How Generative AI Is Turning Promise into Performance

For the past two years, there’s been no shortage of hype around generative AI. But as business leaders have moved beyond pilots and proof-of-concepts, the big question has shifted from “what can it do?” to “what’s the return?”

A recent Financial Times article titled “From AI to ROI” points to some answers, backed by data. Drawing on a new study by researchers at Zhejiang and Columbia Universities, it provides some of the clearest evidence yet that generative AI is creating real productivity and revenue gains for businesses, especially in ecommerce.

The research examined several large-scale GenAI experiments and found results that are difficult to ignore:

• Adding an AI assistant before the point of purchase boosted sales by 16.3% and increased conversion rates by 21.7%.
• A hybrid AI-human system, which passed complex issues to human agents, still outperformed an experienced all-human sales team by 11.5%.

Those are significant gains, and rest assured, there was no replacement of human talent, but AI was used to enhance it.

As the study notes, GenAI’s impact “arose from augmenting existing technologies with minimal or no displacement of labour or capital,” meaning these weren’t cost-cutting improvements, but genuine productivity improvements.

The researchers went on to highlight the mechanisms behind the gains:

“Conversion and engagement metrics increased by 1–22% across workflows, suggesting that GenAI applications contributed to productivity primarily through the creation of consumer value, improving user experience and expanding the market rather than increasing spending per buyer.”

In simpler terms: AI is working best not by replacing people, but by removing friction, making every interaction a little smoother, every search a little smarter, every recommendation a little more relevant.

These results are powerful for anyone looking at how AI can create value across customer touchpoints. What happened in online retail, where friction is costly and attention is fleeting, is now happening across industries.

We’re seeing it in sales conversations, customer success calls, and internal collaboration. AI is becoming a quiet but powerful co-pilot, transforming how information flows and how decisions are made.

Vinton started as a way to bring conversations natively into Salesforce, automatically capturing, summarising, and structuring the insights that emerge in every customer or internal meeting. But what’s most exciting now is seeing how customers are using Vinton to generate their own “AI to ROI” stories.

One customer in Financial Services shared:

“With Vinton, we’ve transformed how we capture, manage, and learn from client conversations – all within Salesforce. The result is greater efficiency, stronger compliance, better data, and a more responsive and insightful client experience.”

A Manufacturing customer said:

“Before Vinton, we were losing valuable insights on every customer call. Our teams were busy, and the quality of notetaking varied – so the data in Salesforce just didn’t reflect what was really happening. Vinton changed that. It captures everything from the conversation – transcripts, summaries, insights – and logs it natively in Salesforce, where we all work. That native integration was critical for us… With rich, searchable conversational data in Salesforce, we have built a solid foundation for AI-powered service. Managers are more informed. Technicians are more efficient. And customer experiences are more consistent. Customer satisfaction is at an all time high.”

Another customer in the Recruitment industry highlighted the same impact on productivity, saving at least seven hours a week per recruiter. But it’s not just about time saved. Using Vinton provides deeper insight, invaluable for coaching and improving client interactions.

And from a High Tech sales leader:

“Since introducing Vinton, we’ve reduced meeting follow-up and admin time by over 75%, with automatic recording, transcription, and tailored summaries and insights for all conversations… We’ve seen improvements, not only in efficiency, but also in sales execution, with a 20% improvement in forecast accuracy and a 25% increase in deal velocity – from initial meeting to deal closure.”

What’s common across all of these examples is the same theme seen in the Zhejiang and Columbia research: AI is reducing frictions. Whether it’s between buyer and seller, recruiter and candidate, or manager and team, those small efficiencies compound into measurable ROI.

The university researchers called this the “creation of consumer value,” the idea that GenAI works by making experiences smoother and more intuitive. That same principle applies inside organisations too.

Vinton reduces the friction of remembering, recording, and re-entering information. It allows every team member to focus more on the conversation and less on the capture.

In doing so, it transforms meeting data into a structured, usable source of truth for tools like Salesforce’s Agentforce. That’s not just productivity; it’s data enrichment. It gives AI the context it needs to act intelligently across the business.

If 2023 and 2024 were about exploring what AI could do, 2025 is shaping up to be the year of proving what it does do.

We’re now seeing clear evidence, from both academic research and real-world use cases, that GenAI delivers measurable, sustainable impact when applied thoughtfully.

The key isn’t automation for automation’s sake. It’s identifying the small moments of friction in your workflows and applying AI where it can add the most value, whether that’s in customer engagement, data capture, or decision support.

At Vinton, we’re proud to be part of that story, helping teams make AI practical and genuinely transformative.