Opal AI in Action: Building Marketing Solutions with Optimizely Agents

Leveraging Optimizely’s Opal AI for Marketing Innovation

The latest version of Optimizely’s Opal AI is designed to orchestrate teams of AI agents across the full spectrum of Optimizely’s marketing platforms and modules. This powerful capability allows marketing teams to rapidly build custom agents and develop agentic workflows without the lengthy and resource-intensive process of collaborating with development teams over several months.

Opal AI has clearly demonstrated its value in key areas such as content personalization and large-scale content creation. Its support for custom code using tools enables marketing teams to address functions that may not be available out-of-the-box, opening the door to a broader range of applications and innovative solutions.

Recently, our team at Rightpoint participated in the 2025 Opal Innovation Challenge. As part of this global initiative, Optimizely’s partners, including Rightpoint were tasked with designing new solutions to solve real-world marketing challenges using Opal AI. Our hackathon team consisted of myself, fellow solutions architect Douglas Couto, VP and Solutions Lead Ben Rothman, and Kelly Noah, Rightpoint’s Senior Design Director and Head of Content Design.

Developing an Automated UX Audit Solution with Opal AI

With Opal’s agentic capabilities launching in mid-2025, our team approached the platform as a typical marketing team would by learning as we went. We set out to address a specific marketing challenge: the labor-intensive process of conducting e-commerce page user experience (UX) audits. Traditionally, this manual task involves combining data on page abandonment, established best practices, and page performance to identify and request UX improvements that could drive higher conversions. This process typically consumes several days each month for a product owner or analytics manager, resulting in significant opportunity cost. Furthermore, since audits are usually performed monthly, UX issues can persist unaddressed for weeks at a time.

We set out to build an automated UX audit tool that would eliminate manual triage, enable on-demand audits, and speed up page optimization. Our Opal AI agents pull data from GA4, Google Lighthouse, and Baymard Institute, coordinating with Azure DevOps to identify opportunities.

This solution required us to refine prompts, build custom tools and construct an agentic workflow capable of producing HTML-friendly output, as well as learn how to effectively transfer data between agents to ensure a seamless process. We also developed prompts to guide each agent in problem-solving and established guardrails with clear instructions to prevent hallucinations.

Our team built two custom tools: the 'GA4 analyzer' using REACT deployed to Cloudflare Workers and the 'Azure DevOps integrator' in Python on Google Cloud.

The result was the Rightpoint Frictionless Funnel Agent Workflow, which orchestrates a team of Opal AI agents to:

• Extract 30 days of page abandonment data from Google Analytics 4, pinpointing where users exit their journeys on specific pages.

• Audit the technical performance of these pages using Google Lighthouse.

• Compare flagged pages against Baymard’s UX best practices.

• Synthesize the collected data to identify and prioritize page enhancements as user stories.

• Create Azure DevOps tickets to initiate page changes.

• Notify the product owner when a user story request is generated.

• Complete all these steps within minutes.

This streamlined agentic workflow lets users quickly instruct Opal AI to run this workflow, replacing hours of manual reporting. Scheduling enables automatic operation without chat requests, freeing product owners for strategic tasks and accelerating page and UX enhancements.

Lessons from the Opal Innovation Challenge

Intuitive

Building the Frictionless Funnel Agent proved Opal AI is easy to use and users with basic agentic knowledge can create workflows without deep technical expertise.

Extensible

Opal doesn't come with all agents by default, but it's easy to build new ones using the built-in tools or add tools to expand its features with basic coding.

Efficient

Our team built our agent in days, not weeks or months. Opal AI’s efficiency, extensibility, and intuitiveness together make it an effective tool for automating and streamlining workflows.

Your partner in creating agentic AI experiences

Marketing teams can also use Opal AI to build and automate much more complex workflows than the Frictionless Funnel Agent. To leverage the full power of Opal across the Optimizely ecosystem, it helps to work with an experienced partner. As one of just five Optimizely Premier Platinum Partners worldwide, Rightpoint has close to a decade of experience helping clients get the most value from their investment in Optimizely.

Contact us to learn more about how Opal AI can help solve your marketing challenges with agentic AI solutions.