Optimizely Mark and AI features
25 min
Mark (Optimizely Agent Platform) is an agent orchestration platform that helps you work smarter across Optimizely One. Mark is directly connected to your workflows, fully aware of your brand, and powered by hundreds of tools to assist with marketing and digital tasks. Usage of Mark features incurs credit charges.
In addition to Mark, Optimizely continues to develop various AI and machine learning (ML) features embedded across its product suite. These features do not incur Mark credit charges. For information on the differences between Mark and generative AI, see the blog article, Mark vs Regular GenAI: A Deep Dive.
For users on Opti ID, administrators can turn off all applicable Mark and AI features from Admin Center > Settings > Generative AI. Users without Opti ID are not eligible for Mark. To turn off a product's AI features, refer to the documentation for that product.
For a full list of Mark and included AI and ML features for specific products, see the following links.
Usage and billing update
Effective May 7, 2025, access to Mark features across Content Marketing Platform, Web Experimentation, Feature Experimentation, Personalization, Content Management System (SaaS), Collaboration, and Optimizely Data Platform will transition to a credit-based usage and billing model.
Mark
Usage of Mark features incurs credit charges.
Prerequisites
- You must use Opti ID to access Mark.
- You must have generative AI enabled in Optimizely.
Mark app
Mark administrators can extend Mark's base capabilities and
- Create custom skills for their organizations.
- Create custom agents with specialized agents.
- Add system tools.
- Create custom tools.
- Add connector tools to Mark from third-parties through Optimizely Connect Platform (OCP).
- View Mark agent usage using Optimizely Reporting.
Mark users and administrators can
- Chat with Mark and view their chat history.
Optimizely Analytics
- Analyze an Agent Visibility Analytics dashboard to get a report of trends, anomalies, and insights from your dashboard data using the Agent Visibility Analytics Insights agent.
Campaign
- Translate the full content (body, subject, preview, and alt-texts) of a specific email using the Email Content Translation agent.
- Send Mark newsletter content from Optimizely Campaign to get structured feedback and detailed recommendations for improvement using the Email Optimization agent.
- Create three highly effective pairs of subject lines and preview texts (focusing on urgency, curiosity, and benefit) using the Subject & Preview Text Ideation agent.
- Analyze provided email content to suggest accessible alt-text and titles for the images included using the Image Metadata Ideation agent.
- Analyze and compares email campaign performance using key performance indicators (KPIs) to identify success factors and performance differences with the Email Performance Comparison agent.
- Generate a standardized campaign brief from a few inputs using the Campaign Brief Generation agent.
- Generate a complete Smart Campaign flow architecture, including node sequences, A/B test variants, timing, and a master send schedule, from a campaign brief using the Smart Campaign Design agent.
- Analyze predictive eye-tracking maps to identify visual hierarchy issues and get prioritized layout recommendations for your email designs using the Eye Tracking Auditor agent.
- Get a visual, step-by-step guide for building a target group using the Target Group Guidance agent.
- Get a real-time dashboard of your active Smart Campaigns, with previews of upcoming messages and recommended send times, using the Campaign Status Insights agent.
- Summarize the performance of your sent Smart Campaigns, flag unusual results, and get recommended follow-up actions using the Campaign Performance Digest agent.
Collaboration
- Use Mark to write experimentation plans and suggest hypotheses and variations.
- Use AI to refine variations and written briefs.
Commerce Connect
- Build and configure product promotions using the Product Promotion agent.
Configured Commerce
- Create product restriction groups in Configured Commerce through a step-by-step wizard using the Restriction Group Creation agent.
- Create global lists or wishlists in Configured Commerce through a step-by-step conversation using the List Creation agent.
- Create promotions in Configured Commerce through a step-by-step conversation, including rules, rewards, and website assignments, using the Promotion Creation agent.
Content Marketing Platform (CMP)
- Generate AI content for a task.
- Get a curated list of keywords to maximize content visibility and search traffic using the Keyword Research agent.
Content Management System (PaaS)
CMS 12
CMS 13
- Build a content model from a URL or an image using the Content Model Creation agent for CMS 13.
- Enhance your content's visibility and LLM discoverability by automatically identifying opportunities for structured data markup using the GEO Schema Optimization agent for CMS 13.
- Analyze your CMS 13 webpages and generate optimized SEO metadata, including meta titles, descriptions, and OpenGraph tags using the SEO Metadata Implementation agent for CMS 13.
