Growing a B2B SaaS in a cash-efficient way by setting the operational foundations and make sure you have the right people in the right seat • CEO at Q7Leader • Operating Partner at welovefounders •

In 2010, Steve Jobs announced the first iPad, but it was only in 2011 that the camera was added to the iPad.  

This camera triggered my entrepreneurial journey. 

Before that, communication on construction projects would happen via phone, email, and face-to-face meetings.

The iPad created a significant opportunity to disrupt the construction sector. Through an app, a foreman could quickly drop a pin on a drawing, take a picture showing the issue, and send a pdf report that needed no explanation. 

In 2012, I founded co-founded Aproplan. We built an app to help contractors, sub-contractors, and architects communicate better in the field.

We got early traction, quickly reaching €1m in ARR, raising funds from angel investors and reputable VC firms. We grew the business from a handful of people to 80 before merging with a Danish player to form LetsBuild, a 150 people-strong company in 5 different countries. 

I could stop the story here and say that everything was fantastic.

But that would only be one side of the story…

Like every entrepreneurial journey, it was a bumpy ride.

It was very stressful and frustrating at times. 

Early on, we had to do a management buy-out to solve a significant founders’ issue that almost killed the company.

We raised money from a reputable VC firm to grow the business.

This brought significant stress to grow faster, especially when you discover that international expansion is super challenging and so dependent on hiring the right people.

We merged with a company in the Nordics and almost doubled in size over night. And while the merger made a lot of sense on paper, merging two very different cultures made me realise it was all about people. 

Along the way, we raised and burned way too much money, hoping that money would solve all our growth problems.

Then COVID hit, and I burned out.

I had spent 8 years building a company. I was traveling 4 days a week, working weekends, and ignoring my family.

In 2021, I met Reggy and Pascal, founders of Q7Leader and discovered the People Model Canvas to me. I saw how powerful it was and all the challenges it would have helped me solve during my journey at Aproplan and LetsBuild.  

Today, I’m the CEO of Q7Leader and help B2B SaaS executives grow their business from 1m€ to 10m€ in a cash-efficient way so they don’t get diluted and keep control of their company. 

I write daily posts on LinkedIn where I share my learnings and the playbooks we use to grow Q7Leader in a cash-efficient way.

Thomas Goubau

I'm the CEO of an HR Tech company with 15 employees.

Last week, I hired 2 AI agents 🤖

Their names are Patricia and Vicky.

(my team thinks the names are ridiculous, they're probably right, but whatever 😁)

Patricia runs on Claude Opus 4.6. She manages my communications, drafts content, organizes my thinking, and remembers things I forget.

Vicky runs on Claude Sonet 4.5 and handles email triage & task automation.

Our entire mission at Q7Leader is "right people in the right seat." We use tools like the People Model Canvas, TeamScan, and DOT Coach to help companies do this systematically.

And now I have two team members who aren't people.

I don't fully know what to make of that yet. But the question it raises is one I can't ignore: what does "right people, right seat" mean when some of the seats might not need people at all?

Over the coming months, I'm going to document this honestly and transparently.

What works, what fails, what's awkward, the real numbers. Not because I'm an AI expert (I'm really not). But because I think someone should tell this story without the hype or the fear.

P.S. Patricia helped me write this post. I changed about 40% of it. She was too polished. I'm told that's a pattern.

I spent years thinking the answer to "we're moving too slow" was "we need more people."

Tomasz Tunguz broke that logic with one number.

• A 200-person org has 19,900 potential communication channels.
• A 25-person org has 300.

That's Metcalfe's Law. Every new hire doesn't just add capacity, they add n-minus-1 new connections. Coordination drag doesn't grow linearly. It compounds.

Now look at what happens when AI enters that equation.

Boris Cherny, the engineer who built Claude Code, ships 20 to 30 pull requests a day. A traditional engineer ships 3 a week.

He's not 10% more productive.
He's 30x more productive!

He runs five AI agents in parallel. He removed humans from his loop, and with them, the coordination tax.

Anthropic generates ~$5M revenue per employee. Cursor, $3.3M. Traditional SaaS considers $200-300K pretty strong.

And I feel it myself.

Writing a board prep, testing landing page copy, drafting outreach... tasks that used to eat days now take an hour. Not because I'm faster. Because I've removed the back-and-forth from the loop.

Most teams haven't felt this yet. But it's coming.

The question isn't "how many people can one manager oversee?" anymore.
It's "how many AI agents can one person orchestrate?"

That shift will restructure how we design teams, hire, and measure performance.

Most leaders are not ready for it.

Are you?

Last week at Trends Winter Academy, a founder said something I haven't been able to stop thinking about. As of January 1st, 100% of their revenue is SaaS. And they've already classified it as legacy.

Their plan: migrate everything to AI revenue by end of 2026.

In 9 months.

