Viki Zabala, Chief Growth & Strategy Officer at First Insight – Interview Series

Viki Zabala, Chief Growth & Strategy Officer at First Insight, brings more than 22 years of leadership across high-growth SaaS, AI, and technology companies, operating at the intersection of strategy, product, and innovation. In her role, she leads First Insight’s unified growth engine across strategy, go-to-market, marketing, product, AI, customer success, and partnerships, shaping the vision behind the company’s Retail Decision Intelligence platform. Known for translating uncertainty into clarity, Zabala has driven hyper-growth, new market expansion, and AI-led transformation by connecting customer insight, decision intelligence, and scalable operating models to deliver measurable business outcomes for global enterprises.

First Insight is an AI-powered decision intelligence platform built for retailers and brands seeking to predict demand, optimize pricing and assortments, and reduce risk across the product lifecycle. By combining real-time consumer feedback with predictive, generative, and agentic AI, the platform helps organizations make faster, more confident decisions across design, merchandising, planning, and in-season execution. Used by leading global retailers and consumer brands, First Insight focuses on turning customer insight into actionable intelligence that improves margins, accelerates speed to market, and strengthens long-term growth.

Your career has consistently sat at the intersection of data, go-to-market strategy, and execution. What moments earlier in your career most shaped how you think about turning insight into real operational decisions today?

I’ve always been focused on one fundamental challenge: how you influence and change behavior at scale.

Early in my career, that showed up in mobile apps and ad tech, where feedback loops are immediate. You learn quickly that data only matters if it changes what someone does next—installs, engagement, conversion. Later, in IoT and experiential platforms, the same truth played out in physical environments: how context, timing, and experience shape human behavior in real time.

Across all of those industries, one lesson stayed constant: insight is only valuable if it’s actionable in the moment a decision is made. If it doesn’t survive the pressure of execution—pricing, go-to-market, inventory, messaging—it’s just information.

That mindset is what brought me to First Insight. Retail is one of the most behavior-driven industries there is, yet decisions have historically relied on lagging indicators and gut instinct. My work has been about closing that gap—bringing the voice of the customer forward early enough, and continuously enough, to engineer better outcomes rather than react to failure.

My focus now is helping organizations make better decisions early enough to matter—so they grow revenue, earn customer loyalty, and consistently outperform the market.

As Chief Growth & Strategy Officer at First Insight, you oversee product, AI roadmap, GTM, and customer success. How does having that unified view change how AI should be designed and deployed inside retail organizations?

When you see the entire system, you stop thinking about AI as a tool and start thinking about it as an operating model.

Product shows you what’s technically possible. Go-to-market shows you what will actually be understood and adopted. Customer success shows you what holds up under real-world constraints—time pressure, cross-functional tension, data quality, and accountability. When those perspectives are unified, AI gets designed around how decisions really happen, not how impressive the technology looks in isolation.

That’s why AI in retail has to function as a system of decisions and actions, not just a system of intelligence. It has to connect customer signals to pricing, assortment, marketing, and planning in a way that aligns teams and accelerates decisions. When AI reduces friction between teams and shortens the distance between insight and action, it starts delivering real value.

Retailers have long relied on seasonal planning cycles and historical data. From what you’re seeing on the ground, why are those models increasingly misaligned with how consumers behave today?

Because those models were built for a world where retail was primarily about optimizing what already existed—not inventing what’s next.

Historical sales and seasonal cycles can help explain performance in established categories, but they’re weak at the two things retailers need most today: responding to fast-changing customer behavior and creating new demand through product innovation and whitespace expansion.

Demand now shifts in real time—driven by price sensitivity, cultural moments, social influence, economic pressure, and channel dynamics. A trend can emerge overnight. A pricing signal can change behavior instantly. Historical data explains what already happened, but it doesn’t reliably tell you how customers will respond next—even for products already on the shelf—when context and sentiment can change at any moment.

At the same time, many retailers are making decisions with aging CRMs and outdated views of who their customer actually is. New competitors, new channels, and younger generations with different expectations and spending power are steadily pulling customers away—often without retailers realizing it until results show up in missed forecasts or declining loyalty. In many cases, brands are optimizing for customers they no longer have—or customers who have already moved on.

