Latest / MBA Training Data / Why most data culture initiatives fail before they start
Transcript
- 0:00Welcome to the Leaders Insights podcast.
- 0:07Today's episode, why most data culture initiatives fail before they start.
- 0:12Most CDOs I speak to have launched a data culture program at some point.
- 0:16Almost none of them can tell me it worked.
- 0:18Why is failure the default outcome here?
- 0:21Because the initiative usually starts with the wrong diagnosis.
- 0:25Organizations see poor data adoption and conclude they need better tools, cleaner dashboards,
- 0:30maybe a data literacy training program.
- 0:32So they spend 18 months rolling out a new platform, run a few lunch and learns,
- 0:37and then wonder why analysts are still being ignored in business meetings.
- 0:41The problem was never the tools.
- 0:43So what is the actual problem?
- 0:45Incentives.
- 0:46People behave according to what gets them promoted, praised, or protected.
- 0:51If a sales director at a company like Schneider Electric builds their entire quarter around gut
- 0:56instinct and it works, nobody is rewarding them for also consulting the demand forecast model.
- 1:02The model becomes decoration.
- 1:05Until you change what behavior is rewarded, the culture doesn't move.
- 1:09That's a structural argument, not a culture argument.
- 1:12Are you saying CDOs are solving the wrong problem entirely?
- 1:16Partially, culture is downstream of structure.
- 1:20The way Spotify approached this around 2023 and 2024 was instructive.
- 1:25They embedded data fluency expectations directly into performance reviews for product managers,
- 1:31not as a checkbox, but as a weighted criterion.
- 1:35By 2026, they report substantially higher rates of model-informed product decisions.
- 1:41That's not a culture program.
- 1:43That's an operating model change that produced a culture shift as a byproduct.
- 1:47But most companies aren't Spotify.
- 1:50A traditional manufacturer or insurer doesn't have that kind of organizational agility.
- 1:54What's realistic?
- 1:56Start with one business unit where the leader is already a believer.
- 2:00Don't try to convert the whole organization.
- 2:02Pick a P&L owner who wants to use data as a competitive weapon.
- 2:06Give them a dedicated data team with a direct line into their decisions and let the results speak.
- 2:12McKinsey's research, and I'll flag this, is vendor-adjacent, so take it with some skepticism.
- 2:18Suggests that pilot units with embedded data teams outperform control groups by 15 to 20%
- 2:24on operational KPIs within 12 months.
- 2:28That kind of visible win is worth more than any all-hands presentation on data strategy.
- 2:33And the CDO's role in that.
- 2:35Because historically, the CDO sits in IT or in a staff function, not close to the P&L.
- 2:41That's exactly where the role has to evolve.
- 2:44The CDO's gaining influence in 2026 are the ones who've repositioned themselves as
- 2:49business partners rather than data custodians.
- 2:52They're sitting in on pricing discussions, supply chain reviews, customer retention calls.
- 2:57They're not presenting data.
- 2:59They're co-owning decisions.
- 3:01The ones still defending governance frameworks in isolation are being quietly sidelined.
- 3:06Final question.
- 3:07One thing a CDO should do differently starting Monday morning.
- 3:11Find the most skeptical senior business leader in your organization
- 3:14and ask them what decision they're about to make.
- 3:17Then show up with something useful for that specific decision.
- 3:21Not a strategy deck.
- 3:22One decision.
- 3:23Build trust at that resolution.
- 3:25To go further, Leaders Insights offers strategic training programs, diagnostic tools,
- 3:30and an executive self-assessment designed for CMOs, CDOs, CFOs, and leaders aiming for
- 3:37board-level performance.
- 3:39Explore everything at MBA-training.com.
- 3:42Thank you for listening to the Leaders Insights podcast.
- 3:45See you tomorrow for another episode.