Documentation Index
Fetch the complete documentation index at: https://docs.abtestly.com/llms.txt
Use this file to discover all available pages before exploring further.
ABTestly
ABTestly is an A/B testing platform built for teams that care about site performance and cost predictability. It runs on Cloudflare’s edge, ships a small inline snippet, and never auto-charges you for traffic overages.What’s different about it
<50mssnippet impact (P95) — about 4x faster than mainstream alternatives. The snippet is small, runs synchronously, and doesn’t depend on third-party tag managers.- Graceful overage handling — if you exceed your monthly tracked users (MTU) limit, your already-bucketed visitors keep seeing their variant (statistical validity preserved), but new visitors stop being bucketed. You get a warning email and decide whether to upgrade. No surprise invoices.
- No data deletion — downgrading, canceling, or being archived after 90 days of inactivity all preserve your data. Reactivation is one click.
- Permanent free tier — 1 site, 1 active experiment, 3,000 MTU. Enough to evaluate the product on a real page.
Who this is for
- CRO professionals running tests across multiple sites
- Solo founders validating landing pages without a Convert/VWO contract
- Agencies managing experiments for client portfolios
Where to start
Quick start
From signup to live experiment in 15 minutes.
Snippet install
Where, how, and why to install the snippet correctly.
Targeting
Locations, audiences, conditions, bucketing.
Plans
Pricing, MTU explanation, grandfathering on downgrade.
How it works at a glance
- You sign up at app.abtestly.com, add your site, and paste a snippet into the
<head>of your HTML. - You build an experiment in the dashboard: name, hypothesis, targeting, variants, goals.
- You click Start. The new config propagates to all edge locations within ~60 seconds.
- Your visitors hit your site, the snippet decides which variant to show, and they see their assigned version on the first paint (no flicker).
- Real-time exposures and conversions stream into your results dashboard. Most tests need 5,000+ exposures per variant for 95% statistical confidence.