- Analyze a document, image, or URL and creates a content item of a specified content type automatically using the Page Builder agent for CMS 13.
- Analyze existing content of a specified CMS 13 (PaaS) content type and produce a markdown skill for creating content of that type using the Content Analysis agent for CMS 13.
Content Management System (SaaS)
- Create content models automatically by having Mark analyze URLs or images using the Content Model Creation agent (CMS SaaS).
- Identify, sort by last edit date, and help resolve duplicate and outdated content (in terms of content guidelines as reference) across a website in CMS (SaaS) using the Content Analysis agent (CMS SaaS) agent.
- Enhance the visibility and Large Language Model (LLM) discoverability of your CMS (SaaS) content by automatically identifying opportunities for structured data markup with the GEO Schema Optimization agent (CMS SaaS).
- Evaluate existing SEO properties, identify optimization opportunities, recommend improvements, and upon your approval, create or update SEO metadata fields in your CMS (SaaS) instance with the SEO Metadata Implementation agent (CMS SaaS).
- Analyze document, image, or URL and creates a content item of a specified content type using the Page Builder agent (CMS SaaS).
- Analyze CMS (SaaS) content of a specified content type to identify structural patterns and produce a reusable skill using the Content Analysis agent (CMS SaaS).
- Audit your CMS (SaaS) content against Web Content Accessibility Guidelines (WCAG) standards and apply approved accessibility fixes using the WCAG Compliance Auditor agent.
- Work with blueprints your team already maintains or create new ones by asking Mark which blueprints exist or whether you have one for a particular purpose. See Work with Mark in the CMS (SaaS) blueprint documentation.
- Ask Mark to list your existing categories or look up a specific one. Mark can also create, update, and remove categories, including nested categories under a parent. Describe the organization you want instead of configuring each category by hand. See Ask Mark to maintain categories.
Content Recommendations
- Use the system tools to retrieve the highest-performing content items, gather performance metrics for the most engaging topics, and analyze the performance of the recommendation widget.
Feature Experimentation
- Brainstorm ideas for flag variables.
- Brainstorm ideas for flag variations.
- Run a pre-launch experiment review that examines your A/B test configuration and provides recommended changes to maximize your odds of reaching statistical significance using the Experiment Review agent.
- Use the Experimentation Program Overview agent to create a report on your Feature Experimentation program performance, including top experiment and win rate data.
- Compare competitor webpages against your own using the Competitive Webpage Analysis agent.
- Detect conflicts between a proposed experiment and any live experiments in the same project with the Experiment Conflict Checker agent.
- Use the Experiment Value Estimator agent to generate a stakeholder-ready report with the projected annualized impact of rolling out a winning experiment variation.
- Analyze a Feature Experimentation project and resolve governance findings in chat with the Feature Experimentation Governance agent.
- Generate a portable SKILL.md file to implement a Feature Experimentation flag in a codebase with the Feature Flag Implementation agent.
- Analyze your Feature Experimentation projects and view an interactive dashboard with your projects health using the Experiment Program Health Review agent.
- Generate an executive value report from Optimizely Program Reporting for Experimentation using the Experiment Program Value Analysis agent.
Opti ID Admin Center
- Monitor Mark credit billing and tracking.
- Receive Mark credit usage notifications.
Optimizely Data Platform (ODP)
- Get answers about an ODP real-time audience, its purpose, how it overlaps with other audiences to achieve your stated goal, and other general troubleshooting queries, or create and update real-time audiences, using the Real-Time Audience Builder agent
- Use the Standard Audience Builder agent to build and troubleshoot ODP audiences through natural language.
ODP + Optimizely Content Recommendations
Note
Mark is available only to United States–based ODP + Content Recommendations customers.
- Enable generative AI topic enrichment for customer records.
Personalization
- Generate copy suggestions for variations in the Visual Editor.
- Modify and update existing website elements, create new ones, and generate and apply enhancement suggestions using the AI variation development agent.
Product Information Management (PIM)
- Use tools to retrieve general statistics from the Dashboard.
Product Recommendations
- Use the system tools to do the following:
- Configure and manage your high-level workspace, including selecting the active site, regional settings, and foundational configurations.
- View and retrieve the algorithms, categories, attributes, and product sets available for configuring recommendations on your site.
- Define the layout and scaffolding of your website to prepare it for recommendation widgets.
- Create, review, and manage the actual recommendation blocks (widgets) shown to the end user.
- Analyze the success, engagement, and revenue impact of your product recommendations.
