Personalizing your content

Context

Personalization delivers a known experience to a known audience. SitecoreAI has two native ways to do it, and they are not interchangeable.

Page personalization changes several parts of a page together for a defined audience.
Component personalization changes a single component while the rest of the page stays put. It is the one that scales across many pages.

Typical work is banners, calls to action, hero content, and message match from a campaign to its landing page.

Before you build either kind, confirm that:

Your site uses Content SDK.
The analytics identifier is enabled. See Manage personalization and analytics for sites.
Goals and KPIs are agreed.
The site scope is set correctly for the languages or sites you want to measure. You can calculate analytics and personalization for one site language, all languages of a site, or several sites together.

Some personalization features, including component personalization and the Code Assistant, are part of a phased rollout. Your organization may not see them until your environment is included. Check what is available before you plan around them.

Execution

Which one to use

The first question is not about the tool. It is whether you already know the answer or are finding out. If you know German visitors convert better with local pricing, that is a rule, and this recipe applies. If you do not know whether one headline beats another, that is a test. Use Testing experiences instead.

Once you know it is a rule, the choice between page and component comes down to how much of the page changes and how many pages need it.

Question	Page personalization	Component personalization
How much of the page changes?	Several components move together as one experience	One component changes, the rest of the page stays put
How many pages need it?	One page, or a small named set	The same component repeats across many pages
Where is it built?	Page builder, directly on the page	The component's decision table
Condition type used	Out-of-the-box or custom conditions built into the audience	Decision table inputs: profile data, out-of-the-box or custom conditions, or custom values
Can inputs be combined?	Yes, with AND/OR and condition groups	Yes, every input on a row combines with AND
Condition limit	Maximum of five conditions per audience	No five-condition limit, but rule order carries the priority logic
What tells you it worked?	Page performance dashboard, variant against original	Analytics once the personalization goes live
When is it live?	When the variant is published	When the rules are published
What happens after?	Stays live as the standing experience	Stays live as the standing experience

The practical difference sits in the condition system rather than on screen. Page personalization runs conditions straight in the audience builder, which is fast but limited to five conditions per audience. Component personalization goes through a decision table, and custom values are defined there. A rule such as "visitor is high intent based on session count", where the input is not available as a built-in condition, is usually a component personalization job. That, more than the visual scope of the change, is what normally decides it.

Two rules apply to both:

A page runs a personalization or an A/B/n test, not both at once. Decide which question you are answering before you build.
A default page supports up to 10 page variants.

If you deliver your front end without the Content SDK or without Page builder, or you want the same experience in a mobile app or kiosk, see headless personalization and A/B/n testing.

Steps to a personalized page

#	Title	What you do
1	Personalize strategy	Work out which audiences gain value from personalization and which pages are the focus. With a backlog of ideas, prioritize audiences by the volume they generate on the site and how valuable they are to the business.
2	Select a page to personalize	Choose the page you want to build personalized content for. Start with simple component swaps such as banners, calls to action, or hero images.
3	Define audience groups	Use the audience builder and the out-of-the-box conditions to specify who sees the personalized content. Conditions might include behavior or demographics. A social campaign with a specific call to action, for example, can be reflected on the landing page so the journey stays continuous. Remember the five-condition limit per audience.
4	Create page variants	For each audience group, create a variant carrying the personalized content: components added or removed, images and copy tailored to that audience.
5	Preview and test	Use preview to check how the content appears to each audience group and confirm that everything behaves as expected.
6	Publish	Publish the page. The personalized content goes live and is served according to the rules you defined.

If a use case is more complex than the current conditions allow, check whether the logic belongs in a decision table on a component before concluding it cannot be built.

Review Personalize for the in-product steps.

Steps to a personalized component

#	Title	What you do
1	Confirm it is a component job	The change touches one component, or the same component pattern repeats across several pages, or the rule needs a custom value that the audience builder cannot hold.
2	Define the inputs	Decide what visitor data the table evaluates: profile attributes, out-of-the-box conditions, custom conditions, or custom values. Inputs are the vocabulary the rules are written in, so agree them before you write rules.
3	Build the rules	Each rule combines its inputs with AND and maps to one output: the component variant that shows when the rule matches.
4	Order the rules	Rules are evaluated top to bottom and the first match wins. If a visitor matches two rules, only the first rule's variant shows. Rule order is where priority between competing signals lives, so agree the priority deliberately rather than letting it fall out of the build order.
5	Configure the variants	Set up the component variants in Page builder. A visitor who matches no rule sees the original component, so make sure the original is a sound default.
6	Preview, publish, measure	Check each variant renders correctly with real content, test the rules against a profile, publish, then read the impact in analytics.

