From Predictions to Engagement: Missing the Real Problem Again

After learning from Extenure, I thought I'd figured it out.
Retention wasn't the problem. Engagement was.

If we could measure how engaged employees felt, we could help managers take action before people started looking for the exit.

So we built engagement analytics: pulse surveys, sentiment tracking, engagement scores by team and location.
Beautiful dashboards. Real-time alerts. Trend analysis.

And once again, nobody cared.

The Pattern I Didn't See

I was so focused on finding the "right" problem that I missed the bigger issue: I kept building the same solution.

Extenure: analytics dashboard.
Engagement platform: analytics dashboard.

Different data, same format. Same promise: "Know what's happening so you can fix it."

I was competing with 50 other dashboards managers already had: scheduling, payroll, inventory, sales, labor costs, compliance reports. One more chart wasn't going to change their day.

The Engagement Survey Graveyard

I pitched to an HR director at a multi-location restaurant chain. She listened politely, then said:

"We already do quarterly engagement surveys. The scores sit in a spreadsheet. Nobody looks at them unless something's on fire."

I tried to explain how our platform was different: real-time, actionable, manager-focused.

She cut me off: "That's what the last three vendors said too."

The Abstraction Problem

Engagement is too abstract to sell.
What does a "67% engagement score" mean? What do you do about it?

I thought we'd solved this with our action recommendations:

• "Schedule more one-on-ones"
• "Recognize top performers publicly"
• "Review workload distribution"

But these felt generic because they were. Real engagement issues are messy and specific: personality conflicts, unfair scheduling, broken equipment, unclear expectations.

A dashboard can't fix those. A conversation can.

What I Should Have Built

Instead of measuring engagement, we should have built tools that created it:

• Structured check-in templates that managers could actually use
• Recognition systems that felt genuine, not automated
• Career pathway visualizations employees could explore themselves
• Feedback mechanisms that led to real changes, not just data collection

But I was still in love with analytics. I wanted elegant data models, not messy human interactions.

The Dashboard Grind

I spent weeks perfecting the engagement dashboard:
Drag-and-drop widgets. Custom date ranges. Exportable reports. Mobile responsive.

I thought polish would matter.
I thought if it looked professional enough, felt intuitive enough, loaded fast enough, people would use it.

They didn't. Because the problem wasn't that engagement tools were poorly designed.
The problem was that engagement was already being tracked everywhere, and nothing was changing.

The Sales Cycle That Never Closed

Every pitch followed the same pattern:

Week 1: "This is interesting, let me think about it."

Week 2: "Can you send me pricing?"

Week 3: "I need to talk to my team."

Week 4: Ghost.

I told myself they just needed more education, more case studies, more proof points.

The truth: they didn't have the problem I was solving.
Or they did, but a dashboard wasn't going to fix it.

What I Learned (This Time)

So we pivoted again.

This time, to something tangible: health code violations, safety issues, cleanliness standards.
Things you could see, point to, and fix.

Finally, a real problem. Or so I thought.