Micro-Interview Series

Episode
#15

Transforming outreach through AI-driven personalization and smart automation.

Micro-Interview Series •
317
 words
Klemen Hrovat
Co-Founder, Chief Revenue Officer
Sellestial

Introduction

Klemen is building Sellestial—an AI-powered HubSpot platform that helps revenue teams turn noisy leads into booked meetings, faster.

What started with personalized email outreach has grown into a full GTM automation layer: contact enrichment, data firewall, and contextual messaging fully integrated inside HubSpot.

In this interview, Klemen breaks down:

• What true personalization actually looks like

• How to keep AI outreach human

• And the most common mistakes RevOps teams make with automation.

A sharp, practical conversation from someone building at the frontier of RevOps + AI.

Everyone’s talking about “personalization” but few agree on what that actually means. In your view, what separates true, value-adding personalization from surface-level tricks like {first_name} and {company_name}?

True personalization (we call it individualization) uses all the relevant context that changes the angle of the message: a quote from their CEO on a recent podcast, insights from their recent job postings, activity timeline from that contact and company in HubSpot, or a fact from their LinkedIn profile.

It proves you’ve done your homework and research on them to understand what matter to them and how you can relate to what they care about. Tokens introduce you; relevance starts a conversation.

We’re entering an era where AI can write faster than we can think. But that also creates noise. What’s one principle you follow to make sure AI-written messages still feel human, relevant and get replies?

Emails should be relevant, not only personalized. With the system and AI we mimic a normal human process in an ideal scenario - what data you would look at.

For example, one would look at recent communication and engagement, their LinkedIn profile (job description and recent posts) and combine that with their understanding of their audience.

So we tailor the message for every recipient; each email is uniquely written for that person.

You’ve helped teams scale personalized outreach in HubSpot without losing authenticity. What’s one mistake you see RevOps teams make when trying to automate too muchand how can they fix it?

From what I’ve been seeing and experiencing recently, several mistakes have occurred.

To pick 2, bad data hygiene (incomplete, outdated, inconsistent) and trying to simplify it too much.

For example, not understanding how to prompt, how to bring all the context in the right way to AI, not iterating and navigating the system enough, and using the wrong LLM models for those kinds of tasks.

Reflections

I really enjoyed this one.

Klemen has a rare way of explaining a noisy topic — AI personalization — in simple, useful terms. What stood out to me most is his practical framing: personalization isn’t just about adding first names or roles. It’s about tailoring the angle of the message using contextual data that actually matters to the recipient. That’s the bar.

What I also found powerful is his point on relevance vs personalization. It’s easy to make AI write 1,000 messages that sound personal. It’s much harder to make those messages relevant and that’s exactly where Sellestial adds real value.

And for anyone building outreach in HubSpot (especially in commercial teams at football clubs), Klemen’s thoughts are a helpful reminder:
Automation without data hygiene is just scaling dysfunction.

Clear, sharp, and very actionable. Thanks again, Klemen.

Enjoyed this? Follow us fore more sharp takes.

Klemen Hrovat
Co-Founder, Chief Revenue Officer
Follow
Klemen Hrovat
Sellestial
Matthias Werner
👉 The CRM guy for football clubs.
Follow me

Want to get featured?

I’m curating micro-interviews with smart voices shaping football and business — think about data, strategy, and systems.

If you're doing sharp work in football, analytics, CRM, or tech-enabled ops, I’d love to hear your take.

Just drop me a message!

Looking forward,
Matthias 👋
Let's connect: