Personalisation in commerce is too often a trade-off today: more personalisation = more performance problems, more tools, more compliance risk. At Laioutr it works differently.
PERSONALISATION — RULE-BASED AND AI-DRIVEN
Personalisation meets performance — and an agent that never sleeps.
Display Conditions in Studio for marketing teams. Personalization Agent for continuous ML-driven optimisation. Edge delivery for performance without flicker. Three layers, one platform.
We think of personalisation as an architectural layer of the frontend platform — with clear rules in Studio, an AI agent for continuous optimisation and edge delivery for performance out of the box. GDPR-compliant, brand-compliant, without flicker.
The definition
What personalisation means at Laioutr.
Personalisation at Laioutr is an architectural layer of the frontend platform — not separate software. On the layer available live today, you explicitly define via Display Conditions in Studio when which component, content or variant is served to whom.
On the AI layer, the Personalization Agent continuously decides which variants work best for which segments, based on real performance data. Both layers run in parallel, complement each other, and deliver personalisation via the edge — without flicker, without client-side hydration issues, without SEO loss.
WHO IT'S FOR
Personalization in two layers - today and tomorrow, together.
We clearly separate what you can already use today from what runs as an AI layer in the background. Both belong to the same platform. Both work together. You build the rules. The agent finds the patterns.
RULE-BASED
Display conditions - rule-based personalization
Directly in Studio you configure per component when it is visible and which content appears. Multiple conditions can be combined AND, OR, NOT. Marketing works without an engineering ticket.
Examples of rules:
Region (country, state, city)
Language and locale
Device type (desktop, tablet, mobile)
Customer segment (logged in vs. anonymous, VIP vs. standard, etc.)
Login state and customer lifecycle phase
Cart value and cart contents
Date / time / season
UTM parameters from marketing campaigns
Custom data fields from your backend or CDP
AGENTIC
Personalization Agent - continuous ML optimization
The Personalization Agent runs in the background, observes behavior and performance and continuously optimizes which variants work best for which segments.
What the agent concretely does:
Segment discovery, finds previously undiscovered customer clusters
Variant optimization, tests which variant converts best per segment
Multi-armed bandit, dynamically distributes traffic to the best variants
Decay detection, recognizes when a variant gets "old" and loses
PERSONALIZATION AGENT
What the Personalization Agent could concretely automate
Personalisation beyond "insert first name" or "detect region". The Personalization Agent handles tasks that, in a classic setup, would keep a dedicated CRO team busy for quarters. Personalisation moves from a sprint to a background process.
Segment Discovery
The agent identifies customer clusters beyond classic segments — behavioural patterns that aren't modelled in your CDP but are conversion-relevant.
Variant Optimization
For each component and segment, the platform continuously tests which variant performs best — headlines, CTAs, layouts, recommendation orderings.
Multi-Armed Bandit
Instead of rigid A/B tests, the agent dynamically distributes traffic to the best-performing variants — learning speed doubled, opportunity cost halved.
Gap Analysis
The agent spots where personalisation is missing — which components and pages aren't yet personalised for which segments, even though it would pay off.
Decay Detection
When a variant grows "stale" over time (conversion drops), the agent detects it and automatically proposes a new variant, where appropriate in collaboration with the Content Agent.
Cross-Channel Sync
Personalisation insights from the web frontend flow back into the CDP, email tools and customer service. One layer, one learning effect, visible everywhere.
DATA SOURCES
What powers personalisation.
Personalisation is only as good as the data it's built on. Laioutr uses five data source categories, each of which can be enabled or disabled per use case.
Implicit frontend data
Region, language, device, screen size, referrer, UTM parameters, returning vs. first visit. Available without cookie consent (edge-detected).
Customer State
Login status, customer lifecycle stage, cart contents, cart value, wishlist, last order — straight from the commerce backend (Shopify, OXID, Shopware, etc.) via the Connect layer.
CDP data
Customer profiles from your CDP (Segment, mParticle, Tealium, Bloomreach Engagement, Klaviyo, etc.) — segment assignment, lifetime value, affinities, predictive scores.
In-session behaviour
What the user has viewed, added to cart or searched for in the current session — real-time signals for same-session personalisation.
Custom Data Fields
Anything else you have and want to connect via the Connect layer — external APIs, loyalty programmes, industry-specific data (B2B terms, price lists, etc.).
Customer Data Platform
How Laioutr personalisation works with your CDP.