And when it comes to innovation, sales history can’t validate a product that doesn’t exist yet—or a customer segment you’re in danger of losing. That’s why so many retailers end up iterating on the past instead of confidently funding the next category, the next feature set, or the next audience. The unlock is bringing the voice of the customer forward—early enough to guide concept creation, pricing power, and positioning—so innovation becomes a repeatable system rather than a gamble.

First Insight’s AI assistant, Ellis, enables natural-language queries around pricing, assortments, and demand. How important is interface design and accessibility in driving real AI adoption versus just technical capability?

Interface is the difference between “AI exists” and “AI gets used.”

Retail decision-making spans far more than one moment—concept research, design, assortment building, price optimization, margin modeling, buy depth, allocation, in-season adjustments, marketing and selling. The challenge isn’t that retailers don’t have questions; it’s that answers are trapped in dashboards, decks, exports, and specialized teams—and by the time they’re delivered, the moment has passed.

Ellis matters because it removes friction between insight and action. Instead of navigating reports or waiting for new analysis, teams can ask strategic and tactical questions in plain language—about concepts, pricing, assortments, segments, markets, competitors—and get clear, predictive answers in minutes. That’s not just usability; that’s decision velocity.

Accessibility also drives adoption across the organization. When the same customer signal is instantly available to merchandising, pricing, marketing, and planning, you reduce internal battles and misalignment. People stop debating whose data is right and start debating what to do next—faster, and with more confidence.

You’ve worked closely with retailers navigating margin pressure, inventory risk, and volatile demand. Where does AI deliver the fastest, most measurable impact today—and where is the hype still ahead of reality?

The fastest impact shows up where decisions are frequent, expensive, and time-sensitive: pricing, assortment selection, demand validation, and inventory risk. When AI helps teams avoid overbuying, hold price with confidence, or exit losing products earlier, the financial impact is immediate and measurable.

Where hype gets ahead of reality is in the idea of fully autonomous retail—or AI replacing real customer understanding with synthetic shortcuts. Consumers are very clear: they value authenticity, transparency, and being heard. AI that distances brands from the customer doesn’t create efficiency—it creates risk.

The winning model today is human judgment augmented by predictive insight, not automation for automation’s sake.

Many AI tools promise predictive capabilities. What does meaningful prediction look like in retail, and how should leaders evaluate whether predictions are actually decision-ready?

Meaningful prediction in retail isn’t a forecast—it’s the ability to close the loop from customer truth to financial outcome.

A lot of AI outputs sound predictive, but they don’t change the business because they never make it into the operating cadence. The quarter misses, the inventory piles up, markdown budgets get spent—and everyone can point to data somewhere that could have helped. The real failure is that decisions weren’t aligned, actions weren’t taken, and the workflow didn’t change.

Decision-ready prediction does three things at once:

1. It’s grounded in how customers actually perceive value—not just sales history—so it can guide decisions from concept through in-season.
2. It ties directly to the economics: demand elasticity, willingness-to-pay, AUR/ASP over the product lifecycle, and the margin implications of holding vs. discounting.
3. It’s operational—embedded in a repeatable process that teams actually follow, not trapped across dozens of tools and siloed dashboards.

A recurring theme we see is the cost of the “long tail” of SKUs. Over-assortment is a silent killer: excess depth, low velocity, buried risk. One of the biggest levers that predictive AI unlocks is the ability to cut the tail — remove underperforming products early and reinvest those inventory dollars into top performers where customer demand and sentiment are highest.

When teams apply this discipline, we see dramatic results:

• inventory dollars are freed up for innovation and high-score opportunities,
• markdown cadence stabilizes and shrinks,
• promotional pressure eases, and
• brand trust increases because customers aren’t trained to expect 50–60% off before they’ll buy.

Leaders should evaluate predictive AI with one question: Does it change where we invest? The highest ROI is not more data — it’s better decisions in how you allocate capital, time, and inventory against real customer demand — early enough to matter.

