Ask any chatbot vendor how to verify their conversion claims. I've done it. The answer is always the same: a case study PDF from 2022 and a line about "up to 30% conversion lift."
I got tired of that answer. Not because the numbers are necessarily wrong, but because there's no way for you to check. Vendors quote industry benchmarks. They show aggregate stats from their best customers. They don't give you a dashboard where you can see which conversations in your store led to actual purchases.
So when I built Emporiqa, I made this a priority: connect chat sessions to revenue. Not estimated revenue. Not projected revenue based on industry averages. Actual cart additions and completed orders, tied to specific conversations.
Message Counts Are Not Revenue
Most chat tools show you how many conversations happened. Message volume. Sessions per day. Average conversation length. Some show you which topics came up.
That's like counting how many people walked into a physical store without checking if anyone bought anything.
You end up in meetings saying things like "the chatbot handled 2,400 conversations last month" and someone asks "great, but did it make us money?" and you don't have an answer. You point to industry benchmarks. You cite vendor marketing. You make the argument that more engagement probably means more sales.
Probably. That word does a lot of heavy lifting in chatbot ROI discussions.
The gap between "conversations happened" and "revenue was generated" is the attribution gap. Closing it requires tracking the full path: a customer chats, adds something to their cart, checks out, and completes a purchase. Without that chain, you're guessing.
What Conversion Tracking Measures
Emporiqa tracks a conversion funnel with four stages:
- Chat session: A customer opens the widget and has a conversation.
- Cart addition: The customer adds a product to their cart during or after the chat. Your store's plugin captures this automatically.
- Checkout initiated: The customer starts the checkout process.
- Purchase completed: The order goes through. Your store notifies Emporiqa automatically when the order completes.
Each step is linked back to the originating chat session. The dashboard shows you the full funnel: how many sessions led to cart additions, how many of those reached checkout, and how many converted to actual purchases. Chat-attributed revenue gives you the dollar headline. The top products table shows which items get added to cart most from conversations. Widget engagement metrics (loads, opens, unique visitors, open rate, engagement rate with 30-day trends) help you understand whether customers are engaging with the chat at all.
How It Works Under the Hood
When a customer adds something to cart from the chat, the plugin records it. When they complete a purchase, your store notifies Emporiqa. The dashboard connects the dots: which chat conversation led to which purchase. Your developer sets this up once during installation — after that, it runs automatically.
Events are deduplicated by order ID, so the same order never counts twice. Sessions older than 7 days are flagged as stale for attribution purposes, because a chat from two weeks ago probably didn't influence today's purchase.
Building a Business Case With Your Data
After 30 days of running Emporiqa with conversion tracking enabled, you have real numbers for the "prove ROI" conversation:
- Total chat sessions
- Sessions where customers added items to cart
- Sessions that reached checkout
- Sessions that resulted in completed purchases
- Total revenue from chat-attributed orders
- Top products that customers add to cart from conversations
- Daily breakdown of the entire funnel
These come from your store, your customers, your products. Not from a vendor's cherry-picked case study. Not from an industry report about stores that may look nothing like yours.
When a stakeholder asks "is the chatbot paying for itself?", you open the dashboard and show them: "Last month, 340 chat sessions led to cart additions. 180 of those completed checkout. Revenue attributed to chat-assisted sessions: $27,000. The chat cost us $59 — capped (340 conversations × $0.25 = $85, hit our default $59 cap)."
That conversation takes 30 seconds. No slides, no industry benchmarks, no hand-waving.
What the funnel tells you beyond revenue
The funnel stages also reveal optimization opportunities. If lots of sessions lead to cart additions but few reach checkout, the problem might not be the chat. It might be your checkout flow. If sessions rarely lead to cart additions despite high conversation volume, maybe the assistant needs better product data to work with. The ROI post I wrote earlier covers how to think about these metrics in the broader context of chatbot economics.
Assisted Conversions vs. Causation
I need to be upfront about what conversion tracking does and doesn't prove.
A customer who chatted with the assistant and then bought something shows up as a chat-attributed purchase. But that customer might have bought anyway. Maybe they were already on the product page with their credit card out and just had a quick question about shipping. The chat didn't cause the sale. It assisted it.
Conversion tracking shows correlation, not causation. A chat session preceded a purchase. Whether the chat was the deciding factor or just one touchpoint in a longer journey is harder to determine.
This limitation applies to every analytics tool. Google Analytics has the same attribution problem. Marketing teams debate "last click vs. first touch vs. multi-touch" attribution constantly. There's no perfect answer.
But here's what I know: correlation with your own data is still more useful than vendor claims with no data. "87 customers who chatted last month also completed purchases totaling $14,000" is a concrete number you can work with. It might overstate the chat's direct impact. It might understate it (some customers chatted, left, came back a week later on a different device, and bought without the session being linked). The number isn't perfect, but it's yours, and it's grounded in what happened in your store.
The Other Side of Measurement: Do Customers Like It?
Revenue attribution answers "is the chat making money?" but not "do customers like the experience?" Both questions matter. An assistant that pushes products aggressively might generate short-term cart additions while annoying customers who never come back.
Emporiqa includes post-chat CSAT ratings: thumbs up or thumbs down. Simple, low-friction. Customers rate the conversation when it ends. The dashboard aggregates these scores and lets you review individual conversations that received negative ratings.
Combining conversion data with CSAT gives you two dimensions of measurement:
- Revenue impact: Is the chat contributing to sales?
- Customer satisfaction: Are customers happy with the interactions?
If revenue is up but satisfaction is down, something is off. If satisfaction is high but conversions are flat, the assistant might need better product knowledge or more effective recommendations. The two metrics together tell a more complete story than either one alone.
What Conversion Tracking Doesn't Do
- It doesn't prove causation. As I covered above, it shows which purchases were associated with chat sessions, not which purchases were caused by them.
- It doesn't track cross-device journeys. If a customer chats on their phone and buys on their laptop later, those sessions aren't linked unless the customer is logged in on both devices.
- It doesn't replace your existing analytics. Your store's analytics platform (Google Analytics, Matomo, etc.) still handles overall conversion tracking. Emporiqa tracks the chat-specific funnel.
- Stale sessions get flagged. If a customer chatted 10 days ago and buys today, the 7-day attribution window means that purchase won't count as chat-attributed revenue. The event is still recorded, but it won't inflate your numbers.
- It requires your store to send data. Cart events are tracked automatically through the widget, but purchase confirmation requires your developer to enable tracking during setup. If it's not enabled, the funnel stops at checkout.
Getting Started
If you're evaluating chat tools and "prove ROI" is on your requirements list:
- Baseline your current metrics. Know your conversion rate, average order value, and support costs before adding chat. You need a "before" to measure against.
- Set up conversion tracking from day one. Don't wait three months and then wish you had data from the beginning. Ask your developer to enable conversion tracking during setup.
- Give it 30 days. You need a full month of data before the numbers mean anything. Weekly fluctuations are noise.
- Review the funnel, not just the total. Where do customers drop off? High cart-add rate but low checkout rate points to a different problem than low cart-add rate.
- Combine with CSAT. Revenue numbers without satisfaction data tell half the story.
Understanding your cost per conversation matters for ROI. Emporiqa is pay-as-you-go: $0.25 per conversation, $25 of signup credit (~100 conversations), and a $59/month default cap you can raise from the billing dashboard — so you always know what you're paying.
A free Emporiqa account includes conversion tracking, so you can see the funnel working with your own products before committing. Upload your catalog and watch real data flow through the dashboard. No card required. Create an Emporiqa account and start measuring.
