TechTable by Vention: Productivity, copilots, and the future of metrics
Back in autumn, when our TechTable guests led discussions on reskilling strategies in the age of AI, one question stayed with us long after dinner ended. Shouldn’t we also rethink how we measure productivity?
After dedicating our latest winter TechTable to this exact topic, the answer was a clear yes.
Moderated by Glyn Roberts, UK CTO at Vention, the evening focused on a question at the centre of today’s innovation agenda. How is productivity being redefined as copilots become part of everyday work? And what should metrics look like when AI plays an active role in how teams deliver results?
Our guest speaker was Jesus Fernandez, Head of Engineering at Lloyds Banking Group. With more than 18 years of experience scaling engineering teams, Jesus shared practical insight into how productivity can be assessed in AI-supported environments.
He opened the conversation by talking about his work leading the Productivity Lab at Lloyds Banking Group. Teams across the UK and Asia are exploring what truly drives developer productivity and how to measure signals that reflect real impact rather than just output.
Redefining productivity in the age of AI copilots
As generative AI tools become embedded across the development lifecycle, traditional productivity metrics are being questioned. Measures such as lines of code, velocity, or ticket volume do not tell the full story.
Jesus and the attendees discussed how AI-assisted workflows are changing the nature of engineering work, including:
- How to measure the ROI of AI tools beyond raw output, looking instead at learning speed, defect reduction, and alignment with business outcomes
- Why the lower cost of building MVPs with AI makes it smarter to build quickly, spot issues early, and refine based on real-world feedback
- Ethical considerations and bias in AI-supported metrics, particularly in cybersecurity teams
Participants shared examples from their own organisations, which highlighted a growing move toward hybrid metrics. These aim to balance speed with long-term value, while avoiding the risks of over-automation.
Need support with your AI adoption
strategy?
Our AI-enabled teams provide +15% faster delivery speed to your project.
Measuring what matters: From numbers to insight
A key part of the discussion focused on whether existing performance metrics are still fit for purpose.
As AI copilots take on routine tasks, leaders are paying much closer attention to measures that capture:
- Strategic impact rather than individual throughput
- The quality of cross-functional collaboration
- Learning and adaptability as teams experiment with new AI tools
Guests talked about approaches already in use at their companies, which ranged from custom dashboards and refined OKRs to entirely new ways of capturing qualitative insight.
Accurate measurement also brings an unexpected benefit. It boosts morale. When teams can clearly see the improvements AI enables, trust in both the tools and their own work increases.
Challenges and opportunities in metrics innovation
Despite the enthusiasm, attendees all agreed that rethinking productivity comes with certain challenges.
One of the biggest is integrating AI-driven metrics into existing performance systems without overwhelming teams. For metrics to have a positive effect, they need to be introduced gradually and with transparency. The larger the organisation, the more complex this becomes.
Another challenge lies in mindset. As traditional benchmarks lose relevance, leaders and teams need a shared understanding of what good performance looks like.
Finally, the group stressed the importance of keeping humans in the loop. Just as AI augments human work, AI-generated insights should support human judgement, not replace it.
Looking ahead: What leaders should watch
TechTable closed with a look to the future.
Attendees expect continued shifts in performance KPIs as AI tools mature, increased investment in systems that make productivity signals easier to interpret, and a growing role for narrative data. Stories and context will matter more in explaining what the numbers really mean.
Source: PR@ventionteams.com