How Cortland experiments in 30 minutes with Opal agents

One team, 209 websites, and a quest to experiment

Aislinn Doss, Product Analyst for Digital Experience at Cortland, is the owner of DXP experimentation at one of the largest multifamily property companies in the US. Cortland manages 75,000 apartment homes across 200+ communities — each with its own website, its own tour scheduling links, and its own contact forms.

Cortland’s digital experience team sits at the intersection of marketing, data and analytics, technology, and operations. Aislinn’s wider team includes product owners, UI/UX designers, and more, with her being the main owner of Optimizely day in, day out.

Cortland had been using Optimizely Web Experimentation for several years before Aislinn joined. Limited developer resources meant simple website changes took longer than they should, so the platform gave the team a way to move independently. Early wins, including a redesigned tour scheduler form and dynamic slide-out templates with community-specific links, proved the value immediately.

But there was no formal experimentation strategy in place. So, it became a question of how to capitalize on this promising program.

“It was really just me creating the ideas, getting them approved, and moving forward,” Aislinn explains. Leadership placed their trust in Aislinn to steer this program that everyone valued, but couldn’t commit time to.

Her first few months were spent learning the platform through Optimizely Academy, and she found Optimizely’s design easy to get started with. But generating ideas and building variations without a developer background required something more.

Cortland tracks three core conversion points on the website: scheduling a tour, submitting a contact form, and clicking the phone link. Aislinn points out that they “have to focus on those three conversions, as it is nearly impossible to track those who visit the website and go on to sign a lease.” And it is those three conversions that Aislinn’s Opal-powered experimentation program has a direct impact on.

AI agents that changed everything

When Aislinn first encountered Opal, it started with one capability: the Experiment Ideation Agent. She could submit a URL, and Opal would analyze every element on the page and generate experiment ideas.

“I look at the website every day. Some things just don’t pop out to me,” Aislinn says. “Having Opal say, ‘hey, this should be changed’ was like having another set of eyes on the page, another teammate.”

The ideation agent didn’t just surface ideas. It helped Aislinn build hypotheses in language that landed with the broader Cortland team, translating experimentation concepts into strategic business terms. For someone driving a program with limited direction, walking into a meeting with a well-articulated hypothesis made all the difference.

But it was the second agent, the Variation Development Agent, that fundamentally changed what was possible.

As someone without a developer background, Aislinn had always hit a wall when experiments went beyond simple UI tweaks. Changing a button color was easy. Restructuring a slide-out template, adding dynamic links, or building an animation? Either these were simply not possible before, or at least took days of trial and error.

She described a specific scenario: They needed to build a template for pop-up on a personalization campaign. Rather than struggling to rework the code manually, she instructed Opal to remove the image from an existing template, add header copy, create a button, and link it to the right URL. Some back and forth was needed, but the result was delivered in a fraction of the time.

Together, the two agents form a connected workflow. The Ideation Agent generates ideas and hypotheses. The Variation Development Agent builds the changes. Cortland still works with Optimizely's RapidEx team to deliver larger experiments, but the faster, more creative tests? Those are built and shipped by Aislinn and her agents.

Why Opal over another AI tool

A key part of Opal’s value is that it lives inside the platform Aislinn already uses every day. No separate tool to learn, no context-switching, and no need to explain her experimentation setup to a generic AI.

“I was already in the platform day-to-day,” she says. “Just having that tool readily available within the platform — it was a no-brainer.”

Aislinn is already exploring what comes next. She wants to start building custom agents, a sign of how naturally the tool has become part of her workflow rather than an addition to it.

The first agent Aislinn wants to build is one that helps her with regional campaign requests. It would monitor and report performance on email campaigns to dormant leads, sharing these results with regional marketing managers, providing customer profiles, and more. Eventually, she envisions the agent receiving requests itself, segmenting audiences in ODP, and creating the email campaigns.

Experimentation at portfolio scale

A major standout from the project is an Affordability Calculator built in collaboration with RapidEx. It takes a prospect’s move-in date, budget, and bedroom preferences and recommends matching units at that community. The result was a 6.8% lift in tour form submissions — Cortland’s most valuable early-stage conversion. The experiment is now set to run as a permanent experience on the website.

Product Recommendations, launched through a widget redesign, saw a similar story. The original version took up too much space and wasn’t driving clicks, so the team rebuilt it into a more condensed, user-friendly format that shows all the specs at a glance.

Aislinn also uses real-time segments from ODP to trigger contextual experiences. One current test detects when someone spends a few minutes on the FAQ page and surfaces a contact slide-out: “Have more questions? Click here to get in touch.” These campaigns run as always-on personalization across Cortland’s portfolio.

The velocity impact from Opal has been significant, too. Before, Aislinn aimed for one to two experiments per month, but couldn’t always hit that mark working alone. Now, she consistently meets and surpasses that target. The experiments are also more varied, more creative, and more ambitious than before.

An industry-leading DXP

Cortland has undertaken a full DXP transformation to continue expanding their communities and best serve residents. The decision was rooted in what the team has already seen: the measurable success of their experimentation program, the potential in Opal, and Cortland’s own wins with AI elsewhere in the business.

That wider AI story is already well underway and runs against the current of an industry that is notoriously slow to adopt new technologies. Cortland has already launched a virtual leasing agent that responds to prospects by confirming schedules, providing availability, and recommending floor plans. They have another AI tool that handles resident emails and creates tasks for on-site teams when humans are needed in the loop.

In an industry that has historically been slow to modernize, where many of the same processes and systems from 10 to 20 years ago are still in use, Cortland is positioning itself as a digital experience leader. Optimizely as the foundation, they’re not just keeping pace with the shift toward modern, AI-powered digital experiences. They’re driving it.

Aislinn’s success with Opal is part of a much bigger organizational bet on AI across every touchpoint. This is a business that trusts its people to best serve residents, without ever losing that human touch needed in every search for a home.