Ep 298 – Amin Palizban (CEO, Topicflow)

1. Managers are at “code red” — and it’s getting worse

We’re at a breaking point. Managers — the single most important lever for performance, retention, and engagement — are now the most disengaged group in many companies.

Why? It’s a perfect storm. Layoffs flattened orgs, increasing span of control. AI boosted individual productivity, which means more output to review. And managers are now expected to lead change, adopt new tools, and coach at a higher level — all at once.

“It’s crazy… they have less resources, more change management, and more direct reports.”

The result: managers are overwhelmed, reactive, and often stuck in approval loops instead of actually leading. If companies don’t address this directly, everything else — performance, culture, retention — breaks downstream.

2. The real problem: we promote great ICs into unprepared managers

This isn’t new — but it’s still happening everywhere.

High-performing individual contributors get promoted into management roles without being trained (or even told) what the job actually is. And the skills required are deeply unnatural: setting expectations, giving tough feedback, coaching consistently.

“A good IC doesn’t necessarily become a good manager.”

Most managers can’t clearly articulate the core dimensions of their role:

• Setting expectations
• Giving constructive feedback
• Recognizing performance
• Removing roadblocks
• Creating psychological safety

And traditional training isn’t solving it — because it’s too slow and too disconnected from real work. By the time a manager needs the skill, the training is long forgotten.

3. Performance reviews don’t improve performance — coaching does

This is one of the most important mindset shifts.

Performance reviews are evaluations. They tell you how someone did. But they don’t actually make someone better.

“All the coaching and feedback… is what has driven you to become better. The evaluation just tells you how well you did.”

The problem is most companies try to combine everything into one bloated process:

• Feedback
• Coaching
• Goal setting
• Evaluation

And it doesn’t work. When someone is being evaluated, they’re not listening to feedback — they’re thinking about their rating.

The better model:

• Continuous, in-the-flow coaching (weekly)
• Separate, clean evaluation moments (quarterly or annually)

If feedback is happening consistently, there are no surprises — and reviews become faster, fairer, and far less emotional.

4. Managers don’t need more tools — they need less “mental load”

The biggest unlock isn’t adding more systems. It’s removing the cognitive burden managers carry every day.

Right now, managers are:

• Trying to remember conversations from weeks ago
• Tracking follow-ups manually
• Reconstructing performance from memory

And that’s where everything breaks.

“Managers are managing from memory… and memory is not very good.”

The future isn’t dashboards or forms — it’s AI acting as a real-time assistant:

• Reminding you what to follow up on
• Summarizing what each employee worked on
• Highlighting missed coaching opportunities
• Nudging you before key conversations

“In your last one-on-one, you mentioned this… follow up.”

This is what real support looks like: not more work — but less to remember.

5. The best performance data already exists — we’re just not using it

Most companies think they need new systems to improve performance management.

They don’t.

The data is already there:

• Meetings and conversations
• Slack and collaboration tools
• CRM, Jira, Zendesk, etc.

The problem is it’s fragmented and inaccessible.

“If you don’t have that, each manager has to go… and remember what this person has done. That’s where bias comes in.”

The first step isn’t transformation — it’s aggregation.

Start by:

1. Capturing conversational data (this is the highest signal)
2. Pulling in work data from core tools
3. Using AI to summarize it into a “work history”

That alone dramatically improves review quality, reduces bias, and saves hours of time.

6. The future is “ambient intelligence” not asking, but knowing

One of the most compelling ideas from the conversation: managers shouldn’t have to ask for insights.

The system should just know.

“Ambient intelligence… giving you these insights without you asking for it.”

Imagine:

• A project finishes → AI suggests collecting feedback from collaborators
• A one-on-one is coming up → AI preps your agenda with key follow-ups
• An employee goes above and beyond → AI prompts recognition

And critically, it happens in the flow of work (Slack, meetings, etc.) — not in another tool.

This is how you move from reactive management to proactive coaching — without adding friction.

Final takeaway: start simple — and build from there

It’s tempting to try to fix everything at once. Don’t.

“Focus on one problem… and not trying to solve five or six problems at the same time.”

The practical starting point:

1. Assess your managers (are they actually equipped for the role?)
2. Fix the basics (expectations, feedback, recognition)
3. Start capturing and summarizing work data
4. Layer in AI support gradually

This isn’t a massive transformation overnight. It’s a series of small, compounding upgrades that unlock a completely different way of managing.

And the companies that figure this out first?

They’re going to have a massive advantage.