You shouldn’t have to take an AI tool’s word for its results. This guide walks you through three layers of measurement – all available from day one – so you can verify with your own data that Corelex is driving more conversations, more add-to-carts, and more completed orders.
The Corelex Analytics tab #
From the moment the plugin is active, Corelex logs every visitor interaction in real time. The Analytics tab in your WordPress admin gives you a live view of how visitors are engaging with the widget — no extra setup required.
What you can see #
Top products #
Products your visitors ask about most — direct signal for what drives purchase intent.
Zero-result rate #
Percentage of queries where the AI couldn’t find a match — your content gap indicator.
Unanswered questions #
The actual queries that returned nothing — tells you exactly what to add to your catalogue.
How to find it #
Go to WordPress admin → Corelex → Analytics. Use the date range selector to filter by week, month, or a custom period. All data is stored in your Google Cloud environment — no third-party analytics service is involved.
What to look for first: A high zero-result rate (above 15–20%) means visitors are asking questions your product catalogue or knowledge base can’t answer. Use the unanswered questions list to identify gaps and either add products or update your content.
Google Analytics 4 – Corelex in your own dashboard #
If your store already has Google Analytics 4 installed, Corelex starts sending events to it automatically. No additional configuration needed. Every significant interaction – from opening the widget to clicking Add to Cart inside the chat – fires as a named GA4 event, prefixed corelex_ so they’re easy to filter.
This means you can analyse Corelex-driven behaviour using the same GA4 reports, segments, and funnels you already use for the rest of your store.
Events Corelex sends to GA4 #
|
Event name |
When it fires |
|---|---|
|
corelex_chat_open |
Visitor opens the chat widget panel |
|
corelex_message_sent |
Visitor sends a query (includes turn number) |
|
corelex_add_to_cart |
Visitor clicks Add to Cart directly inside the chat (includes product ID, name, price) |
|
corelex_product_card_click |
Visitor clicks a product card surfaced by the AI |
|
corelex_zero_result |
A query returned no matching products or content |
|
corelex_whatsapp_handoff |
Visitor tapped the WhatsApp handoff button |
|
corelex_dwell_trigger_fired |
Product page dwell trigger opened the widget automatically |
|
corelex_cart_abandonment_trigger_fired |
Cart abandonment trigger activated after an add-to-cart |
|
corelex_exit_intent_shown |
Exit intent overlay appeared |
|
corelex_exit_intent_cta_click |
Visitor clicked a CTA inside the exit intent overlay |
|
corelex_exit_intent_lead_submitted |
Visitor submitted the lead capture form |
|
corelex_search_intercepted |
A WooCommerce search was redirected into the chat widget |
How to use these in GA4 #
- In GA4, go to Reports → Engagement → Events. Filter the event name column by
corelex_to see all Corelex interactions at a glance. - Create a custom segment for sessions that included at least one
corelex_chat_openevent. Compare this segment’s purchase conversion rate against all sessions – this is your baseline assisted-conversion view. - Build a funnel exploration:
corelex_chat_open→corelex_message_sent→corelex_add_to_cartor native WooCommerce purchase event. This shows how many visitors flow from chat interaction to purchase. - Mark
corelex_add_to_cartas a conversion event in GA4 settings. You can then compare this conversion against your standard WooCommerce add-to-cart event to see the relative weight of chat-direct vs. organic adds.
Note on attribution:
corelex_add_to_cartonly fires when a visitor clicks the Add to Cart button inside the chat widget. If a visitor chats with the agent and then navigates to the product page and adds to cart there, GA4 will record it as a standard WooCommerce add-to-cart – not a Corelex event. This is intentional: the distinction lets you separate direct chat-driven conversions from assisted ones.
The PoV pilot – controlled A/B conversion proof #
Steps 1 and 2 tell you how visitors interact with Corelex. The PoV (Proof of Value) pilot answers the harder question: does the widget actually cause more conversions, or would visitors have converted anyway?
The pilot runs a controlled A/B test on your own store. During the pilot period, every visitor is assigned to one of two groups – treatment (widget active) or control (widget hidden) – using a deterministic algorithm tied to their anonymous visitor ID. The same visitor always sees the same experience, so results are not polluted by the same person switching groups between visits.
What gets measured #
For both groups, Corelex tracks how many visitors were exposed and how many added a product to their cart. At the end of the pilot, the results panel in your WordPress admin shows:
|
Metric |
Control (no widget) |
Treatment (widget active) |
|---|---|---|
|
Visitors exposed |
N |
M |
|
Add-to-cart events |
X |
Y |
|
Conversion rate |
X / N % |
Y / M % |
|
Conversion lift |
– |
+Z% |
For the treatment group, a further breakdown shows how many add-to-cart events came directly from inside the chat widget vs. from the product page — so you can see whether the widget is influencing decisions or simply acting as a shortcut button.
How to set up a pilot #
Contact is if you’re interested to run a proof-of-value pilot.
