Adam Mills

By Adam Mills

AI Editor

0

Share

Newsweek is a Trust Project member

See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.


Enterprise software companies used to organize product development around the release cycle. Teams built the product, shipped it and improved it over time based on customer feedback. But as AI becomes more embedded in software, what happens after deployment is becoming harder to separate from the product itself.

AI-enabled products depend more on context and often perform differently across customers, workflows and business settings. That makes real-world use more important to how companies decide what to improve, adjust or build next.

Read More on AI

“Deployment used to be the finish line. Now it’s the starting gun,” Ken Fine, CEO of Affinity, a software company that helps private capital firms manage business relationships, told Newsweek. “When AI is in your product, the product is being co-authored in production, and post-deployment optimization is its own parallel cycle that never stops running.”

That changes the rhythm of product development. Product teams once had more distance from implementation and customer success. Road maps could be shaped by sales feedback, support tickets, surveys and periodic customer reviews. But AI-enabled products may depend more on timely information from the places where they are actually being used.

Usage patterns, workflow breakdowns, support interactions and implementation feedback are no longer just post-sale service issues. They can become inputs into what the software needs to do next.

Meredith Whalen, chief product, research and delivery officer at IDC, a technology market intelligence firm, told Newsweek that AI agents can learn from real-world use, making monitoring data, governance and human oversight “part of the product lifecycle, not just operations.”

Tiger Tyagarajan (L), former CEO of Genpact, spoke with Dr. Ranjit Tinaikar (R) during Newsweek’s “AI Impact Forum” webinar on how AI is blurring software and services models.

...

A similar point came up during Newsweek’s recent “AI Impact Forum” webinar, “Will Today’s Software Companies Be Tomorrow’s Services Companies?” The discussion between Dr. Ranjit Tinaikar, host of the series and former CEO of Ness Digital Engineering, and Tiger Tyagarajan, former CEO of Genpact, focused broadly on how AI is changing the boundary between software and services.

But the discussion also pointed to a more practical issue: speed. Tyagarajan said AI changes much faster than traditional software because each interaction can inform the next one. “Every transaction is a learning [opportunity],” he said.

That makes closeness to customer use more valuable. If a company is too far from how its software is actually being deployed, it may struggle to build products that keep improving in the right direction.

The conversation also turned to how leaders gather customer input. Tinaikar described traditional mechanisms such as surveys, focus groups and periodic reviews as too slow for the pace of AI-enabled change. Tyagarajan pointed to models where companies collect feedback from customer interactions continuously, then use AI systems to identify patterns across those signals.

The challenge is deciding what that information should change. More data does not automatically create better software. Companies still need to determine which signals matter, how quickly to act on them and how to separate noise from evidence.

Whalen told Newsweek that different signals serve different purposes. “Workflow breakdowns and process data help to identify where AI should execute or assist,” she said. Usage patterns can help companies prioritize fixes, while support interactions can reveal “intent, sentiment, and failure modes” that inform model updates and policies.

In that model, deployment is not the end of the product story. More than ever, AI software’s success depends on how it fits into data, workflows and human processes inside the enterprise.

That also changes how software companies organize around product development. If insights from implementation and customer success become central to development, those teams may benefit from a clearer path into product decisions. That can mean different reporting lines, different incentives and faster coordination between the people building software and the people closest to the customers using it.

Fine said companies that try to keep those functions apart will feel the cost.

“Our CSMs and implementation consultants are often the first to see how our AI behaves in a real customer environment, and we treat what they observe as product signals, not service problems,” Fine said. “The winners will be those that route post-deployment learning back into product strategy as a core operating rhythm.”

Customers increasingly expect AI software vendors to understand not just whether a product works, but how it performs inside real workflows.

...

Brian Stimpfl, CEO of S-Docs, a Salesforce-native document generation and e-signature company, told Newsweek that one of the most dangerous enterprise software failures is also one of the quietest: “non-adoption.”

A company may buy software, allocate budget and later discover that workflows were never adopted or that expected outcomes never materialized, he said. The warning signs are not always obvious.

“What are customers actually doing, versus what they told us they would do?” Stimpfl said. “When those two things don’t match, that influences your product roadmap right there.”

Software companies may gain an advantage less by releasing the most features than by learning fastest from the customers already using the product. In that environment, customer success becomes more than retention, implementation becomes more than setup and feedback becomes more than a quarterly review.

Whalen said companies that cannot turn customer usage, workflow data or implementation feedback into product changes could struggle with “slower time to value, higher support costs, lower retention, and loss of differentiation” as buyers gravitate toward platforms that can turn customer data into measurable results.

Fine said the risk comes down to speed.

“The companies that fall behind are those that cannot redefine ‘quickly’ to mean machine speed while still applying human judgment,” he said.

For more on how AI is changing the relationship between software, services and customer feedback, watch Newsweek’s recent “AI Impact Forum” webinar discussion with Tiger Tyagarajan above.

You can also sign up for the next “AI Impact Forum” webinar, "The Trillion-Dollar Question: Who Wins and Who Loses in the Services Economy?" That session takes place on May 28 and will feature Noshir Kaka, senior partner at McKinsey & Company.

View 0 Comments

Request Reprint & Licensing

Submit Correction

View Editorial & AI Guidelines