message” form instead of a live conversation. There’s nothing more frustrating than typing a question at 11:00 PM and getting an automated “we’ll get back to you” that never materializes. Transparency beats ghosting every time. Some of our most successful clients actually disable the widget after 6:00 PM and re-enable it at 8:00 AM, ensuring that every single conversation that starts during open hours gets a real human response within the five-minute window. That focused approach consistently outperforms a scattered, over-extended attempt at round-the-clock coverage.

The handoff problem that kills momentum

This is where a lot of well-intentioned chat projects stall. A prospect chats for ten minutes, asks about pricing, gets a helpful answer from an AI, and then… nothing. The conversation ends, the context disappears, and the prospect has to start over on a contact form or worse, lose interest entirely. We’ve tracked this enough times to know the numbers: leads captured by chat but never followed up on drop off by sixty percent or more. The lift vanishes because the loop stays open. If you’re using AI to qualify a lead, you need a clear path to hand that qualified conversation off to a human who can take the next step—whether that’s scheduling a call, sending a proposal, or updating a CRM record. Without that handoff, every chat is a dead end, and your conversion rate will reflect that.

I want to walk through a specific example that illustrates this perfectly. We worked with a boutique fitness studio that wanted to use chat to book trial classes. Their AI did a great job of answering basic questions—membership tiers, class schedules, pricing. A prospective member would start a chat at 10:00 AM, the AI would explain the unlimited pass option, and the prospect would express interest in signing up for a Wednesday class. The AI would confirm the class time and ask for an email address to send a confirmation. Perfect, right? Except the AI didn’t have the authority to actually book the class. It just stored the email in a chat log. Twenty-four hours later, the prospect hadn’t received a confirmation, hadn’t heard back, and had assumed the studio wasn’t serious about onboarding new members. They never returned.

The fix was simple but required a change in workflow. We integrated their chat platform with their booking system so that when the AI flagged a qualified lead—someone ready to book—the system automatically created a pending reservation in their studio software and sent an email notification to the front desk. The front desk then sent a personalized confirmation email within the hour. The conversation that once went cold now resulted in a booked trial class 90% of the time. The handoff from AI to human, facilitated by integrated tools, was the difference between a missed opportunity and a new member. That’s the power of a defined exit point for every conversation.

What makes this tricky is that many small businesses use chat as a support tool rather than a sales tool. The conversation ends when the question is answered. But if your goal is conversion, the conversation needs an exit strategy. What happens next? Do you send a follow-up email? Does the lead get added to a nurture sequence? Is a calendar link generated? We advise our clients to define the “next step” before the chat even begins. If a prospect is interested in pricing, the goal might be to get them to request a quote. If they’re asking about services, the goal might be to schedule a discovery call. Whatever the objective, the chat flow should be designed to deliver the lead to that outcome, not just resolve the immediate inquiry. When the handoff is an afterthought, the momentum dies. When it’s built into the flow, conversion follows.

Proactive vs reactive: which actually moves the needle

One of the first questions we get is whether to set up proactive triggers—those “can I help you?” messages that pop up after a visitor spends thirty seconds on a page—or to let customers initiate. Both have their place, but our benchmarks show proactive triggers roughly double the engagement rate of passive contact forms. The key is relevance. Triggering based on page intent—someone on a pricing page for more than forty-five seconds gets a message—converts at about twice the rate of random triggers. Reactive chat, where the visitor starts the conversation, tends to have higher intent already, but fewer of them because the bubble is easy to ignore. A blended approach works best: proactive triggers to start the conversation at the right moment, then seamless hand-off to reactive dialogue once they’ve engaged.

Let me give you a concrete breakdown of how we structure this for different types of sites. For an e-commerce client selling handmade jewelry, we set a proactive trigger for anyone who spends more than sixty seconds on the “About” page and then navigates to a product page. The message that pops up is friendly and specific: “Loving the [ specific product ]? I can tell you more about the materials and shipping time if you’d like.” Because the trigger is tied to specific behavior, the conversion rate on those triggered conversations is about 3.5%, compared to 1.2% for visitors who initiate on their own. The proactive message doesn’t feel like a generic sales pitch; it feels like a helpful assistant noticing their interest.

