I was interviewing a VP of Operations. He was telling me about a property he had turned around. Collections were a mess, expenses were leaking, occupancy was soft, the team was demoralized. He told me he had restructured the leasing team, cut three line items in the operating budget, and brought collections current in ninety days. Sounded right. Sounded experienced.

Then I asked for a specific example.

What was the first thing you fixed? Who did you fix it with? What did the rent roll look like the month before, and what did it look like ninety days later? Who pushed back? What did you try that didn’t work?

The story collapsed. He hadn’t lied. He had been in the room when it happened, not at the table where the work got done. He could recite the order of operations. He couldn’t tell me why each step worked at that property, on that team, in that submarket. The vocabulary was real. The experience was secondhand.

An empty chair experience.

I didn’t hire him. I almost did.

Why specificity is the sharpest filter you own

“Give me an example” costs you ninety seconds. It pays for itself the first time it saves you from a bad hire.

The reason it works is mechanical. Memory has a grain. People who did the work remember the maintenance tech who flagged the leak before the inspector did. The line item on the T12 that didn’t match the bank statement. The Tuesday the lender called and they had to decide in twenty minutes. Real memories come with names, dollar amounts, dates, and arguments they lost.

The fingerprint of someone who did the work is in the nouns. Carlos. Unit 312. The August roll. The HVAC bid that came back at $12,000 when we’d budgeted $8,000. The fingerprint of a watcher is in the categories. The team. The property. Operations. The business plan. Listen for the grammar of the answer. The grain tells you everything.

The second question doubles the filter strength

Anyone smart enough to interview for a VP role has a polished story or two. The trap is asking once and being satisfied.

Ask twice. Then ask a third time. People who did the work have a deep bench because they’ve lived through dozens of these moments. Fakers have one, maybe two, rehearsed to a shine.

And the question that breaks the act faster than any other: what did you try that didn’t work?

Anyone can tell you the polished version of a win. Only people who did the work remember the dead ends. The bid that came in late. The leasing manager you promoted who wasn’t ready. The expense category you cut too aggressively and had to walk back. Failures live in the memory of people who were responsible for them. Watchers don’t carry them, because the failures weren’t theirs to begin with.

When you ask for a specific failure and the candidate hesitates, that long hesitation is the most diagnostic moment in the interview. They’re not searching for the story. They’re searching for a story that won’t damage them.

The industry rewards proximity and calls it experience

Why are there so many half-knowers in senior roles in the first place?

Because the comp structure can’t tell the difference between contribution and proximity.

Bonuses are tied to property performance. Everyone on the multifamily org chart gets paid when the asset hits its numbers, whether they touched the decision or not. Promotions follow tenure and visibility, not causation. References get returned by the people who hired you, not the line staff who watched you work. LP capital flows to sponsors whose track records were built in a market that paid for showing up.

Show me the comp plan and I’ll show you the behavior.

The VP I interviewed wasn’t a con artist. He was the right answer to a comp plan that paid for being in the room. The system paid him as if he ran the turnaround, and after enough of those paydays, the line between watching and doing starts to blur even in his own head.

That’s the dead tree principle in management form. Ten years of experience that’s really one year he watched ten times.

Where else this tool earns its keep

The example test is the cheapest filter in any room where someone is making a claim.

Vendor case studies. Every proptech demo includes “we helped Property X increase NOI 18%.” Ask for the specific decision their software prompted that the operator wouldn’t have made anyway. Nine out of ten stories collapse at the first follow-up, because the software correlates with the result but didn’t cause it.

AI fluency. The half-user says “we use AI for underwriting.” The real user can describe what the tool gets wrong and what they built around it. I ran a T12 extraction system in production for months before I realized it was sanding rough edges off small expense categories in a way that smoothed the underwrite. I pulled it back and put a human eye on those lines. The example that proves you’re using the tools isn’t the part that worked. It’s the part that didn’t.

One last thing

Run this test on yourself.

Pick a deal you say you executed well. Pick a turnaround you say you led. Now produce the specific examples that would survive someone like me asking three follow-up questions. The named people. The exact decisions. The things you tried that didn’t work.

If the examples are there, you did the work. If they’re not, you had an empty chair experience.

The question is too useful to only ask other people.