To kick things off, can you please tell us a bit about your professional journey and how it led you to your current role as SVP of Product Strategy & Operations at Optimizely?
I started my career in analytics, helping brands use data to make better decisions. From there, I moved into the marketing tech space and helped build a predictive analytics startup that was eventually acquired by Optimizely in 2019. Since then, I’ve taken on a mix of product strategy and operations work — everything from M&A and analyst relations to product launches and commercialization. It’s a good mix of big-picture strategy and making sure the day-to-day runs smoothly.
Agentic AI is changing how customers interact with the digital world. How do you see AI agents reshaping the way brands design and optimize their websites?
Most websites today are built for humans. But AI agents don’t care about your hero video or your scroll animation. They want clear webcopy, structured data, clear intent, and fast answers. Brands will need to think in layers: one experience for humans, another for agents. It’s not about rebuilding everything; it’s about surfacing the right data in a way agents can use. For example, if an AI is shopping for a customer, your product details need to be machine-readable, not buried in really complicated marketing jargon. The brands that nail this will win when AI-driven traffic becomes the norm, not the exception.
As AI moves beyond chatbots and starts making decisions on behalf of users, what should businesses be doing now to ensure their sites are not just human-friendly, but AI-friendly too?
Stop assuming all traffic is human. Audit your site like an LLM would: Can it find pricing, specs, or policies without scraping paragraphs? Tools like schema markup help, but the bigger shift is mindset.
If a customer’s agent can’t book a flight or compare products on your site, you’re invisible to a rapidly growing part of the market. This isn’t about just jumping on the AI bandwagon. It’s about recognizing that AI-powered tools are becoming a real proxy or extension for users. Start small: optimize key journeys, like checkout or chat support for both humans and agents, and work from there.
With AI generating a growing share of web traffic, what new metrics or performance indicators should companies start paying attention to?
If an AI hits your site, how fast can it get what it needs? Track things like API response times, structured data completeness, and success rates for agent-driven tasks. Bounce rate doesn’t matter anymore, AI won’t “bounce,” it will just stall. Also, watch for traffic patterns. If your website suddenly gets hit with a bunch of weird traffic, like automated requests or spikes in API calls, it’s a sign that agents are stumbling around. They’re not browsing like humans, they’re grabbing data fast. Most sites aren’t built for that. They’re built to look good, not to work efficiently for machines.
You’ve likely seen a wide range of AI capabilities in product development. In your view, how “good” does AI really need to be before people start to trust it with important decisions?
People don’t need AI to be perfect. We need it to be better than the alternative. If an agent saves me 10 minutes vs. doing it on my own, I’ll tolerate a few mistakes that I can notice and correct myself. The bar is lower than you’d think, but the cost of failure has to be low, too. For high-stakes decisions (think anything medical, legal), AI needs pretty clear guardrails. For everything else? As long as it works most of the time, saves time, and gets smarter with each try. Trust in this kind of thing isn’t about 100% precision, it’s built on consistency.
Where do you personally draw the line between human input and machine autonomy when thinking about AI in customer experience?
I think you should draw the line at real, brand risk. If an AI’s decision could damage an actual customer relationship, like sending a tone-deaf email or misquoting pricing, keep a human in the loop. But for repetitive tasks (experimentation, content tagging), there’s way more room to let AI work solo. AI should handle the grind, humans should handle the judgment calls.
There’s growing interest in AI’s role in enhancing — not replacing — human creativity. How do you think teams can best leverage AI to unblock their creative process without losing the human touch?
AI’s best role in creativity is as a sparring partner. Stuck on some copywriting? Generate 20 options and cherry-pick the best. Hit a wall on campaign ideas? Use AI to brute-force angles you wouldn’t think of. The key is understanding what AI is good at, and it’s great at iteration, bad at final calls. The human touch isn’t in the whole output; it’s in the editing.
What’s your personal approach to staying ahead of AI trends and making sure your strategy is future-proof in such a fast-moving landscape?
At this stage, execution matters more than innovation. The winning solutions will be boringly reliable, not flashy. That’s how you know they’re working. The market is sorting this out quickly. Vendors adding AI randomly as an afterthought, just because it sounds or looks nice, will struggle. Those who are really building it into their core offerings have the advantage, not because their tech is more advanced, but because it’s more useful.
What advice would you give to business leaders who are hesitant about fully embracing AI — especially those unsure of how it fits into their customer strategy or brand identity?
Start with a simple question: What’s the most repetitive thing you do? Automate that. Prove ROI, then grow from there. AI isn’t an all-or-nothing bet, it’s a growing, rotating toolkit. And if you’re worried about brand fit, remember: AI doesn’t (and shouldn’t) define your strategy; it executes it.
Lastly, any final thoughts on where you believe we’re headed as AI continues to take on a more active and independent role in how people — and agents — experience the internet?
We’re heading toward a world where the internet has two users: people and their agents. The winners will be the platforms and brands that serve both equally well. That doesn’t mean homogenizing the web; it means designing for intent, not just eyeballs. The biggest shift won’t be technical, it’s psychological. Businesses need to stop thinking of AI as a feature and start seeing it as a customer segment.
For more expert articles and industry updates, follow Martech News