For a B2B software site, the strategy looks different. We use proactive triggers on the pricing page specifically. If a visitor hovers on a pricing tier for more than forty-five seconds, the chat bubble appears with a message like: “Need help deciding which plan fits your team size? I can walk you through the differences.” Because the intent is already there—they’re looking at prices—the conversation quickly moves to a qualification stage. These triggered chats have a much higher handoff rate to sales calls because the visitor is already in buying mode. Reactive chats on the other hand, while high-intent, often get lost in the noise of a busy website. The visitor might start a chat, get distracted by another tab, and never follow through. Proactive triggers capture attention at the moment of interest, but they have to be right. Random triggers, or triggers that fire too aggressively, feel desperate and drive people away. We’ve seen bounce rates increase when a chat pops up immediately on landing, before the visitor has a chance to understand what the site is about. Timing and context are everything.

The blended approach we recommend starts with proactive triggers to lower the barrier to entry. It’s easier for a visitor to click “yes” or type a quick question than to seek out a contact form. Once they’re engaged, the conversation naturally shifts to a more reactive mode—they’re driving the conversation now, asking their own questions, and the human or AI response feels like a continuation rather than a push. This flow—proactive initiation, reactive engagement—maximizes the number of conversations that start while keeping the quality high enough to convert. It’s the sweet spot between being invisible and being pushy.

Integrating chat with your stack

If you’re not connecting chat data to the rest of your business tools, you’re leaving lift on the table. I’ve seen it time and again: a chatbot gathers a prospect’s budget range, decision timeline, and specific use case, but if that information never makes it into the sales CRM, it’s just a note in the wind. Integration doesn’t have to be expensive or custom-built. Many chat platforms offer native connections to HubSpot, Salesforce, or even Google Sheets. The goal is to capture the conversation context so the next person who talks to that lead knows where they left off. That continuity is what turns a one-off chat into a pipeline entry. Without it, every conversation starts from scratch, and your conversion costs creep up.

I had a client—a home cleaning service—who was manually typing key details from every chat into a spreadsheet. It was tedious, and after a few weeks, the spreadsheet was abandoned, and the context was lost. When I asked why they weren’t using the integration features already built into their chat platform, they didn’t even know it was possible. Their platform had a two-click connection to their Google Calendar and a Zapier link to their CRM. We set it up in an afternoon. Now, when a chat ends with a qualified lead, the system automatically creates a new contact in the CRM, tags them as “chat-sourced,” and populates the initial notes with the budget and timeline discussed. The next time a salesperson reaches out, they can see exactly what was discussed. No re-introduction needed. That single integration increased their booking rate from chat-sourced leads by about 25% because the follow-up was timely and informed.

The technical barrier is lower than most people assume. If you’re using a modern chat platform, look for the “integrations” or “connectors” section in your settings. Many offer drag-and-drop connectors for popular CRMs like HubSpot or Mailchimp. If you’re on a tighter budget or using a simpler tool, Zapier or Make (formerly Integromat) can bridge the gap between your chat widget and your existing software for a modest monthly fee. The key is mapping the data you actually need. You don’t need to send every single chat transcript. Just the essentials: visitor name, email, the topic they asked about, and any qualification criteria you use (budget, timeline, decision maker). That minimal data set is enough to make the next conversation productive. When the context travels with the lead, the sales cycle shortens, and the conversion rate climbs.

Measuring what actually matters

Vanity metrics will mislead you. Total chat volume means nothing if none of those chats turn into anything. What we tell clients to watch are three numbers: conversation-to-lead conversion rate (what percentage of chats result in a qualified lead or booking), average response time (keep this under five minutes if you can), and qualified handoff rate (of those leads, how many get a follow-up action). Those three give you a clear picture of whether chat is contributing to your bottom line or just adding noise. If any one of those is trending downward, that’s your signal to adjust staffing, refine triggers, or tighten the integration before the problem spreads.