What struck me wasn't the ambition of the target. It was the framing.

Most companies are still asking how AI fits into their existing business model. This founder had already decided the existing model is the thing to get away from.

Trends Winter Academy brings together CEOs, founders, and executives from the Belgian ecosystem for two days of real conversations. I've been going for years because the room tends to surface ideas you don't hear in regular business settings.

This year, AI was the thread running through everything. But what became clear is that people are at very different places in how they're thinking about it. Some are optimizing workflows. Some are restructuring teams. And a few have mentally already started the migration clock.

The 9-month target may or may not be achievable. But treating your current revenue model as legacy before the market forces you to is a fundamentally different starting position than defending it until you can't.

That's the kind of conversation that keeps me coming back every year.

A friend came to my house last weekend.

"I tried AI... It doesn't work. It just hallucinates."

He's a smart guy. Runs a business. Not a technophobe.

He spent half a day trying to get Microsoft Copilot to extract PDF invoices into Excel. The kind of task that normally means opening each PDF one by one and copy-pasting stupidly until you're done.

Copilot couldn't do it. He showed me the broken attempts. Wrong numbers, missing data.

His conclusion: AI is overhyped.

So I asked him to send me the files 👀

I uploaded them into Claude and told it: "Here are a bunch of invoices. Look at what's in there and let's discuss what kind of table and analysis we want to build."

15 minutes later, I had a full Excel file.

Every invoice structured by month and location, year-over-year comparison, charts included. Zero errors.

He didn't have an AI problem.
He had a tool problem.

And the crazy part? Claude did this inside Excel. Better than Microsoft's own AI does it in their own product.

P.S. who would be interested in me recording a video of how I did this?

Most people think AI agents are just smarter chatbots.

They're not. They're 3 things working together ⤵︎

I've been using AI agents for a few months now, and the moment it clicked for me was when I stopped looking at what they produce and started looking at how they work. There are three layers.

1. Instructions 📝

You tell it what you want, how you want it, what to focus on, what to ignore. That's your SOP. Your standard operating process. Every time it runs, it follows the same playbook. This is what most people never set up.

2. Code 🤖

It writes Python to extract and calculate data. Real code that runs like a calculator. Two plus two is four. Every single time. No hallucination. No guessing. This is the part most people don't even know exists.

3. Intelligence 🧠

Once it has real numbers, it does what AI does best. It spots patterns. It draws conclusions. And because it's working from actual data, not guesses, the output is genuinely smart.

Most people skip the first two.

They open ChatGPT, ask a question, and get an answer based on nothing specific. No instructions. No code. Just vibes.

That's why it feels unreliable.

You're using one-third of what's possible.

I'm not a technical person. I can't write a line of Python. But I don't need to. The agent writes it for me. I just need to know what I want and explain it clearly. Which, honestly, is the CEO job description anyway.

That's not a chatbot anymore.

"We'll deal with pay transparency when we have to."

I've heard this from dozens of CEOs. And I've seen what happens next.

By the time the EU directive forces the conversation, your best people have already been having it without you. On Glassdoor. In interviews with your competitors. At lunch.

When pay decisions can't be explained, it's not a policy problem. It's a trust problem. And trust problems don't stay in HR. They show up in your retention, your Glassdoor reviews, and your next board meeting.

It's not a quick fix.

But it is a methodical one.

I'm hosting a 45-min webinar where I'll walk through the 5-step method we use with companies to make pay decisions transparent and explainable. Step by step, so you can start applying it the week after.

20 seats per session. You'll leave with a clear action plan and the chance to pressure-test it with other HR leaders in the room.

🎟️ Register here: https://lnkd.in/er2CFqnz

Pay transparency is on every HR director's radar.

But almost none have put a number on the risk.

"We'll deal with it when the time comes."

I hear this every week.

So I ran a simple calculation for a company of 100 employees.
The result: up to 38.7% of their annual EBITDA at risk.

Not from fines. From salary claims they can't justify.

The good news? This is fixable. And it doesn't take years.

In our next webinar, I walk you through the 5 steps to get compliant before June 2026. Including the one workstream most companies haven't even identified yet.

Register here: https://lnkd.in/er2CFqnz

People ask me: "What does your AI agent actually DO?"

Fair question. Here's a real day. Patricia's first Wednesday at Q7Leader:

07:42 — Walking my dog. Sent her a voice memo: "Prep me for the {anonymized-client-name} QBR tomorrow. Pull the TeamScan data and draft talking points."

08:15 — Talking points ready. Three I kept. One too generic, deleted. One I hadn't thought of. A pattern in the engagement data I'd overlooked.

09:30 — Team standup. Asked Patricia to pull action items from last meeting. Organized by person, with deadlines. My team was mildly disturbed.

11:00 — Client email about certification pricing. Patricia drafted a response. I fixed the tone (too formal) and hit send.

