How A Chief Product Officer Is Infusing AI Into The Product Roadmap

This blog first appeared as Steve Wunker's piece for Forbes

AI has rapidly become a key priority for product teams in all manner of software companies, but it’s often not easy to incorporate the technology. AI capabilities are changing fast, there are many distinct forms of AI to consider, relevant training data may not be easily accessible, and outputs aren’t always reliable.

Aarthi Ramamurthy is familiar with these challenges, and she’s routinely addressing them. Ramamurthy is Chief Product Officer at CommerceHub, a platform that processes an average 2.4 billion transactions a day to match up brands such as the online shops for Home Depot and Nordstrom’s with over 40,000 suppliers. The platform handles over $50 billion in commerce annually, which is about 30% more than eBay in the U.S.. Following stints at Meta, Clubhouse, Microsoft and two e-commerce start-ups, she’s seen a range of product departments and is bringing those lessons to CommerceHub.

Ramamurthy told me that she starts in an essential place. “I begin with empathy for our customers. I know what the retail ecosystem goes through and all the complexity in it.” Without the deep customer understanding, the plan lacks direction and risks being technology in search of a problem to solve.

“I also leverage this technology myself,” Ramamurthy adds. “I can have hundreds of documents to distill and summarize and AI can be very useful to look through a large corpus of information, highlight what is relevant and useful for our customers, and help me prioritize them. Then I can match available resources to the Jobs to be Done.”

She continues, “Customers don’t think about something being an AI feature or not. They know what the problem is and they’re looking for a solution. AI by itself is not a differentiator for us. We always stay focused on AI being a means to an end. The differentiation comes from doing something faster, better, or higher quality using the technology.”

How does Ramamurthy think about where to deploy deep learning in her product suite? “We started internally on insights dashboards. We want to think about data privacy externally, so we began on internal data such as what are customers asking for and how quickly suppliers are getting products. Then we’re figuring out patterns over time. Because we handle so much of the transaction, we can look across the whole loop in a way that no one else can.”

“AI is moving so fast,” she says. By doing quick, high ROI deployments, CommerceHub is enabling bigger use cases over time, and it’s staying flexible to adapt as AI’s capabilities keep shattering barriers.

In short, CommerceHub is leveraging AI by focusing on three principles:

1. Start with the customer first.
2. Examine use cases for customer value, risk, ease of execution, and potential learning value.
3. Stay flexible to learn fast, build capabilities, and exploit new AI breakthroughs.

​Rarely has a technology emerged that can enable so many advances in a product organization so quickly. As CommerceHub shows, the full potential is realized through a deliberate yet nimble approach. The company knows that the technology will continue to advance rapidly, and it’s organized its efforts to suit.

​By Steve Wunker
​