Responsible AI is often discussed at a high level. In retail specifically, what does practical, responsible AI adoption look like when decisions directly affect pricing, consumers, and brand trust?

Responsible AI in retail starts with one simple principle: use AI to deepen the customer relationship, not exploit it.

This isn’t about hyper-tracking individuals, surveillance, or harvesting data for its own sake. Responsible AI is about bringing the voice of the customer into every decision at scale — so products, pricing, messages, and experiences reflect what people actually value. In many ways it’s a form of co-development: customers guide what gets created, how it’s positioned, and what feels fair.

Practically, responsible AI looks like:

• Grounding decisions in real customer input — both quantitative and qualitative (“what she/he/they said”).
• Building transparency and guardrails for high-impact decisions like pricing, promotions, and segmentation.
• Ensuring fairness across segments and markets, so AI doesn’t inadvertently favor one group while disadvantaging another.
• Keeping humans in the loop for judgment, accountability, and the creative nuance AI can’t generate itself.

Used this way, AI strengthens the customer relationship instead of eroding it. Customers feel heard at scale. Teams make better decisions faster. And brands build trust — because they’re no longer reacting to the market; they’re acting with it.

You’ve led both marketing narratives and product strategy. How should retailers rethink internal storytelling around AI so it’s seen as a decision partner rather than a threat or black box?

Retailers should stop telling the story that AI is “smarter analysis” and start telling the story that AI is customer centricity at scale.

The internal friction in retail isn’t just silos—it’s silos making high-stakes decisions with different truths: marketing has engagement signals, merchandising has sales history, pricing has margin pressure, planning has inventory constraints. That’s where the battles happen.

AI becomes a decision partner when it creates a shared language across functions: the voice of the customer, translated into predictive guidance that informs product, price, assortment, and how to sell—end to end, from concept to conversion.

And it’s important to be honest about the role of humans. AI doesn’t invent the next breakthrough idea—it learns patterns. Humans bring creativity, taste, brand intent, and cultural intuition. AI makes that creativity sharper by shortening feedback loops and pressure-testing decisions before the market does.

As AI becomes more embedded in planning and in-season decision-making, how do you see the role of human judgment evolving rather than disappearing?

Human judgment becomes more important—and more leveraged—because in-season is where retail profit is won or lost.

Markdowns are one of the biggest costs in retail. Retailers often budget for them because they’re forced to clear unsold inventory. The reason markdowns are so painful is timing: discount too early and you destroy margin; discount too late and you miss the window to convert demand.

With predictive AI and humans in the loop, teams can model elastic demand curves and understand how ASP/AUR should evolve over the lifetime of the product—based on sell-through, customer perception, and market signals. That enables smarter moves: when to hold price, when to discount, and by how much—without overcorrecting.

And in-season decisions aren’t just pricing. AI can inform promotions and marketing in-season by factoring in cultural moments, influencers, trend acceleration, and shifts in customer personas—alongside product perception and price sensitivity. Humans then apply judgment: brand intent, risk tolerance, and the creative choices that AI can’t originate.

The future isn’t automation. It’s faster, more customer-informed decisions—where AI scales the listening, and humans lead the meaning.

Looking ahead, how do you expect agentic and generative AI to reshape retail workflows over the next two to three years—not theoretically, but operationally?

We’re moving from systems of intelligence to systems of action.

Operationally, generative AI will make insight accessible across roles and levels—summarizing, comparing, explaining, and answering questions instantly. Agentic AI will increasingly take on the repetitive work that slows organizations down: preparing scenarios, assembling executive-ready briefs, monitoring signals, flagging risk, and coordinating next-best actions.

But the most meaningful change won’t be that AI “runs retail.” It will be that retailers finally tighten the loop between the customer and the enterprise. Teams will move faster, cut through internal friction, and make better decisions earlier—before trends peak, before markdowns cascade, and before missed opportunities become quarterly misses.

The retailers who win won’t be the ones with the most AI experiments. They’ll be the ones who build a repeatable operating cadence where customer truth, predictive intelligence, and human creativity work together—from concept to conversion.

Thank you for the detailed interview, readers who wish to learn more should visit First Insight.