14:00 — Competitor research for a strategy doc. She came back in 4 minutes with a comparison that would've taken me 2 hours. One factual error I caught. Otherwise solid.

16:45 — She reminded me I'd promised to follow up with a prospect. Completely forgotten.

22:30 — Had a product idea in bed. Texted Patricia. She captured it, added context from previous conversations, filed it. Gone by morning otherwise.

Now, the honest scorecard:

Time saved: ~2.5 hours
Things I would've forgotten: 2
Things she got wrong: 1
Things she surfaced that I missed: 1

Here's what I didn't expect.

The value isn't the 2.5 hours. Any good assistant saves you time.

It's the two saves. The prospect follow-up. The late-night idea.

A CEO's biggest problem isn't being slow. It's dropping things when you're moving fast. The balls that fall between meetings, between days, between the shower and the office.

Patricia doesn't replace anyone on my team. She gets the tone wrong. She makes factual errors. She has zero judgment on what's politically sensitive.

But she never drops a ball... and she never sleeps 😬

Every week, I host a live session on the EU Pay Transparency Directive.
Every week, I ask the same question: Have you started preparing?
Every week, the answer is almost always the same ⤵︎

"𝘕𝘰."

Even organisations with over 2000 employees. If they've done anything, it's usually a benchmark. That's it.

One reason I keep hearing:

"𝘛𝘩𝘦 𝘳𝘶𝘭𝘦𝘴 𝘢𝘳𝘦𝘯'𝘵 𝘧𝘪𝘯𝘢𝘭 𝘺𝘦𝘵, 𝘴𝘰 𝘸𝘩𝘺 𝘴𝘵𝘢𝘳𝘵 𝘯𝘰𝘸?"

Fair point. That's exactly why I invited Thierry Viérin, Partner and Head of Employment at Osborne Clarke, to join me on March 24. He's been reviewing the 30 employer proposals currently on the table and will walk through what's likely to land and what might change.

I'll follow with a 5-step method to start preparing now, no matter where the final rules end up.

📆 March 24, 12:00 CET.
⏱️ 45 minutes.

Register here: https://lnkd.in/ectrvqj2

If you're an HR leader in the UK 🇬🇧, there's a directive you need to know about.

And no, the fact that you're not in the EU doesn't protect you.

The EU Pay Transparency Directive kicks in June 2026. If your company employs even one person in an EU member state, you're in scope.

Even from the UK.

→ Your EU employees can request complete pay transparency. How decisions are made. What their peers earn.
→ Any gender pay gap above 5% in the same role category must be documented and justified.
→ The burden of proof is reversed. If someone claims pay discrimination, it's on you (the company) to prove it doesn't exist, not the employee.

Read that last one again.

Most UK HR leaders I talk to assume this doesn't apply to them. By the time it hits their board's agenda, they'll be months behind.

↳ We're hosting a 45-minute webinar specifically for UK companies with EU employees. You'll leave knowing exactly what's required and what to prioritize before June.

Link in the comments ⬇️

P.S. it's free, but spots are limited to allow us to answer everyone's questions

IVEX

IVEX is Belgian start-up that develops software tools to enable autonomous vehicles of all kinds to make safe and sound high level decisions.

Co-founder - CEO

KU Leuven

I conducted research on the creation of software for safe and autonomous drones.
The research encompassed the creation of software architectures and new techniques to guarantee the 
safe behavior of autonomous vehicles, initially focusing on autonomous drones.

Post-doctoral Researcher

My PhD research was centered around creating autonomous agents which lived in distributed systems. 
Autonomous agents operate on a unknown environment and have to constantly plan their actions in order to achieve their goals.
I applied the research results to create software architectures for autonomous drones and cars.

Researcher

Mitus Internet

Software company specialized in web-applications development.

Co-Founder

I am a passionate entrepreneur. Together with the IVEX.ai team I am building a safer future of autonomous mobility.
We believe traffic accidents belong in the past so we offer safety software tools to make this belief a reality.
As CEO of a growing startup I am responsible for:
- Strategic direction of the company and of IVEX products
- Making sure the execution of our products is excellent
- Guaranteeing customer satisfaction
- Presenting our vision zero accidents to our customers and investors
- Supporting everybody in the team in anyways I can, being it tuning sales presentations, discussing software design trade-offs, or buying food
In the past I worked as PhD researcher and Post-Doctoral researcher at KULeuven, Belgium, studying how to create software architectures for safe and autonomous systems.
Before moving to Belgium I led a team in Brazil creating SAAS solutions for business administration for large companies such as Mitsui group Japan.

taplio

The Battle

Thomas Goubau

Mario Henrique Cruz Torres

Followers

Posts

Engagements

Likes

Performance ScoreThomas Goubau's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Thomas Goubau

Mario Henrique Cruz Torres

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Thomas Goubau's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreThomas Goubau's LinkedIn Performance