Rebuild the Rails: Why AI only works when your core does
I recently visited an auto dealership for routine service. The intake process featured impressive AI tools—video inspection with transcripts and highlights. But when it came time to approve the work, the service advisor turned to a monitor running what looked like Windows 98. We're bolting new tech onto old rails.
AI Adoption is Real—But Value is Uneven
The numbers tell a compelling story: 71% of organizations now regularly use generative AI in at least one function, up from 33% in 2023. Yet three in four companies struggle to achieve and scale value from AI initiatives.
The gap doesn't exist in demonstrations. It exists in foundational infrastructure.
The Dealership Lesson: Features vs. Foundations
Core systems—Dealer Management Systems, ATS, CRM, ERP—serve as operational backbones. When these fail, everything else grinds to a halt. The 2024 CDK outage forced dealers nationwide to revert to manual paperwork.
Adding AI features without modernizing core systems creates faster noise rather than genuine value.
The Hidden Tax of Legacy
The cost of maintaining old systems is staggering:
- IT spending on legacy maintenance consumes 70-80% of budgets in some sectors
- Banks spend up to 70% of their IT budget maintaining legacy platforms
- 80% of enterprises changed their mainframe strategy specifically to accommodate AI
Most Transformations Still Fail—Humans, Not Models, are the Blocker
Approximately 70% of transformations miss their goals. The culprit isn't technology—it's change management.
Projects with excellent change management are up to 7× more likely to meet objectives. Yet there's a perception gap: 74% of leaders believe they involve employees in transformation decisions, while only 42% of employees agree.
From "Speed Up the Steps" to "Rebuild the System"
Organizations ask the wrong question: How do we use AI to make the current process faster?
The right question: How do we rearchitect the workflow so AI sits at the core?
A Four-Point Framework
- Start with the system of record. Consolidate the data layer first. Without a single source of truth, AI generates conflicting outputs from conflicting inputs.
- Give AI the initiative—keep humans in command. Let AI handle volume; retain human judgment for context, storytelling, and closure.
- Instrument everything. Real-time metrics—pass-through, time-to-fill, quality-of-hire, DEI mix—piped back weekly enable continuous learning.
- Treat change as a first-class workstream. Executive sponsorship, manager enablement, role-based training, and transparent communication aren't optional. They determine success.
A Practical Rubric: Core, Flow, Value
Before launching any AI initiative, we apply a three-part framework:
- Core: Is the system of record clean, connected, and extensible?
- Flow: Does AI own high-volume steps end-to-end with humans in decisive moments?
- Value: Can lift (speed, quality, cost, experience) be measured weekly and fed back?
If a proposal cannot pass these gates, it doesn't launch.
Bringing It Back to the Service Bay
The AI inspection video at the dealership wasn't the problem. Imagine if the core platform were modern: intake auto-populates from appointment data, approvals happen digitally with one tap, parts and pricing sync in real-time, payment and warranty reconciliation happen automatically.
Same staff. Same customers. Different throughput.
How We Live This at WorkGenius
We've modernized our foundational systems so AI functions as the backbone rather than an add-on. The platform runs on clean, connected data with a single source of truth, keeping recruiters in command where judgment matters.
Bottom Line
AI isn't everything. Core systems + change are the force multiplier. Get those right, and the "wow" moments won't stop at a demo—they'll show up in your throughput, your candidate experience, and your P&L.
Onward,
Daniel
