Optimizely: How Road Scholar Effectively Runs a MarTech Stack Feeding Every Medium and Channel with a Lean Team
Most marketing stacks do not fail because a team picked the wrong headline. They fail because content lives in one system, tests live in another, the customer profile lives in a third, and the catalog still needs a human to walk hundreds of pages of proofs. Acquisition and retention share a budget. Mission work does not. Mobile traffic does not convert like a laptop. Last-click reporting keeps funding the channel that closed the form, not the one that started the trip.
That problem is sharper when the audience is older, the offer is experiential, and the organization is a nonprofit. You still have to find people, keep them, and prove that the next experiment is worth the people hours. A Digital Experience Platform (DXP) is supposed to be the answer: one place to publish, test, personalize, and learn. Most teams still end up with a Frankenstein of logins.
What Optimizely Actually Is
Optimizely now markets the suite as the AI Platform for Marketing: content, experimentation, commerce, personalization, data, and an Artificial Intelligence (AI) agent layer. Agent Platform, Agentic Experimentation, and Agentic CMS cover the three paths to growth (marketer experience, customer experience, and agentic experience). It combines the original Optimizely testing company with Episerver’s Content Management System (CMS) business. Episerver acquired Optimizely in 2020 and rebranded the combined company as Optimizely in 2021, then kept buying adjacent products (content operations, a Customer Data Platform (CDP), Business-to-Business (B2B) commerce) rather than asking customers to stitch those jobs together themselves.
According to Optimizely’s published figures, more than 10,000 brands use the platform. The company cites a Forrester-validated 446% platform Return on Investment (ROI). Plans are packaged per customer, and the agent layer is credit-based. Product walkthroughs and customer stories live on the Optimizely YouTube channel.
How Road Scholar Uses the Whole Platform
At Opticon in New York City last week, I sat down and interviewed Mark Fagiano, Associate Vice President of eCommerce at Road Scholar. After reading about how well-run and sophisticated his marketing operations are, I joked that I’d love to have his job one day. He smiled and told me he wasn’t leaving any time soon!
Road Scholar is a nonprofit experiential learning organization (not to be confused with the Rhodes Scholarship program). Road Scholar started 50 years ago in New Hampshire, putting older adults in college dorms for summer courses. Today the organization is global: classroom programs, international trips, a cruise program, and, after 18 months with no in-person programs during the pandemic, a substantial online lecture business. Road Scholar enrolls over 100,000 people a year and continues to publish course guides to target their average participant, who is 50+ years old.
While that target audience still wants a physical catalog, they also browse on phones, convert on desktops, and call a contact center with questions. Marketing has to cover every channel and medium… including Google, display, retargeting, Search Engine Optimization (SEO), email, Connected Television (CTV), direct mail, and those oversized books, while donations usually arrive at enrollment rather than from a separate fundraising funnel. Mark’s brief is acquisition and retention, with a slim nonprofit margin and mission work (scholarships for caregivers, awards for educators) that does not always show up as ROI.
Road Scholar started with Optimizely about ten years ago on the CMS, moving off a homegrown stack because the team still needed customization but also needed support, flexibility, and tools that didn’t have to be invented in-house. As Optimizely absorbed content operations, data, and experimentation into one company, Road Scholar adopted the products on offer: the content marketing platform, the customer data platform, experimentation, and the Agent Platform (formerly Optimizely Opal) baked into the suite.
Experimentation helps them find blockers on the site and ship fixes in code or in the platform without waiting on a quarterly redesign. Mark said rapid tests across landing pages and copy have meaningfully raised conversion. The content marketing platform is how marketing, communications, and the program team share one flow for editing, comments, and approvals, which matters when the artifact is a catalog with a lot of pages to prove.
The stubborn gap was mobile. Most traffic arrives on a phone. Most conversions still happen on a laptop. Leads are almost entirely mobile. Through testing, the team found the checkout was not built for a thumb, then rebuilt the flow.
We’ve doubled our conversion rate through mobile over just the last two years.
Mark Fagiano, Associate Vice President of eCommerce, Road Scholar
People still browse on mobile and finish on desktop. That is fine if the phone can capture the lead and the checkout no longer blocks the ones who stay. Mark said the contact center used to take most reservations. Over about ten years, that flipped to the website, which is the point of a CMS you can actually experiment on: the customer self-serves, and the contact center spends time on the questions only a human can answer.
Over the last 18 months, Road Scholar pushed AI agents through marketing and communications until it was daily work, then started handing it to other teams. They built agents so a lean staff can cover more of the mission, including gaps in programming and trends in pricing or enrollment. They also defined four participant personas and use synthetic versions of those personas inside the Agent Platform to pressure-test copy and landing pages before a human has to.
