Driven by Data: The Podcast

Welcome to another episode of the Data Debrief, the companion show to Driven by Data: The Podcast, where hosts Catherine Dowden-King and Kyle Winterbottom unpack Tuesday's episode, share what's been on their minds, and explore the realities of leadership, culture, and capability across the data and AI landscape.This week, Catherine and Kyle reflect on the conversation with Justin Borgman, co-founder, CEO, and chairman of Starburst, diving deeper into why AI adoption keeps stalling at scale, the real cost of pointing powerful tools at the wrong problems, and what it means to build differentiated business capability rather than just better infrastructure.They cover:Why Justin's refreshingly candid starting premise — that messy, fragmented data is simply the reality most organisations are working in — cuts against the vendor instinct to promise a clean, unified solution, and why that honesty lands differently coming from a SaaS founderThe recurring pattern of businesses spending two to three years consolidating data into a single source of truth, only to arrive at the same "so what?" question — and why Starburst's founding premise of using data where it lives challenges the orthodoxy of centralisation as a prerequisite for valueThe context problem that no platform solves on its own: how the same number pulled from the same source can mean two completely different things depending on interpretation, and why that ambiguity at enterprise scale can quietly corrode trust in data across an entire organisationJustin's observation on where AI is delivering the clearest, most demonstrable value right now — coding — and what that signals for how skill sets in software development, data science, and adjacent technical roles are likely to evolve faster than most organisations are prepared forThe entry-level talent question neither businesses nor education systems have yet answered: as AI absorbs the work that once built foundational experience, where does the next generation of senior leaders come from, and who quality-assures the outputs of people who have never done the work themselvesCatherine's take on AI as fire: extraordinarily useful when understood and controlled, capable of running out of control very quickly when deployed at enterprise scale through FOMO rather than focus — and why a CFO's instinct to shut it all down is an entirely predictable response to cost spiralsKyle's reflection on the speed problem at the heart of this AI cycle: unlike previous technological revolutions, where the pace of change gave industries time to adapt and reskill, this one is moving fast enough that many organisations and individuals haven't yet worked out what adaptation even looks likeA moment from Catherine's farming background and the latest series of Clarkson's Farm that brings the AI transition into sharp relief — precision agricultural technology that looks futuristic to most farms but is closer than people think, and what the emotional weight of replacing a working horse with a tractor tells us about how humans really respond to transformationKyle's thought of the week: prompted by a pattern he's been tracking across executive search processes throughout 2026, Kyle reflects on a frustrating gap between capability and communication at the senior leadership level. The people not getting the roles aren't failing on technical grounds — they're losing out on energy and enthusiasm, inability to be concise, talking around questions rather than answering them, and failure to give specific examples. Kyle's concern is that these aren't just interview problems: they're signals of how someone will perform in front of a board or a CEO, and the skills that fix them can be self-taught and improved quickly. Catherine adds a practical tip for building confidence in high-pressure communication situations using AI tools like ChatGPT or Claude as a low-stakes rehearsal partner — and shares a striking example from a full studio broadcast that shows how dramatically even experienced communicators can disappear under pressure.This episode explores why the data and AI industry's biggest bottleneck isn't the models — it's the foundations, the focus, and the people trusted to lead the work and make the case for it.

Welcome to another episode of the Data Debrief, the companion show to Driven by Data: The Podcast, where hosts Catherine Dowden-King and Kyle Winterbottom unpack Tuesday's episode, share what's been on their minds, and explore the realities of leadership, culture, and capability across the data and AI landscape.This week, Catherine and Kyle reflect on the conversation with David Krauza, VP of Enterprise Data Strategy, Products & Governance at Comcast, diving deeper into the "arts and crafts" trap that derails data programmes, the discipline of building stakeholder trust before you need it, and what it really takes to drive the bus rather than ride it.They cover:Why David's mandated business school module ended up shaping his outlook on data leadership, and the recurring pattern of guests whose commercial thinking was forged outside a purely technical backgroundThe "arts and crafts project" analogy David's boss used to describe technically impressive work that never moves the needle, and why naming the difference between process-enjoyment and outcome-focus matters as much in data as it does in any creative pursuitWhy so much of the data and AI ecosystem gravitates toward exciting new models, tools, and techniques without tying the work back to specific goals, decisions, and KPIsDavid's framing that you have to help people before you need their help, and why the leaders who consistently land the biggest roles are the ones already putting into their networks and communities long before they need anything backCatherine's take on why this same principle defines external brand building, and why leaders who wait until they're job hunting to invest in relationships are always playing catch-up against those who started years earlierKyle's view that building relationships with future stakeholders is not a side project for a data leader, it is the job, every bit as much as overseeing platform delivery, governance, and architectureWhy David reframed the trust gap many data leaders face as a sequencing problem rather than a communication problem, and what that distinction means for how and when leaders should be reaching outThe "bus riders and bus drivers" analogy at the heart of the episode title, and why organisations hire a data leader precisely because they don't already know the answer, making it the leader's job to shape direction rather than simply execute instructionsWhy being a strong, detailed communicator changes the entire dynamic of a hiring conversation, and how that same skill plays out with stakeholders once someone is in the roleCatherine's practical tip for building interview and communication confidence using AI tools like ChatGPT or Claude as a low-stakes practice partner, and why consistent repetition beats waiting for natural talent to show upKyle's thought of the week: prompted by a message from a CDO at a crossroads in their career, Kyle reflects on why the CDO role isn't disappearing or resurging industry-wide so much as it's becoming entirely dependent on whether a business's leadership views data as a commercial value-creation function or a technology delivery capability. Where it's the latter, that responsibility increasingly sits with the CIO, and Kyle notes the early signs of broader transformation-style mandates emerging that fold CDO, CIO, and Chief AI Officer responsibilities into a single board-level role.This episode explores what it actually takes to drive value rather than just deliver outputs, the discipline of investing in relationships long before you need them, and why naming the gap between busywork and real impact is often the first step to closing it.