How experts are using AI to turn call transcripts and historial retail data into real-time growth levers

For decades, business intelligence has been a one-way street. You ask a question, plead with engineering to build a tracker, wait months for data to accumulate, and only then can you find an answer. But a new approach powered by AI is making that linear model obsolete, turning stagnant data archives into living assets you can query with questions you haven't even thought of yet.

In this future-is-now scenario, John Hughes, Chief Growth Officer at commerce platform giant Vintage Cash Cow, explains how his UK-based resale platform is pioneering a non-linear approach to data that empowers strategic agility.

Escaping the roadmap: The breakthrough, Hughes explains, is that AI can now retroactively analyze historical data with entirely new criteria, liberating businesses from the traditional product development cycle. “It can really free up your product roadmap,” he says. “We can now use our existing data to answer questions we didn't even know we had. Previously, to capture new information, we'd have to add a new reason into our software and wait for enough events to happen before drawing any conclusions.”

Rewriting history: Hughes offers a potent example of this in action. "Now, you can just give the AI an instruction: 'Here's my dataset of 50,000 call transcripts. I used to have four ways of classifying a call, but now I want to add a fifth. Please go reclassify all of them for me,'" he says. The result is transformative. "Suddenly, you discover this 'new' process has actually appeared in 25% of your calls from the past year," Hughes continues. "You can get that insight instantly, without needing to change the product or retrain your agents."

From analysis to action: This same underlying technology—AI that can listen and comprehend—can be applied to automate real-time agent workflows. Hughes describes building a system of sequential, specialized agents. “We have a call listening agent that is trained to pick up whether the customer mentioned where they heard about us,” he details. “At the end of the call, it hands that information over to another agent that is trained to make sense of that data and put it into our CRM.”

Cognitive burden: The human-centric result of this automation is profound. By offloading the administrative work, agents' focus can be freed up entirely. “Any agent recognizes how distracting it is when you're talking to a customer but also having to click between screens in a CRM or make a note,” Hughes says. “By removing that cognitive burden, you allow them to focus fully and be more present in the call.”

Providing value: Ultimately, whether analyzing past data or automating present tasks, the goal is the same: clarifying what work is uniquely human. “A lot of people think AI is coming to take their jobs, but it’s going to make you better at your job,” he says. “You just don't realize which parts are your job and which parts aren't. Your real job is not the burdensome admin tasks; your job is providing value.” He concludes, “That call in the call center is not repetitive. It's an interaction between two different humans every single time.”