How GE HealthCare tests for quality leads, not just more clicks

Most experimentation advice is written for consumer businesses: lots of traffic, short purchase cycles and a conversion that happens in a single session. GE HealthCare lives in a different world. Its visitors are healthcare professionals and health system buyers, its journeys can stretch across a year, and a single qualified lead can represent a high-value opportunity.

Tom Gores, Director, Product - Digital Experience & E-commerce at GE HealthCare, has run experimentation programs in both worlds. On this episode of The Experimentation Edge, he tells host Ashley Stirrup what changes when you move from B2C retail to B2B healthcare, and why the most important metric is not always the one that moves fastest.

From socks to capital equipment

Tom's first A/B testing program started about 12 years ago at Kohl's, when testing tools were becoming friendly enough for marketers to use. A tool few people were using landed on his desk, and he quickly realized "the gold mine that we were sitting on." He grew that program, moved into a more technical IT role, and joined GE HealthCare just over 5 years ago. His scope now goes beyond A/B testing to optimization, content authoring and publishing, and UX across a digital ecosystem with GEHealthcare.com at the center.

The contrast with retail is stark. At Kohl's, the job was to capture attention and convert now, with promotions and flashy experiences. In healthcare, buying cycles can run from months to a year, with many visits along the way. "It's not always about the instant click making the sale. It's much more in it for the long game." Some deals are self-service e-commerce; bigger ones, like capital equipment, may start online, move to a sales rep and close offline.

Lower traffic, higher stakes

Tom's team is a centralized experimentation group. Stakeholders across the company bring ideas, often raw ones, and the team turns them into executable tests that can be made "bigger, better" and learned from.

The economics shape everything. "The volume of traffic might be lower than we had in a retail environment, but the stakes are higher." A retail order might be $10 of socks after coupons. At GE HealthCare, "each qualified lead carries much higher stakes."

The test that won on clicks and lost on quality

Early in his time at GE HealthCare, the experimentation culture was less mature than at Kohl's, though people clearly wanted more of it. Like any young program, the team went looking for quick wins. Contact Us form submissions always surfaced as a top KPI, so they added a floating Contact Us badge to the homepage.

Clicks went up significantly. Then the business development teams weighed in: junk leads went up too. "You can get more clicks for the metric that you care about if you put it in someone's face. But does it really mean quality leads? Probably not." Unqualified leads created churn and extra work for sales teams who had to vet them out.

The lesson was about balance: driving the click versus doing what is best for the business. In B2C, Tom says, you can throw ideas at the masses and see what sticks. "In B2B you gotta be a lot more strategic and a little bit more tactical about who you're delivering what."

Making ideas bigger, better and broader

People usually arrive with an idea for their own page. Tom's first move is to "put some quantitative juice behind it": how much traffic does that page get, and how much action happens there?

Then the team asks whether the idea can be generalized. If someone wants to redesign one product page, could the same theme be applied to every product page template? Testing it that way puts the change on a roadmap for the base template and turns an isolated tweak into something that affects thousands of visitors instead of hundreds.

Plan for failure before you launch

For leading indicators, Tom starts with engagement: are customers interacting with the content? Then come form fills that start a lead gen relationship, with different page templates serving different purposes, some for research and some for conversion.

His bigger point is foresight. Before spending time, money and resources to build a test, play out the scenarios. What if it fails? Coming back to leaders with "this didn't work" carries negative energy. Having a plan avoids it: "If it works well, here's the plan. If it doesn't, here's how we're gonna iterate." That keeps experimentation framed as a way to learn and make better decisions. "Oftentimes, you get a win from the thing that you don't end up building."

Where experimentation grows next

Tom sees 3 areas for growth:

• People and culture: keep highlighting big wins and put experimentation at the forefront of big projects, "so that we don't have to go back retroactively."
• Data: "Not everyone has the cleanest data and data systems and data sources that connect." Better foundations enable personalization and more advanced use cases.
• Technology and capabilities: tools like customer data profiles that give one view of a customer across channels, connecting an online visit to the next conversation on another channel.

Being a global company adds complexity, including global privacy laws that limit what can be done and seen. Those, Tom says, are challenges to solve in order to deliver a great experience.

The metrics that matter

As the leader of the whole ecosystem, Tom treats CSAT as a North Star. Relevant, engaging, high-impact experiences all culminate in whether visitors find the site satisfactory and would recommend it. CSAT is measured through a recurring survey with the same questions, compared against the last quarter or year. Alongside it sits conversion rate: how many visitors fill out forms.

The newest metric is discoverability in AI. "Customer behavior is significantly changing toward using AI for research. We gotta make sure our brand shows up when they're performing those prompts." Organic traffic still matters, but positioning content so LLMs surface GE HealthCare is now part of the job.

The takeaway

B2B experimentation rewards patience and precision. Measure the quality of what you drive, not just the volume; make good ideas bigger before you test them; plan for failure up front; and track the long journey with metrics like CSAT and AI visibility, not just the instant click.

🎧 Listen to the full episode →