Two points are worth stating plainly to stakeholders:

Put anything reusable into a custom condition from the first decisioning use case, not the fifth. Logic copied into many decision tables drifts apart as the number of tables grows.
The visitor who matches nothing gets the original component. The original is part of the design, not a fallback nobody looked at.

Review Component personalization and Decision tables for the in-product steps.

Planning decides the outcome

For each prioritized audience, confirm:

How the audience will be recognized in the product.
Whether the change is a page job or a component job.
What content is needed and what the call to action is.
Which dependencies sit with marketing, technical teams, campaigns, or UTM setup.

Prioritize high-traffic entry points (homepage, category landing pages, product detail pages) and conversion-critical pages (pricing, sign-up flows, lead forms) first. Plot audiences by the traffic they generate against their value to the business. The ones high on both come first.

Two habits carry the rest of the programme:

When a test has declared a winner, fold the winning version into the decision table or the personalization rules. Do not leave the test running at 100/0 traffic as a permanent workaround, because a winner parked in a test is logic nobody governs.
When the same targeting keeps reappearing, define it once as a custom condition and reuse it rather than rebuilding the rule on each page.

Insights

Starter use cases

Use these as a backlog seed. The phase column follows the crawl, walk, run model in Conversion optimization strategy.

Category	Phase	Use case	Audience	Goal and KPI	Tool
Campaign	Crawl	Mobile-optimized campaign experience	Visitors on mobile devices	Increase conversion rate. KPI: signup conversion rate, engagement uplift	Page
Campaign	Crawl	Day-of-week promotional banner	Visitors arriving on specific days	Increase click-through. KPI: purchase, conversion rate	Component
Campaign	Crawl	First-time paid traffic incentive	New visitors from paid campaigns via UTM	Grow the subscriber base. KPI: signup conversion rate	Page
NGO	Crawl	Regionalized fundraising content	Donors in EU, NA, and Asia	Increase engagement and funds. KPI: lower bounce, more time on page, higher conversion	Page
Commerce	Crawl	Affinity-driven homepage personalization	Wholesale and retail customers	Shorten the path to purchase. KPI: add-to-cart volume and conversion rate	Component
Commerce	Walk	AI-assisted persona personalization	Shoppers segmented by inferred intent	Shorten time to market. KPI: repeat visits, higher basket values	Component
Loyalty	Walk	Tier-based member experience	Logged-in members by loyalty tier	Increase member satisfaction and time on site. KPI: repeat visits, basket value	Component
B2B	Walk	Re-engage playbook downloaders	Visitors who downloaded a playbook in the past seven days	Turn leads into prospects. KPI: demo booking form submissions	Component
B2B	Walk	Industry-adaptive hero and content order	Visitors with an inferred or declared industry	Increase relevance and conversion. KPI: weighted conversion, engaged sessions	Page and component
B2B	Walk	Customer against prospect split	Known customers identified by login or email match	Route each group to the right next step. KPI: support deflection, demo requests	Component
B2B	Run	Next-best-content recommendations	Visitors with an established affinity profile	Increase depth of engagement. KPI: click-through, downstream conversion	Component

Four of these need technical groundwork before the personalization itself:

Regionalized fundraising content depends on geolocation captured through the SDK.
Tier-based member experiences need the member tier passed in as customer data, plus a custom condition.
The B2B re-engagement case needs custom event tracking for downloads, plus a condition for events within a time window.
The customer against prospect split depends on identity resolution being in place, not just an event.

Personalize pages with AI

With content customized for different audiences, visitors see content that matches their expectations. For example, you can use the AI capability of Sitecore Stream to rewrite text written in American English into British English. Review Personalize pages with Stream for more.

Using affinities in personalization

Testing experiences

Conversion optimization strategy

Delivering components through Design Studio

Personalize

Component personalization

Specifying variables for conditions

Manage personalization and analytics for sites