Web Experimentation
- Generate copy suggestions for variations in the Visual Editor.
- Modify and update existing website elements, create new ones, and generate and apply enhancement suggestions using the AI variation development agent.
- Run a pre-launch experiment review that checks your experiment configuration and sends recommended changes to maximize your odds of reaching statistical significance using the Experiment Review Agent.
- Use the Experimentation Program Overview agent to create a report on your Web Experimentation program performance, including top experiment and win rate data.
- Use Idea builder to generate specific, actionable experiment ideas based on your page, goal, and research using Mark.
- Compare competitor webpages against your own using the Competitive Webpage Analysis agent.
- Detect conflicts between a proposed experiment and any live experiments in the same project with the Experiment Conflict Checker agent.
- Use the Experiment Value Estimator agent to generate a stakeholder-ready report with the projected annualized impact of rolling out a winning experiment variation.
- Generate an executive value report from Optimizely Program Reporting for Experimentation using the Experiment Program Value Analysis agent .
MCP servers
Model Context Protocol (MCP) is an open standard that connects AI clients to external tools and data. Optimizely MCP servers let your AI client access your Optimizely data and act on it through natural-language requests, so you can query and manage your Optimizely products without leaving the client.
You need an MCP-compatible AI client, such as Claude Desktop, Cursor, or Visual Studio Code. Select a server from the following list for connection instructions:
Included generative AI features
These features do not incur Mark credit charges.
Commerce Connect
Note
When Mark is available for Commerce Connect, these features may transition and incur credit charges.
- Generate text for a product property.
Content Marketing Platform
- Generate AI images from a task.
- Ask AI to write marketing plans and suggest meta-descriptions, keywords, and so on. (For only customers who are not on Opti ID.)
Collaboration
- Bring your own AI (BYOAI) to deploy your own LLM or AI engines to generate content.
Product Information Management (PIM)
Note
When Mark is available for Product Information Management, these features may transition and incur credit charges.
Included machine learning features
These features do not incur Mark credit charges.
Content Recommendations
- Uses Natural Language Processing (NLP) to match the content that shares the greatest similarity of an individual user's profile in real time.
- Identify content gaps and saturation points using NLP with the Insights dashboard.
- Generate real-time interest profiles based on content consumption.
Digital Asset Management (DAM)
- Automatically suggests tags for uploaded images based on machine-learning algorithms.
- Generate cropped variations of images in the library using an automatic smart focal point.
Feature Experimentation
- Uses Stats Engine algorithms to reduce errors and get faster experiment results.
- Uses Stats Accelerator machine-learning algorithms to manipulate traffic distribution and minimize time to statistical significance.
- Uses a multi-arm bandit (MAB) intelligent algorithm to identify top-performing variations and funnel more visitors to the best-performing variation to drive higher conversions.
Optimizely Data Platform
- Prevent churn and win back customers using ODP's customer win-back model, which shows the relationship between days since engagement and the likelihood of remaining your customer.
- Use the Predicted Time Until Next Order report to evaluate customer data from the last 180 days to predict a customer's likelihood to order in the next 42 days. This report uses the Order Likelihood and Days Until Next Order modeled attributes.
Optimizely Graph
- Enable semantic search to return similar vocabulary results.
Personalization
- Use Stats Engine algorithms to reduce errors and get faster experiment results.
- Use Stats Accelerator machine-learning algorithms to manipulate traffic distribution and minimize time to statistical significance.
- Use a multi-arm bandit (MAB) intelligent algorithm to identify top-performing variations and funnel more visitors to the best-performing variation to drive higher conversions.
- Create a predicted intent audience condition using adaptive audiences that capture a percentage of your visitors most interested in a specific topic.
Product Recommendations
- Uses Optimizely's algorithms to examine visitor behavior and recommend products.
Web Experimentation
- Use Stats Engine algorithms to reduce errors and get faster experiment results.
- Use Stats Accelerator machine-learning algorithms to manipulate traffic distribution and minimize time to statistical significance.
- Use a multi-arm bandit (MAB) intelligent algorithm to identify top-performing variations and funnel more visitors to the best-performing variation to drive higher conversions.
- Create a predicted intent audience condition using adaptive audiences that capture a percentage of your visitors most interested in a specific topic.
- Use contextual bandits to deliver personalized variations to each user by dynamically reallocating traffic based on the primary metric and user attributes.
Note
If you use Opti ID, administrators can turn off generative AI in the Opti ID Admin Center. See Turn generative AI off across Optimizely applications.