//
AI-powered A/B testing and personalization built in Paris — GDPR-first by design
Kameleoon is a Paris-based experimentation and personalization platform founded in 2008. It delivers web experimentation, feature experimentation, and AI-driven personalization through a hybrid client-side and server-side architecture. Patented AI Predictive Targeting adjusts visitor experiences in real time without requiring cookie consent for many use cases. Trusted by 700+ enterprise clients across Europe and North America.
Headquarters
Paris, France
Founded
2008
Pricing
EU Data Hosting
Yes
Employees
51-200
30-day free trial available
Pay-as-you-go
Contact Sales
Billing: annual
Kameleoon is a Paris-born experimentation and personalization platform that has spent nearly two decades becoming the European alternative to Optimizely and VWO. Founded in 2008 as a web personalization company and incorporated as Kameleoon SAS, the company has built its entire platform around a single premise: enterprise-grade A/B testing and AI personalization should not require teams to choose between performance and compliance.
The product covers three distinct use cases — web experimentation (client-side A/B and multivariate tests), feature experimentation (server-side flag-based testing across backend systems and mobile apps), and AI-driven personalization — all managed from a unified workspace. This breadth distinguishes Kameleoon from tools that specialize in just one layer. A product team can run a front-end visual test and a server-side API experiment in parallel, with results appearing in the same reporting interface.
What separates Kameleoon technically is its hybrid architecture. Client-side tests run in the browser using a lightweight JavaScript snippet. Server-side tests run via SDKs deployed in the application stack. Kameleoon enables hybrid experiments that combine both simultaneously — for example, using client-side audience data to target a server-side variant — without requiring separate tool configurations or data pipelines.
The AI layer is built around two patented capabilities: AI Predictive Targeting, which models visitor behavior in real time to predict conversion probability and adjust personalization without cookie consent for many standard use cases, and Prompt-Based Experimentation (PBX), which allows teams to describe a test idea in natural language and receive a generated experiment design, variant copy, and targeting configuration. Predictive Impact Scoring then ranks pending experiments by estimated revenue lift before any traffic is allocated.
The technical case for Kameleoon at the enterprise level starts with the hybrid architecture. Many web experimentation tools run exclusively in the browser — quick to deploy but limited to front-end changes and unable to test backend logic, API responses, or mobile app features. Server-side tools do the reverse. Kameleoon runs both simultaneously and allows teams to bridge them: a client-side audience segment can drive targeting decisions for a server-side test without needing a second tool or a custom data integration. For engineering teams running platform-wide experiments, this eliminates the overhead of managing separate tooling stacks.
PBX is Kameleoon's most forward-looking feature. Instead of navigating the experiment builder manually, a product manager or marketer describes the hypothesis — for example, "test whether moving the checkout button above the fold increases conversion on mobile" — and the AI generates the full experiment configuration including variant mockups, targeting rules, and KPI selection. The generated test can be edited before launch. This workflow removes a meaningful time cost from experiment design and makes the platform accessible to teams without a dedicated CRO specialist.
Predictive Impact Scoring adds a prioritisation layer. Drawing on aggregated data from thousands of past experiments across the Kameleoon customer base, it models expected revenue impact for each pending test idea, allowing product and marketing teams to sequence tests by estimated business value rather than intuition.
Kameleoon's AI Predictive Targeting analyses each visitor's behaviour in real time — pages viewed, scroll depth, device, time on site, traffic source — and models the probability they will convert. Personalized experiences are served based on these behavioral predictions rather than stored personal data or third-party cookies, allowing many personalization use cases to operate under legitimate interest or with minimal consent requirements depending on the specific implementation.
This matters for European companies navigating CNIL requirements in France, the ICO's guidance in the UK, and GDPR enforcement across member states. Kameleoon documents its consent management approach and offers a built-in module for handling cookie preferences. The platform also supports HIPAA-capable configurations on the enterprise plan, relevant for healthcare and pharmaceutical clients experimenting with patient-facing digital products.
Feature experimentation extends A/B testing logic into the product engineering workflow. Teams can wrap new features in feature flags, roll them out to defined user segments (1%, then 10%, then 50%), collect experiment data at each stage, and roll back instantly if metric degradation is detected. Kameleoon's SDKs support common server-side environments including Node.js, Python, Java, Go, PHP, Ruby, and .NET. The same targeting engine used for web personalization applies to feature flag conditions, so product and marketing experiments can share audience definitions.
