Rebuilding a Website for Answer-Based Content: Lessons from a Marketing VP

By Calvin Scharffs

After decades of a market that favored pithy website content, AI upended everything. Like many VPs of Marketing and CMOs, I woke up to a new reality that seemed to emerge overnight. Visibility—the currency we all worked so hard to amass—was vanishing. As marketers, we understood that unless we adapted quickly, our website would offer zero presence in the upper funnel. That, in turn, would mean our entire demand engine would solely depend on the sales team. For anyone responsible for their company’s new customer pipeline, that isn’t just risky, it’s unsustainable.

Looking around, I saw no shortage of articles explaining why this happened or touting the rise of answer-based content. What I couldn’t find was practical help. How do you actually do it? How do you turn a polished but sparse website into one that answers real customer questions and earns visibility in AI-driven search?

I wrote this for other leaders carrying pipeline responsibility, because when the upper funnel disappears, you don’t just lose visibility on the web, you lose the ability to predict growth reliably. 

Getting into the answer-based mindset required us to strip our site back to the first principles of customer relationships. We stopped worrying about sounding clever and started asking: what are our clients actually trying to solve, and what proof do we have that we’ve solved it before? Simply asking that question laid out for us the work we needed to do.

The Listening Tour

We started by interviewing clients, but not with quick check-ins. We engaged in-depth conversations about the challenges they face day-to-day and how those challenges shape their plans. Then we asked how they saw us solving those problems. Our internal teams, who worked with these clients, were on hand to provide us with the technical insights and field-tested strategies they deployed to solve those challenges.

The case studies that emerged became gifts that keep on giving. Obviously, they gave us a more nuanced understanding of our value from the customer’s perspective. They also shaped the foundation of our revised website copy. And later, when we built a buyer’s guide, those same client challenges formed the basis of what we told prospects to ask any vendor.

The Buzz Word Rebellion

After the listening tour, we took a hard look at our site and saw plenty of words that rang hollow. In our business, account support is critical, and we—like thousands of others—described ours as “bespoke.” But what does that actually mean? 

To answer that, we interviewed our internal teams about their day-to-day work and the behind-the-scenes steps required to keep client projects humming along. Out of that came our onboarding guide, which helps customers understand exactly what to expect once an I/O is signed. It also gave us sharper, more grounded content for our “how we work with you” section of the site.

And that’s when it clicked: shifting from pithy copy to content that shows how we deliver is one of the most valuable moves we can make to build the upper funnel in today’s world of AI Overviews. The differentiator isn't words like “bespoke.” It’s demonstrating that tangible processes are in place to deliver value.

Why Are You Spending Cycles on This?

Another serendipitous moment came when a member of our executive team asked me point-blank: “Why are you spending cycles on the website?” That question forced me to pause and rethink. Instead of treating it as a tactical refresh, my team and I thought about all that we wanted to accomplish. We then reframed the project around seven strategic goals we identified:

1. Speak directly to the reader’s concerns and persona
2. Show how we actually support clients
3. Be honest about the problems our clients brought to us, and the ones we encountered along the way, and be specific in how we solved them
4. Make each page generative engine optimization (GEO) friendly so we can begin amassing visibility once again
5. Make the site a “discoverable” experience with lots of resources so that people can learn about the industry
6. Showcase our team’s expertise without falling back on “trust us, we’re experts” language
7. Demonstrate the degree to which we take ownership for getting the work done

While those goals were good they were still abstract. To achieve them we needed to define each goal in concrete terms (for example, requiring every page to open by naming the persona it serves). From there, we mapped the specific page elements that demonstrated each goal in practice. This resulted in detailed checklists for each goal. For instance, we asked: Does this page explain what we do in our customers’ own words? Does every claim have tangible proof to back it up, such a case study, a quote, a reference? Does it show, not just state, how we solve problems?

The Goals Checklist

At this stage, we began to rely heavily on AI tools. With our first drafts in hand, we uploaded our goals checklist to Microsoft Co-Pilot and then ran each draft through it with the question: “Does this draft meet each goal, and if not, what’s missing?” 

Generative AI is, at its best, a word calculator, and in this case, it did its job well. It flagged where we lacked elements or where we had them but failed to express them clearly enough to meet the goal. Each round produced a punch list of fixes, which became the basis for our next draft. In all honesty, this was a bit tedious, but it was a tremendously fruitful endeavor.

From SEO to AEO to GEO

We approached optimization as a progression. First, we optimized SEO, ensuring the fundamentals were strong so the site remained discoverable in traditional search. Then we layered in AEO, rewriting content around clear questions and answers so visitors and search engines alike could understand the value we deliver.

The last step was to stress-test everything for GEO. Would an AI model understand what this page was about and surface it as an answer? Was it structured in a way that lined up with how LLMs read? 

Again, the AI tools were invaluable here. We uploaded the text and asked for a rating on a scale of 1 to 10, along with a detailed list of what’s working and what’s not. Each round produced a list of fixes, and yes, it got a bit tedious, but we inched up towards near-perfect scores. 

And here’s the ultimate validation: when our internal AI Council ran the revised site through its copy evaluation framework—built to score website copy, SEO, and overall quality— our copy came back a perfect 10 out of 10. That wasn’t just a vanity metric; it was concrete evidence that the process gave us clearer messaging, stronger SEO and GEO signals, and pages that AI models actually understand.

Impact on Visibility

Within a few months of publishing the new answer-based and GEO-aligned content, our search visibility grew in ways we had not seen in years. Total keyword coverage more than doubled, average ranking improved by more than four positions, and top-ten placements nearly quadrupled. Just as important, the gains concentrated around the high-intent terms that matter most for our service areas.

That shift has proven more valuable than raw volume. Inbound leads are stronger and more consistent because prospects are finding us through the exact problems we solve. Paired with new efforts to highlight our team’s thought leadership, the site now produces a steady stream of qualified interest in AI readiness, GEO strategy, and performance marketing. The impact has been far greater than that of a traditional website refresh, confirming that structured content and entity-focused optimization change how search engines and AI systems understand our expertise.

In terms of AI visibility,  the site received an AI Visibility score of 35 (out of 100) with 92 of its pages cited and multiple mentions inside generative panels. This level of AI sourcing is well above the normal benchmarks for B2B service websites and confirms that structured, entity-driven content is becoming a trusted input for AI engines.

The Payoff

We can bemoan the loss of visibility, but what AI-driven search has really done is take away the shortcuts we marketers have long relied on. Answer-based content forces us to articulate real value from the customer’s perspective and to back it up with proof. That’s a good thing.

Yes, the process can be onerous at times, but to me it’s one of the most valuable exercises a brand can go through. We ended up with better website copy. More importantly, we ended up with a clearer understanding of our customers, sharper messaging, and a site built to stand up in a world where both humans and machines decide what content gets seen.

Calvin Scharffs is the VP of marketing at Orange 142