The plugin stores the account key and embed settings in wp_options under zendesk_chat_settings. Chats themselves live in Zendesk. Most production setups wire the Zendesk chat.ended webhook into a custom zendesk_chat post type so the WP side has the chart-ready data SleekView needs.

Without the webhook, the WP side only stores embed settings and identify rules. Charts are limited to coverage of the widget across wp_posts. For agent, department, rating, and duration cards, configure the chat.ended webhook to POST into the zendesk_chat CPT first.

The webhook saves pre-chat form responses as postmeta on each zendesk_chat post. Any select-style field becomes a groupBy dimension for a donut and any numeric field becomes a sum or average aggregation, so reason mix and visitor segment charts come from the same dataset.

Yes. SleekView honors WordPress capabilities for the zendesk_chat post type. An agent role sees their own chats while a support lead sees the full team's. The capability filter applies identically to chart cards and the underlying SleekView table view of the same dataset.

Yes. The webhook saves the originating page URL as zendesk_page postmeta. SleekView joins that to wp_posts.ID via slug or URL match and renders a top-pages bar resolved to post_title, useful for marketing and content teams reviewing widget placement and intent.

Charts run aggregate queries on view load. Open the dashboard and the numbers reflect the current state of zendesk_chat including any webhook posts written up to that moment. There is no caching layer to refresh and no scheduled job that lags behind the live queue.

Yes. Clicking a slice or bar opens the SleekView table with the same filter applied. From the agent bar with the lowest rating, one click lands on the exact list of zendesk_chat posts that drove the average, each with the transcript link one further click away.

Charts query zendesk_chat using the same indexes WordPress maintains for post-type and postmeta lookups. On sites with tens of thousands of chats archived, dashboards load in well under a second from the admin context with no extra database tuning required.