Kameleoon operates on a fully custom pricing model with no published rates for enterprise tiers. The Starter plan is publicly documented at $495 per month for up to 50,000 Monthly Unique Visitors, and includes a 30-day free trial. Enterprise pricing scales with monthly unique visitor volume and unlocks server-side experimentation SDKs, AI Predictive Targeting at full capacity, HIPAA capability, and dedicated customer success support.
There is no self-serve sign-up for enterprise evaluation. All demos and trials go through a sales qualification process. This is standard for enterprise software at Kameleoon's price point and allows the team to configure the trial environment appropriately — but it creates a meaningful barrier for teams doing early-stage vendor comparisons without budget approval.
For context, Optimizely's pricing is similarly opaque at the enterprise level. VWO publishes plans starting around $399 per month for basic web testing but requires enterprise pricing for full feature flag and personalization access. Kameleoon's $495 Starter plan positions it as accessible for mid-market teams willing to self-qualify through the trial process.
Kameleoon SAS is registered in Paris (SIREN 507 754 570) and stores all data in EU-based infrastructure. The platform includes a consent management module that integrates with standard cookie management frameworks, supports tag-based consent for client-side scripts, and offers cookieless personalization for use cases that fall under legitimate interest under GDPR Article 6(1)(f).
The GDPR-first positioning is not marketing language — it reflects the company's French regulatory environment and customer base. Major European retailers, media companies, and financial services firms, where data protection officers have significant influence over procurement, make up a large share of Kameleoon's 700+ enterprise clients. The platform is designed to satisfy procurement checklists that require documented DPAs, EU data residency, and consent audit trails.
The company does not publish formal ISO 27001 or SOC 2 certifications as prominently as some US competitors. Enterprise prospects with specific certification requirements should confirm the current audit status during the sales process.
If you are a digital product team or growth marketing team running more than a handful of experiments per month and need both front-end and back-end testing in one platform, Kameleoon is one of very few EU-built options that delivers the full stack.
If your company has strict GDPR requirements and your data protection team scrutinises cross-border data transfers, Kameleoon's French infrastructure and cookieless personalization architecture resolve concerns that US-hosted tools create.
If you are an e-commerce business or publisher running primarily visual front-end tests at modest traffic volumes, VWO or AB Tasty offer comparable web testing features at lower or more transparent starting prices.
If you need a self-serve tool you can sign up for without speaking to sales, Kameleoon is not the right fit at the enterprise tier — the evaluation process requires sales engagement.
If your engineering team is building a server-side feature experimentation program alongside web testing, the unified platform avoids the overhead of a separate LaunchDarkly or Split.io subscription alongside a web testing tool.
Kameleoon has quietly built one of the most technically complete experimentation platforms in Europe. The hybrid architecture, AI Predictive Targeting, and Prompt-Based Experimentation represent genuine product innovation — not features bolted on to a legacy A/B testing tool. The GDPR-first positioning is backed by real EU infrastructure and a consent management module that satisfies most European procurement requirements. The main trade-offs are the opaque pricing model and the sales-gated evaluation process, which create friction for teams in early comparison stages. For European enterprises where compliance, server-side testing, and AI personalization are all on the requirements list, Kameleoon is a genuinely strong contender.
Yes. Kameleoon is a French SAS storing data in EU infrastructure and includes a built-in consent management module. The platform supports cookieless personalization for many standard use cases, which can reduce or eliminate cookie consent requirements depending on the implementation. A Data Processing Addendum is available for enterprise customers.
PBX is Kameleoon's AI-assisted test creation workflow. Teams describe a hypothesis in plain language, and the AI generates the full experiment configuration — variant designs, targeting rules, and KPI definitions. Predictive Impact Scoring then ranks pending ideas by estimated revenue lift based on thousands of historical experiments in the platform's dataset. It is designed to reduce the specialist knowledge required to design statistically sound experiments.
Both platforms compete at the enterprise level with hybrid experimentation capabilities. Kameleoon's advantages are its EU infrastructure and GDPR-first architecture, AI Predictive Targeting with cookieless support, and the PBX test-creation workflow. Optimizely has a larger product suite including CMS, commerce, and content experimentation layers, a larger partner ecosystem, and stronger name recognition with US enterprise procurement teams. Both require custom pricing.
Yes. Kameleoon supports feature flags and progressive rollouts as part of its feature experimentation offering. Flags can be targeted to defined user segments, rolled out incrementally, and connected to experiment metrics. SDKs are available for Node.js, Python, Java, Go, PHP, Ruby, and .NET.
Yes. AI Predictive Targeting operates on real-time behavioral signals rather than stored personal data or third-party cookies, allowing many personalization use cases to run under legitimate interest without requiring a cookie consent event. The platform's consent management module handles configurations where cookie consent is required for specific use cases.