Let’s look at how to interpret these numbers in practice. Imagine a home services business—let’s say an HVAC company. They start tracking their metrics and find that their conversation-to-lead conversion rate is sitting at 8%. That’s decent, but they want more. They dig into the data and notice that whenever average response time creeps above seven minutes, that conversion rate drops to 5%. When they get it back under five minutes, it bounces back up to 9%. The takeaway? Response time is a leading indicator. They don’t need to wait for the next quarterly review to make a change; they can adjust staffing schedules weekly based on that metric alone.

Qualified handoff rate is where many businesses see the biggest gap. They might have a 10% conversation-to-lead conversion rate, but only 40% of those leads ever get a follow-up call or email. That means 60% of the potential business generated by chat is evaporating because nobody’s reaching out. That’s a signal that the integration or the handoff process needs work. Maybe the AI is qualifying leads but not flagging them for human review. Maybe the team is too busy with phone calls to follow up on chat leads. Whatever the cause, it’s a solvable problem. By zeroing in on these three metrics—conversion rate, response time, and handoff rate—you stop guessing and start making data-driven adjustments that actually move the needle.

Common mistakes we’ve seen (and how to avoid them)

Installing chat and walking away. That’s the big one. If there’s no plan for who answers and within what timeframe, the widget becomes a liability rather than an asset.

Random proactive triggers. Popping up a message on every page, or worse, at the wrong time (like right after a user clicks “add to cart”), feels desperate and drives people away. Context matters.

Treating chat like a ticket system. If your only goal is to resolve the inquiry and close the thread, you’ll miss the sales opportunity. Chat is a conversation, not a ticket.

Ignoring the handoff. We said it before, but it’s worth repeating: if a lead leaves the chat and has to re-explain themselves to a human, that momentum is dead.

Skipping the follow-up. A chat that ends without a next step—a calendar link, a proposal, a discount code—is a missed opportunity. Every conversation should have a clear outcome.

These are the pitfalls that trip up even well-meaning businesses. Let me elaborate on a couple of them because the consequences are often underestimated. “Installing chat and walking away” is the most common, and it’s exactly what it sounds like. A business pays for the service, the developer sets it up, and then nobody owns the inbox. Messages pile up. The team assumes someone else is checking them. Eventually, the chat gets disabled out of embarrassment, or it just sits there, dormant, doing nothing but adding a line to the monthly bill. The fix is simple: assign ownership before you go live. One person, even if it’s only for two hours a day, should be the designated chat holder. That ownership creates accountability.

“Treating chat like a ticket system” is another subtle but costly mistake. In a support context, the goal is indeed to resolve the issue and close the ticket. But in a sales context, the goal is to move the person forward. If a chat ends with “Your question has been answered, thank you,” and no next step is offered, you’ve wasted the opportunity. The conversation should feel like the beginning of a relationship, not the end of a support case. We recommend ending every chat with a proposed next step. “I’m glad I could clarify the pricing for you. Would you like me to send a formal quote, or do you have any other questions before you decide?” That small pivot keeps the door open and makes it easy for the prospect to say yes to the next thing.

Comparing chat lift to other low-cost channels

How does chat stack up against a Google Ads campaign or a monthly newsletter? In our experience, the cost-per-lead from well-staffed chat is usually lower, especially for service-based businesses where the sale happens through conversation rather than a checkout cart. Ads bring volume, but chat brings qualification. A newsletter nurtures over time, but chat can accelerate a decision in the moment. The real differentiator is speed: most businesses can’t match a five-minute response time with an email sequence or an ad landing page. That immediacy is where the lift comes from, and it doesn’t require a big budget—just a committed few hours a week and the right integration in place.

Let’s put some numbers to this comparison, because that’s usually what business owners want to see. A typical Google Ads campaign for a local service business might cost $50 to $150 per lead, depending on the industry and competition. Those leads are often cold—they clicked an ad, maybe filled out a form, but there’s no conversation happening. Now compare that to a well-staffed chat setup. If a team member is covering two peak hours a day, five days a week, and they’re averaging three qualified conversations per hour with a 10% conversion rate to a booking or quote request, that’s 30 qualified conversations per week. If the cost of that staff time is factored in—let’s say it’s the equivalent of $30 per hour in labor cost—you’re looking at about $150 in labor costs for 30 leads, or $5 per lead. Even if you factor in the chat platform subscription, you’re still well under the cost per lead of Ads. And those chat leads are warmer. They’ve had a real conversation. They’ve had their questions answered. They’re further along in the buying journey.

