AI product strategy in 2025: Top advice from AI-first founders

In late 2024, we surveyed over 200 leaders in AI across the tech industry, over half of whom were founders, to better understand how today’s new age of AI innovation impacts their day-to-day decision-making and overall product strategy as they headed into 2025.

The results were insightful:

• 97% of respondents think AI will become a long-term necessity that will be as “essential as the internet”
• 89% were somewhat to very concerned that their competition’s AI strategy will outpace their own
• 60% reported cost savings for business and customers after implementing AI

We also sat down for a series of candid conversations with top AI-first founders to learn what they are doing to successfully ride this AI tide and come out ahead—and what they’ve learned since their journey began. 

In particular, the founders examined what a top product strategy looks like, including how to develop an AI-first product strategy, how to adjust and iterate product roadmaps in line with rapid industry change, and why a continued focus on customers paves the way for success. 

Check out the entire Assembly Required series here

On developing your AI product strategy based on customer needs or technology capabilities

AI product strategy encompasses many components, like choosing AI partners, determining pricing, and prioritizing team focus areas, but zeroing in on end-user value remains a top priority for these AI-first founders. Each founder, though, has a slightly different take on how they’re balancing customer value with the innovative technology available to build with as they set their overall product strategy.

Jason Boehmig, founder and CEO of Ironclad, discusses how a winning product strategy in AI solves for end-to-end workflows for their customers. He describes how integrating AI helped Ironclad unlock a key contract management workflow that was a core issue for their users and ultimately, continue to disrupt an increasingly competitive software category.

Victor Riparbelli, co-founder and CEO of Synthesia, advises that you need to truly understand your product’s unique value driver to develop a successful product strategy. For Synthesia, this meant focusing first on the user experience and customer problems they wanted to solve for, and let their technology decisions follow from there, rather than fixating on the technology they wanted to implement or build with. 

Co-founder and CEO of Fireflies.ai Krish Ramineni approached building product strategy similarly to Synthesia. When sitting down with our team, he discussed how he learned that the problem you’re trying to solve has to come first—and the technology second—to successfully solve problems for your users, even if you’re setting a strategy based on where the industry is going, rather than what the technology can do today.  

Replicate founder and CEO Ben Firshman has witnessed firsthand how thousands of developers are building on the Replicate API. In his experience, today’s developers are often building AI-first products with a desire to incorporate a specific technology or, like the founders above, they are starting with a particular user need and letting their technology decisions flow from there.

On making bets, adjusting, and iterating as the market changes rapidly

Each of the founders we spoke with agreed—there’s never been a faster pace of change than what we’re currently seeing in the AI space. In order to keep up and succeed, they agreed that today’s leaders must be prepared to make a bet that they feel strongly about, and then be willing to adjust, iterate, and repeat as they get signals from the market. 

Edo Liberty, founder and CEO of Pinecone, experienced this firsthand when he first went to market with the concept (and name) of vector databases. He shares how it was a customer insight that helped him unlock how to describe what his product does and how that helped him see traction with the product.

Victor Riparbelli, Synthesia co-founder and CEO, also describes how even though your initial product vision may seem spot on, the market may change—sometimes drastically—while you’re working on getting your product to market. Or, when you try to put your vision into production, it might flop. Flexibility and a willingness to adjust are key to long-term success in today’s market. 

Ironclad founder and CEO Jason Boehmig also discusses how the AI strategy that initially sounds compelling isn’t always the right strategy to pursue. Since AI innovation is moving at such a rapid pace, the market is often playing catch-up to the reality about how to best utilize AI, explains Boehmig. Being okay with admitting that your first instinct was wrong, and being open to changing your plans as the industry progresses, has been key to Ironclad’s success. 

Riparbelli goes on to explain how the best product strategies are driven by a deep sense of curiosity and open-mindedness. Had his team not been open-minded, he explains, they would have ended up with an entirely different product—one without true market fit—and been nowhere near where they are today.

On keeping the focus on customers above all else 

Above all else, these founders agree that your roadmap has to be focused on your customer’s needs, rather than fluctuating wildly to incorporate new advancements in the industry. 

Fireflies.ai co-founder and CEO Krish Ramineni explains that if you start by figuring out how to make your customers happy, you’ll be able to figure out how to monetize that product. There’s a natural equilibrium, he continues, between quality, cost, and customer satisfaction and when you find it, you’ll know your product strategy is successful.

Boehmig also describes how the biggest breakthroughs for Ironclad have come through focusing on how to generate significant value for their customers, not just on making a flashy or cool AI product. 

Riparbelli drives home the point that customer feedback should greatly influence your product roadmap. For Synthesia, this meant anchoring their roadmap and product strategy directly to customer feedback trends they saw through support channels or direct conversations.

Source: Read More