Opal is in everyday use. We built a lot of agents around Opal, so we can make sure that we’re getting the most efficiency out of the people that we do have.
Mark Fagiano, Associate Vice President of eCommerce, Road Scholar
Attribution is not inside Optimizely alone. Demand hits the Customer Relationship Management (CRM) system, then a third-party model that assigns value across touches and devices. Mark’s point is the same one every multi-channel team eventually learns: last click starves the catalog and the first ad, and overfunds the form. The platform’s job is to keep the site, the tests, the profiles, and the content calendar in one system so the model isn’t matching dirty data.
Every Product in the Optimizely Suite
The suite is not a single SKU. The marketing site currently lists these products. Road Scholar did not invent this list. Mark said they adopted the products Optimizely offers. Use the list as the suite map, then decide which pieces you actually need.
- Agent Platform: The no-code AI layer across the suite (formerly Optimizely Opal). Orchestrate agents and virtual teammates for search, personalization, experimentation, and operations, with human review.
- Analytics: Warehouse-native experimentation and product analytics. The pitch is that you do not copy the warehouse into another silo. Built for product managers, analysts, and marketers who need the Stats Engine next to the tests.
- Campaign: Still sold, even though it is not on the main product grid. Agentic email and omnichannel journeys, with segments from the data platform and connectors into Salesforce and Dynamics.
- Commerce: Content, commerce, and experimentation for B2B and Business-to-Consumer (B2C). Configured Commerce is the SaaS B2B product (quoting, account pricing, self-service; formerly Insite). Commerce Connect is the headless API path with the CMS for custom brand experiences.
- Content Management System: The CMS (marketed as an agentic CMS) for creating, publishing, and optimizing experiences without waiting on a developer for every change. Software as a Service (SaaS) and Platform as a Service (PaaS) are both still sold. Visual editing, content modeling, edge delivery, embedded Digital Asset Management (DAM), and AI translation sit here. This is the product Road Scholar started on.
- Content Marketing Platform: Plan, brief, collaborate, and ship campaigns. Calendars, workflows, and approvals in one place. Formerly Welcome. This is the system Mark described for catalog proofs across marketing, communications, and the program team.
- Customer Data Platform (CDP): Unify profiles, score propensity and churn, build real-time segments, and activate them into the CMS, experiments, and ads. Listed on pricing even when it is not on the nine-up product hub.
- Digital Asset Management (DAM): A single repository with AI tagging, brand rules, and reuse, native to the CMS and to experiments. For content and marketing ops. Optimizely’s published Road Scholar story puts tens of thousands of assets here.
- Experimentation: A/B, multivariate, bandits, a visual web editor, and server-side tests. Web Experimentation URLs now resolve to this product. This is the engine Road Scholar used to find mobile checkout blockers and to raise conversion on landing pages without a full rebuild.
- Feature Management: Flags, progressive rollout, kill switches, runtime config, and SDKs for backend, mobile, and edge. Feature Experimentation URLs resolve here. This is the product/engineering counterpart to the marketer-facing experiment tools.
- Personalization: Real-time one-to-one changes (recommendations, on-site swaps, geo, social proof, abandon recovery) with the same Stats Engine as testing.
Taken together, that is content to plan and store, a CMS to publish, tests and flags to change the experience, personalization to tailor it, commerce to transact, a data platform to know the person, analytics to prove it, campaign to message them, and an agent layer to do more of the busywork. Graph (content query and delivery for headless and AI) sits under the CMS, not on the product grid. Content Intelligence still appears in data-platform docs. Neither is a flagship SKU you should shop as a standalone logo.
How the Pieces Connect
The intended loop is simple to say and hard to buy as separate tools. The content marketing platform plans and creates. Digital asset management stores the files. The CMS publishes. Experimentation and personalization change what people see. Commerce takes the order. The data platform unifies the audience. Analytics argues about the result. The agent platform sits across those products and, through connectors, across much of the rest of the stack.
That is why Mark keeps returning to one system. Road Scholar still has a contact center and a third-party attribution model. The website, tests, profiles, and catalog workflow don’t have to be four vendors with four identity graphs.
If you are starting out, you do not have to turn on every product on day one. Road Scholar did not. They bought the CMS, then grew into the rest as Optimizely grew into a platform. Ask for a demo of the suite, then turn on the piece that is actually blocking conversion, whether that is mobile checkout, catalog approvals, or a profile you cannot activate.
