SplitWit-style AB integrations typically write each experiment to a custom post type such as splitwit_experiment with variant definitions either as child posts (splitwit_variant) or as a serialized array in wp_postmeta. Targeting rules (URL patterns, user roles, traffic split) live in postmeta, and exposure or conversion events land in a log table like wp_splitwit_exposures.

The default admin for an AB tool inside WordPress is usually a wizard for editing one experiment at a time. Listing experiments by status, traffic share, primary metric, and current exposure count is rarely a first-class view. Spotting tests that have been running for weeks without enough samples, or tests still targeting a URL pattern that no longer exists on the site, means clicking into each one.

SleekView reads the SplitWit custom post type plus its wp_postmeta rules and joins with the exposures log table. Experiment, variant count, primary metric, traffic split, exposures in the last 7 days, and status appear in one row. Edits route through the integration's own save hooks so randomization tables and target audiences refresh as expected.

AB testing is high-leverage and low-discipline. The first few experiments are documented carefully, but as the program scales, dozens of experiments end up running on different pages, with different audiences, against different primary metrics, and with different teams owning each. Most AB integrations inside WordPress give you a wizard for editing one experiment at a time and a basic list table that shows title and status.

That works for three experiments and breaks at thirty. The teams operating the program end up exporting spreadsheets, keeping shadow Google Docs of which tests are live, and missing tests that quietly stopped collecting data after a route change or a theme update. SleekView treats the AB integration's records as the structured data they actually are.

Experiment posts, targeting rules, traffic splits, exposure counts, and primary metrics become joinable columns. Growth catches dead tests early, marketing scopes views per campaign, and QA validates that targeting still matches the live site after every release. The result is an AB program that scales past three concurrent tests without the spreadsheets.