If your team already uses a CDP — Segment, mParticle, Tealium, Bloomreach, Klaviyo — then that's the right customer data source. We don't compete with it. We integrate deeply.
Via pre-built apps, customer profiles from your CDP flow into Laioutr as input for Display Conditions in Studio and as a training signal for the Personalization Agent. Performance and conversion data from the frontend flow back into the CDP, so your customer profiles grow richer the more Laioutr works. You keep your CDP as the single source of truth. We deliver the frontend layer that serves the profiles to the customer — performant, in real time, without code mapping.
Segment · mParticle · Tealium · Bloomreach Engagement · Klaviyo · Customer.io · Twilio Segment · Custom via REST/GraphQL
Performance
Personalisation without flicker, without performance loss.
Classic personalisation tools have a blind spot: performance. Doing personalisation client-side risks flicker (FOUC), hydration issues and LCP regressions. We solve it differently — at the edge. LCP under 1.5 s even with full personalisation.
Personalisation at the edge
Personalised content is delivered at the edge — from the very first byte sent to the browser. No client-side logic that swaps content after the fact.
Server-Side Personalization Hints
Even dynamic personalisation (e.g. customer state) is prepared server-side before the HTML renders. No hydration mismatch, no layout shift.
SEO stays SEO
Personalisation is transparent to search engines — Google sees the default variant, users see the personalised one. No cloaking risk, no hreflang confusion.
Personalisation x A/B testing
Personalisation and A/B testing — one layer, two disciplines.
Classically, personalisation and A/B testing are two tools and two sprints, two reportings, two sets of component variants. At Laioutr they live in the same layer. Display Conditions in Studio serve both personalisation (component only for segment X) and A/B testing (variant A vs. B with random distribution).
The Personalization Agent runs as a multi-armed bandit, combining both disciplines by dynamically distributing traffic to the best-performing variants per segment. The result: instead of classic A/B tests with a fixed split, you get continuous optimisation that handles personalisation and testing in a single operation. Learning speed doubles, opportunity cost halves.
Conditions = rule-based personalisation and explicit A/B testing.
Personalization Agent = ML-driven optimisation and dynamic traffic distribution
Both run on the same edge layer — no double performance burden
Performance
GDPR-compliant and brand-compliant — a prerequisite, not an add-on.
Personalisation in Europe is unthinkable without clear compliance and brand governance. We make both a prerequisite of the platform, not a bolted-on feature. AI without compliance is risk. Compliance without AI is stagnation. We deliver both.
GDPR & Compliance
EU hosting available; data stays in the chosen region
Built-in cookie consent layer (TCF 2.0-compatible, controllable per data source)
Configurable per personalisation rule which
data sources are permitted
Customer profiles from the CDP are not used for model training; content stays with you
DPA (Data Processing Agreement) included in the contract by default
Audit logs for all personalised deliveries
(compliance-relevant during audits)
Brand guardrails
Tone, style and imagery rules configurable per brand
Forbidden words and taboo topics are automatically excluded by the Personalization Agent
Approval workflows per personalisation variant (what goes live directly, what goes to review)
Cross-brand protection: no mixing of content across brand boundaries
A/B test results can be isolated per brand
Audit trail for every variant generated or selected by the agent
Performance
What personalisation looks like in everyday commerce.
Six concrete examples from real commerce setups — not theoretical workflows, but tasks that keep marketing teams busy today.
Region-specific content
The hero banner shows winter products to DE visitors and summer products to AU visitors. Controlled via Display Conditions, no engineering required.
Who: Marketing teams
First-time customer vs. VIP
Anonymous first-time visitors see the brand story and top products. VIP customers see exclusive offers, wishlist reminders and new releases first.
Who: Commerce teams with a clearly segmented customer base
Black Friday / seasonal campaigns
Black Friday banner active from 28 Nov to 1 Dec, automatically reverting to default afterwards. Cart value > 100 EUR? Show the free-shipping banner.
Who: Marketing teams with high campaign frequency
Recommendation optimisation
Instead of rigid "customers also bought" lists, the Personalization Agent continuously selects which recommendation logic converts best per segment.
Who: Commerce teams with a large assortment
Cross-brand personalisation
Brand families can share personalisation insights without violating brand boundaries. What works for Brand A is tested on Brand B — with brand guardrails as protection.
Who: Multi-brand holdings
Mobile-only optimisation
Mobile visitors from 3G/4G regions are served a lighter hero visual and more compact components. Conversion holds, performance rises.
Who: Commerce teams with high mobile traffic
FAQ