Uniform Context

Uniform Context is a personalization and optimization SDK that enables you to classify visitors, personalize content, and run A/B tests across web and mobile experiences. It's designed from the ground up to work with modern front-end frameworks like React and Vue, runs on the edge for sub-50ms response times, and never collects personally identifiable information.

Classification is the process of building a real-time profile of each visitor based on their behavior, attributes, and external data. This profile is the foundation for all personalization and testing decisions.

Uniform tracks and classifies visitors using three complementary mechanisms:

  • Enrichments - Build behavioral profiles by tracking content engagement. When a visitor views product pages, reads articles, or interacts with specific content categories, their interest scores update automatically. Over time, this creates a rich picture of each visitor's preferences and intent.

  • Signals - Detect specific visitor behaviors and context in real time. Signals fire based on criteria like the current page, UTM campaign parameters, query strings, cookie values, page view counts, or analytics events. Multiple criteria can be combined with AND/OR logic to create sophisticated detection rules.

  • Quirks - Store key-value attributes about a visitor, such as geographic location, device type, customer segment, or any custom classification from your own systems. Quirks are set automatically by edge integrations (like GeoIP data from CDNs) or programmatically from any data source.

These mechanisms work together to create a multi-dimensional visitor profile. For example, a visitor might have enrichment scores indicating interest in specific product categories, active signals from a paid campaign landing page, and quirks populated from your CDP identifying them as a returning customer. All of this context is available instantly for personalization decisions.

For detailed configuration, see the classification guide.

Personalization lets you tailor the content and components each visitor sees based on their classification profile. You define multiple variations of a component and set criteria for when each variation should appear. Uniform evaluates those criteria against the visitor's profile and selects the best match.

Uniform supports several approaches to variation matching:

  • Strongest match - Automatically selects the variation that best matches the visitor's strongest dimension score. Ideal for intent-based personalization where you want to show content aligned with a visitor's primary interest.

  • Top-down criteria matching - Evaluates variations in order, selecting the first one whose criteria are met. This gives you precise control over priority and supports complex rules combining multiple dimensions with comparison operators.

  • Custom matching - Implement your own selection logic through Uniform's plugin system when you need bespoke personalization algorithms.

Every personalization component includes a default variation that's shown when no criteria are matched, so there's always a meaningful experience for every visitor.

Each personalization component has an analytics tracking name that's sent to connected analytics tools along with the selected variation. You can set a control group (a percentage of visitors who see the default, non-personalized experience) to measure the actual impact of your personalization strategy. Uniform ships with a Google Analytics 4 connector and a generic event interface that adapts to any analytics tool.

For detailed configuration, see the personalization guide.

A/B testing lets you expose different content variations to segments of your audience, measure which performs better, and roll out the winner. Unlike personalization, which targets specific visitor profiles, A/B testing randomly distributes variations based on configurable percentages.

Each A/B test component contains multiple variations with assigned distribution weights. Uniform generates a random number for each visitor to determine which variation they see, ensuring consistent assignment across their session.

Test results flow to your connected analytics. When you have enough data, you can declare a winning variation in the Uniform dashboard. Once published, the winner is shown to all visitors going forward.

For detailed configuration, see the A/B testing guide.

Most personalization tools predate the modern front-end era. They manipulate the DOM directly, which conflicts with virtual DOM-based frameworks like React, Vue, and Svelte. This conflict causes visible content swapping, layout shifts, and flicker that degrades the visitor experience and hurts Core Web Vitals.

Uniform Context is built from the ground up to work within your framework's rendering lifecycle, not against it. Personalization decisions are resolved before the page renders, whether that's on the edge, on the server, or on the client, eliminating flicker entirely.

Context runs on your infrastructure at the CDN edge, delivering personalized responses in under 50ms. This means you keep the performance, scalability, and security characteristics of static sites while adding full personalization and A/B testing capabilities.

Uniform supports edge-side personalization on any edge worker platform, including:

If your CDN has edge compute capabilities, it can run Uniform Context.

Visitor classification data is stored in the visitor's own browser and never transmitted to Uniform's infrastructure. No personally identifiable information is collected, stored, or processed by Uniform. This simplifies compliance with privacy regulations and reduces your data liability surface.

The Context engine can be hydrated with data from any source. Connect your existing CDP, profile API, data warehouse, or any custom data provider to enrich visitor profiles with known customer attributes. This works with commercial platforms like Salesforce, as well as homegrown systems. Data flows one way into the tracker, where it's combined with real-time behavioral data for personalization decisions.

By default, Uniform won't collect visitor activity data until consent is granted. You control when and how consent is captured, integrating with your existing consent management solution. You can also configure different consent levels for different classification activities.

For more details, see the compliance and consent guide.

Context isn't limited to the web. You can power personalized experiences on native mobile devices as well:

  • React Native - Use the Uniform SDK directly in React Native apps, with the same component-based personalization model as web applications.
  • Any mobile platform - For native iOS, Android, Flutter, or other platforms, Uniform provides an SDK-free approach through API middleware that runs personalization logic at the API boundary. Your app sends visitor context via HTTP headers, and receives a fully personalized response. This middleware can be self-hosted on Cloudflare Workers, Akamai EdgeWorkers, or similar platforms.

For more details, see the mobile applications guide.

Context supports a wide range of personalization triggers out of the box:

  • Behavioral - Content engagement, page views, navigation patterns, cart activity, and custom events
  • Geographic - Country, region, and city via CDN-provided GeoIP data
  • Campaign - UTM parameters, referrer data, and marketing campaign attribution
  • A/B test targeting - Use test assignment as an input for further personalization
  • Request context - Cookies, HTTP headers, query strings, and any data available on the incoming request
  • External data - Customer segments, purchase history, loyalty tier, or any attribute from a connected CDP or profile API

These dimensions can be combined freely to create precise targeting rules, from simple geo-based content swaps to sophisticated multi-dimensional personalization strategies.

Last modified
Previous
AI
Next
Canvas