Newsletters are great for long-term relationship building, but they don’t create the same immediate lift. A subscriber might open a newsletter week after week before they’re ready to buy. Chat cuts through that timeline. It’s a “right now” channel. For a business that needs pipeline filled this month, chat is the more efficient use of a limited marketing budget. The key is that it has to be staffed and integrated. An understaffed, unintegrated chat channel is actually more expensive than Ads because you’re paying for the tool and the labor without getting the conversion results. But when it’s done right, it’s a high-ROI channel that competes favorably with anything else you’re running.

The quiet truth about AI and conversion

There’s a lot of hype around AI chat these days, and honestly, some of it is warranted. A well-trained bot can handle the fifty percent of questions that are routine—hours of operation, basic pricing, appointment availability—freeing up humans for the nuanced conversations that actually close deals. But AI alone won’t move the needle if there’s no one on the other end to close. We’ve seen sites where the bot does great work, but because no human ever reviews the qualified leads the bot surfaces, those leads go cold. The sweet spot is AI for scale and speed, human for qualification and closure. When that balance clicks, conversion lift follows.

I want to be specific about what AI should and shouldn’t do in a conversion-focused setup. On the routine side, AI should absolutely handle the repetitive questions that eat up human time. “What are your hours?” “Do you service [zip code]?” “What’s the price for the basic package?” These are the questions that, if answered by a human, take five minutes of their time and add no strategic value. AI can answer them in five seconds, 24/7, without getting tired. That’s a win. But here’s the catch: the AI needs to be configured to recognize when a conversation has moved from routine to qualified. It needs to flag the conversation for human review when the prospect asks about pricing specifics, mentions a timeline, or expresses interest in moving forward. If the AI answers those questions but doesn’t hand them off, you’re back to the handoff problem we discussed earlier.

On the human side, the role shifts to qualification and closure. A human should review the flagged leads, respond within the five-minute window if the prospect is still online, or send a follow-up email within the hour if they’re not. The human adds the nuance—the ability to read between the lines, address objections, and ask the right questions to uncover the real need. That’s where deals get closed. I’ve seen businesses that went all-in on AI, thinking it would replace their sales team, and ended up with a bunch of automated conversations that never resulted in a booking. The AI was efficient, but there was no one to close. The sweet spot, based on our experience, is a tiered approach: AI handles the first line, filters the noise, and qualifies. Humans handle the qualified leads, the conversations that have actual buying potential. When that division of labor is clear and the tools are connected, conversion lift follows naturally.

One change you can make this week

If you’re ready to see whether chat can move the needle for your site, start small. Pick one peak hour block during your busiest traffic window. Assign one team member to cover it, set a five-minute response goal, and connect the chat platform to whatever CRM you already use. Track the conversation-to-lead conversion rate for that window over two weeks. If it moves, you’ve found a lever. If not, you’ve lost very little time and can adjust the next block. The mistake most people make is trying to boil the ocean—staffing 24/7, building complex trigger logic, integrating every tool under the sun before they’ve proven the basics work. Start with one focused window, get the data, then scale what works.

This is the practical advice we give to every client who feels overwhelmed by the prospect of a full chat implementation. It’s tempting to want to have the perfect setup from day one—24/7 coverage, AI handling everything, perfect CRM integration, beautiful proactive triggers everywhere. But that’s a recipe for paralysis. The businesses that see the fastest results are the ones that start with a single, manageable change. Pick your busiest hour. Maybe it’s Monday morning from 9:00 to 10:00 AM, or Thursday evening from 5:00 to 7:00 PM. Get one person scheduled in that slot. Give them the mandate to respond to every chat within five minutes. Connect the chat to your CRM so at least the lead info travels through. Run that for two weeks. Look at the numbers. Did conversion go up? Even a small lift—5% or 10%—is proof of concept. That’s your signal to add another block, or to optimize the one you have. If it didn’t move, you haven’t lost much. You’ve just learned what doesn’t work for your specific traffic patterns, and you can try a different window or adjust the response process.

