Roxana Irimia

"she’s the marketing roxstar" | Head of Marketing @ lempire | click FOLLOW to save hours of Marketing, Copywriting and Branding research

Sharing insights on marketing, GTM, product marketing, copywriting, personal branding, and pipeline growth straight from what I build, test, and ship in real life.

No fluff. No BS. Just proven, sharp strategies from 10 years in SaaS - from working in content, product marketing, to leading marketing at lemlist ($0 → $30M ARR, bootstrapped), and now scaling global marketing at Chili Piper ($40M+ ARR).

Head of Marketing

Leading the marketing team at lempire, $20M+ in ARR, $150M valuation, bootstrapped.

Suite of products helping people grow their businesses.

→ Book meetings with 17%+ of your prospects [Powered by AI]
www.lemlist.com

→ Keep your emails away from spam
www.lemwarm.com

→ Your personalized scheduling tool
www.lemcal.com

→ Build your personal brand on LinkedIn in 10min a day
www.taplio.com

→ Build your personal brand on Twitter
www.tweethunter.com

Most marketers are trained to be “easy to work with.”

They accept “a tweak here,” a “suggestion there,” a “quick brainstorm,” a “can we also mention this,” a “what if we rephrase,” a “just one small change.”

They often turn clear ideas into group projects, and marketing ends up being a collage of inputs.

The best marketers I’ve seen know how to say “no.” They protect the strategy, the message, the direction, the vision. 

Not to be difficult, but to keep the work sharp and clear.

Because when everything is made to please everyone, nothing stands out anymore.

VP Marketing at ChiliPiper 🌶️ | ex-Head of Marketing at lemlist | “she’s the marketing roxstar” | FOLLOW to save hours of Marketing, Content, Product Marketing and Branding research

If I can give 1 piece of advice after 10 years in SaaS marketing, it’s this:

don’t build your marketing around channels, build it around problems.

Channels change, algorithms change, formats come and go.

But buyer problems stay the same.

People still need to notice you, trust you, and prove value internally.

Don’t base your strategy on LinkedIn, ads, events, or email. They’re just tools.

Instead, create playbooks that solve 1 clear buyer problem.

Example 👇

If your goal is to drive attention, build a playbook for attention.

1. Define the message and POV you want people to remember
2. List all the channels that can spread that message

If one channel slows down or stops working, you don’t change the message.

You just change where it shows up.

In 2026, marketing is about clarity, speed, and being human.

My bets:

→ 0- or 1-field forms
→ fewer tools, tighter stack
→ emails triggered by real intent
→ short paths to book a meeting
→ orchestration over campaigns
→ direct conversations, not fake automation
→ real insights from people doing the job
→ distribution matters more than content
→ customers as the main marketing channel
→ founders and employees as the main voices
→ one clear ICP, one clear outcome, repeated everywhere
→ raw, clear videos explaining one problem, one solution

What really pushes you forward in your career isn’t talent, it’s your ability to take ownership.

In a world where everyone can do things with AI, your advantage is not what you create.

Your advantage is whether people can trust you to take something messy, turn it into something real, and not drop it when it gets hard.

That’s what makes you hard to replace.

The most dangerous sentence in a company is: “we’ve always done it this way.”

It’s not really about process. It’s fear in disguise.

→ Fear of changing
→ Fear of being wrong
→ Fear of losing what feels safe

The sentence protects comfort, not customers.

It blocks change and keeps things average while the world moves on.

Every real change begins when someone asks: “why are we still doing this?”

This quick exercise will tell you more about your growth than any traffic report or conversion dashboard.

1. Open your CRM
2. Look at your last 50-100 inbound leads
3. For each one, note how long it took your team to reply

Now ask yourself:

→ How many were contacted in under 5 minutes?
→ How many were contacted instantly?

That number shows you how many opportunities disappear before sales even get a chance.

You can launch more ads.
Rewrite your landing pages to drive more traffic.
Launch new campaigns to generate more leads.

Or you can answer people while they’re still warm and show interest.

Most teams choose the first. The best teams fix the second.

In 2026, speed = growth.

P.S. In the comments you'll find a simple way to calculate how much pipeline you’re losing based on how fast you reply

This is how trust can 2x your pipeline.

(not brand trust, but decision trust)

Pipeline doesn’t break randomly. It breaks at very specific moments, for very specific reasons.

1. Pick the exact moment you’re losing pipeline

Not “we need more leads,” but the real drop-off.

E.g., low demo conversion, high no-shows, stalled evals, deals lost to “do nothing,” or deals stuck in legal for weeks.

If you can’t name the moment when pipeline leaks, you’re guessing, not fixing.

2. Name the fear behind that moment

No-shows = “this probably won’t be worth my time”
Stalled evals = “I’m not confident this will work for us”
Losses to competitors = “they feel safer to buy”
Losses to do-nothing = “status quo feels less risky than change”

Until you name the fear, everything you ship is generic.

Create 1 trust asset per objection, not a content campaign.

If “worth my time,” build a 60-second pre-demo page that shows exactly what they’ll see, who it’s for, and the outcome they should expect.

If “will it work,” create a proof pack with 3 customer examples from the same ICP, same use case, real numbers, and setup details.

If “safe to buy,” publish a comparison page that clearly states your trade-offs and when you’re not a fit.

If “status quo,” build a simple cost-of-delay calculator with 3 inputs so the buyer can justify urgency internally.

Inject those assets into the funnel, not your content calendar.

The pre-demo email gets the pre-demo page.
The first sales follow-up gets the proof pack.
The eval stage gets the comparison page + clear implementation plan.
Procurement gets security, data handling + “what happens in week 1-4” doc.
Trust works when it shows up exactly when doubt shows up.

Repeat this weekly, 1 bottleneck at a time.