PROVIDER Β· European experimentation and personalisation
AB Tasty
Founded in 2013 in Paris, AB Tasty combines A/B testing, personalisation, and feature flags with visual editing so marketing ships experiments itself, fitting teams validating landing pages across channels.
4.1pts
Pros
Cons
Product Information
Detailed Notes
Overview
AB Tasty was founded in 2013 and is headquartered in Paris, one of the European platforms most often compared with VWO and Optimizely. Its structure is testing plus personalisation plus feature flags, letting marketing teams ship experiments without waiting for engineering.
For Testing & Optimization work the practical value is visual editing plus audience targeting, where operators show different landing page versions by source, device, and behaviour to test whether channels need different above-the-fold messaging, which moves acquisition efficiency far more than minor button tweaks.
The limits mirror SaaS peers, with tiered traffic pricing that climbs at scale, experiment scripts adding requests that need careful loading, and no guarantee of correct conclusions without methodology. Chinese interface and local support are limited.
Core Strengths
- Operating since 2013 with A/B testing, personalisation, and feature flags in one platform
- Visual editing changes pages without code so marketing ships experiments independently
- Audience targeting by source, device, and behaviour validates channel-specific first-screen strategy
- Experiment reports with significance assessment compare conversion across variants
Product Ecosystem
Experiments and visual editing
A/B, multivariate, and split tests with a visual editor for copy and layout. Make changes large enough, since button-colour differences need enormous samples to reach significance.
Personalisation and audiences
Different versions by source, region, device, and behaviour. Validate per-channel segments first and then decide whether each channel needs its own landing page, which saves production cost.
Drawbacks
- Tiered traffic pricing climbs noticeably as traffic grows
- Experiment scripts add requests, hurting first paint and CWV if loaded carelessly
- No guarantee of valid conclusions without methodology and adequate sample size
- Limited Chinese interface and local support, requiring English documentation skills
Use Cases
- Marketing-led conversion work (β β β β ) - Visual editing frees operators from engineering queues
- Validating first screens across channels (β β β β ) - Segment experiments test per-channel landing strategy
- Product teams needing feature flags (β β β β) - Usable but shallower than specialist platforms
- Low-traffic sites (β β ) - Sample size rarely reaches significance
Pricing
| Plan | Price | Scope |
|---|---|---|
| Testing plans | Monthly | Tiered by traffic with baseline experiments |
| Personalisation and flags | Plan dependent | Audience targeting and progressive rollout |
| Enterprise | Custom quote | High traffic quotas, SSO, and support |
Plans tier by traffic with upgrades on overage. Include the performance cost of the script, since careless loading that degrades CWV can cancel out the conversion gain.
FAQ
-
How does it compare with VWO? Both are marketing-side platforms, with VWO covering more (recordings and heatmaps) and AB Tasty focusing on personalisation and targeting. The deciding question is whether you need audience-level personalisation.
-
How long should tests run? At least one full business cycle, usually one to two weeks, avoiding promotions. Stopping early is the most common error and systematically favours the early performer.
-
Does it hurt SEO? Split tests without URL changes have limited impact, but avoid serving different content to crawlers and pin the original version for bots.