The “boil the ocean” mistake is so common because it feels like the only way to get real results. But in our experience, the compound effect of small, focused improvements always beats the all-at-once approach. You wouldn’t try to redesign an entire website in a single day. You’d test one page, see if the conversion rate improves, then move to the next. Chat should be treated the same way. Start small, prove the model, then scale. That’s how you build a chat channel that actually drives revenue, rather than just adding a noise feature to your site.


Conclusion

I’ve been in the trenches with enough businesses now to know that chat isn’t a magic button. It’s a channel, and like any channel, its value depends on what you do with it. The businesses that see real conversion lift are the ones who treat it as a priority—staffing it, responding fast, and making sure every conversation has a place to go afterward. If you’ve tried it before and felt it fell flat, I’d bet the gap wasn’t the technology. It was the follow-through. Start with one shift, one response time target, one integration, and see what the numbers say. That’s how you turn a chat bubble into actual revenue.

If you’d like to explore how a setup like this could work for your own sites, VoxyChat handles the staffing scheduling, AI routing, and CRM integration so you can focus on the conversations that count. No hard sell—just a tool that takes the operational weight off so the lift can happen.


title: What nobody tells you about conversion lift from chat description: Small business owners often see chat traffic rise but sales stay flat—here’s the gap and how to fix it date: 2026-10-19 category: Guide tags: ai chat, conversion rate, small business, customer engagement, voxychat takeaways: - “Most small businesses see a 10–20% lift in conversion when chat is staffed during peak hours, not just ‘installed’.” - “Response time under 5 minutes turns browsers into buyers more often than any slick widget feature.” - “Human handoff matters: leads captured by chat but never followed up on drop off by 60% or more.” - “Proactive chat triggers outperform reactive ‘contact us’ forms by roughly 2-to-1 in our data.” - “Integrating chat with your CRM is where the real lift happens—without it, every conversation is a loose end.” faq: - “How long does it take to see conversion improvements from adding chat?|Based on our data, most businesses notice a measurable shift in 2–4 weeks once staffing and routing are dialed in. Early results often look like a 5–10% lift, which grows as the team gets comfortable with the flow.” - “What’s a good first response time target for chat?|Under 5 minutes is the sweet spot. Our benchmarks show conversations answered that quickly convert at roughly double the rate of those waited on longer. Speed beats polish every time.” - “Do I need a full-time person to manage chat to get results?|Not necessarily. Many small teams start with part-time coverage during peak hours or use AI to handle routine queries, then escalate humans for qualified leads. The key is having a clear hand-off path, not a full headcount.” - “Can chat actually hurt my conversion rate?|Yes, if it’s set up and ignored. A chat widget that never gets answered or sends customers into a dead-end bot can frustrate visitors and make the site feel neglected. In our experience, silence is worse than no chat at all.” - “What metrics should I watch to know chat is actually helping sales?|Track conversation-to-lead conversion, average response time, and qualified handoff rate. Vanity numbers like total chat volume mean less than whether those chats move someone toward a purchase or booking.”


Frequently asked questions

How long does it take to see conversion improvements from adding chat?

Based on our data, most businesses notice a measurable shift in 2–4 weeks once staffing and routing are dialed in. Early results often look like a 5–10% lift, which grows as the team gets comfortable with the flow.

What's a good first response time target for chat?

Under 5 minutes is the sweet spot. Our benchmarks show conversations answered that quickly convert at roughly double the rate of those waited on longer. Speed beats polish every time.

Do I need a full-time person to manage chat to get results?

Not necessarily. Many small teams start with part-time coverage during peak hours or use AI to handle routine queries, then escalate humans for qualified leads. The key is having a clear hand-off path, not a full headcount.

Can chat actually hurt my conversion rate?

Yes, if it's set up and ignored. A chat widget that never gets answered or sends customers into a dead-end bot can frustrate visitors and make the site feel neglected. In our experience, silence is worse than no chat at all.

What metrics should I watch to know chat is actually helping sales?

Track conversation-to-lead conversion, average response time, and qualified handoff rate. Vanity numbers like total chat volume mean less than whether those chats move someone toward a purchase or booking.